Citation
Multiple-region equilibrium world trade model : the orange industry

Material Information

Title:
Multiple-region equilibrium world trade model : the orange industry
Creator:
Brenes, Esteban R., 1957-
Publication Date:
Language:
English
Physical Description:
xvi, 348 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Economic models ( jstor )
Elasticity of demand ( jstor )
Import prices ( jstor )
Imports ( jstor )
Market demand ( jstor )
Market prices ( jstor )
Mathematical variables ( jstor )
Modeling ( jstor )
Statistics ( jstor )
Trade regionalization ( jstor )
Consumption (Economics) -- Mathematical models ( lcsh )
International trade -- Mathematical models ( lcsh )
Orange industry -- Mathematical models ( lcsh )
City of Gainesville ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1992.
Bibliography:
Includes bibliographical references (leaves 339-347).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Esteban R. Brenes.

Record Information

Source Institution:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
027910122 ( ALEPH )
26699670 ( OCLC )

Downloads

This item has the following downloads:


Full Text












MULTIPLE-REGION EQUILIBRIUM WORLD TRADE
MODEL: THE ORANGE INDUSTRY
















By

ESTEBAN R. BRENES


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

1992


UNIVERSITY OF FLORIDA LIBRARIES
















ACKNOWLEDGEMENTS


I wish to give particular thanks to Dr. Ronald W. Ward, chairman of my doctoral committee, for the guidance, assistance and constructive criticism given to me during the preparation of this dissertation. Similarly, I would like to express my gratitude to the other members of my advisory committee--Drs. Kenneth R. Tefertiller, James L. Seale, Jr., Max R. Langham, and Terry L. McCoy, for their advice and

support in the preparation of this document. Special thanks are given to Sharon Bullivant and Melissa Bracewell in helping me put the dissertation in its final form for the Graduate School. Special thanks are also given to Bill Messina for his support and encouragement during the preparation of the document.

I also appreciate the funding I received from the the United States

government through the Fulbright Scholarship Program. I want to thank my friends from Liberty Church of Gainesville for their support,

understanding, and prayers. I give special thanks to my parents for they taught me to appreciate education and always gave me the necessary support to accomplish my academic achievements.

Finally, to my wife and son, I owe the most gratitude. They were

patient, understanding, and supportive all these years, and thanks to them I finished this document. So, I would like to dedicate this dissertation to my wife and son.


ii

















TABLE OF CONTENTS


ACKNOWLEDGEMENTS LIST OF TABLES LIST OF FIGURES ABSTRACT . . CHAPTER

1 INTER


. ii vi x xv


NATIONAL TRADE AND AGRICULTURAL PRODUCTS:


THE ORANGE INDUSTRY .


1


Introduction . . .
International Trade of Agricultural Pro Problem and Objectives . . . .
Scope . . . . . . . . .
Methodology . . . . . . .
Overview . . . . . . . .

FRESH ORANGE WORLD PRODUCTION AND TRADE . .

Introduction . . . . . . .
Production Analysis . . . . .
Trade Flow Analysis . . . . .
Partner Region Perspective . .
Region Perspective . . . .
Conclusions . . . . . . .

LITERATURE REVIEW . . . . . . .

International Agricultural Trade Models Trade Models: The Orange Industry .

WORLD FRESH ORANGE TRADE MODEL . . . .

Introduction . . . . . . .
Fresh Orange Trade Model . . . .
Demand Side . . . . . .
Supply Side . . . . . .
Export Supply Equations . . .
Equilibrium Conditions . . .

iii


1
. 3


ducts


. . . 23 . . . 26 . . . 27 . . . 28

. . . 29

. . . 29 . . . 29 . . . 34 . . . 39 . . . 50 . . . 56

. . . 58

. . . 58 . . . 68

. . . 75

. . . 75 . . . 77 . . . 77 . . . 83 . . . 84 . . . 85


2


3


4


. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .


. . .

. . .

. . .

. . .









Price Linkage Equations . .
CRES Model Restrictions . .
Model Implications . . . Trade Data Base . . . . . .


ECONOMETRIC PROCEDURE AND EMPIRICAL RESULTS . . .

Introduction . . . . . . . . .
Econometric and Estimation Procedure . . .
Empirical Results and Implications . . .
Empirical Results . . . . . .
Empirical Results: Graphical Analysis
Empirical Results: Statistical Analysis
Empirical Results: Economic Analysis .


. 98

. 98 . 98 107 107 134 190 195


Application for Policy Purposes . Conclusion . . . . . . . .


ECONOMIC IMPLICATIONS FROM SENSITIVITY ANALYSIS .


Introduction . . . . Sensitivity Analysis Procedure Rationale for Region, Equation, Variable Selections . . Sensitivity Analysis . . .
Total Market Demands .
Export Supplies . . .
Product Demands . . .
Summary . . . . Conclusions . . . . .


7


APPENDIX A APPENDIX B APPENDIX C APPENDIX D APPENDIX E APPENDIX F


and


SUMMARY AND CONCLUSIONS . . . . . . . .

Introduction . . . . . . . . .
Data Limitations . . . . . . . .
Estimation and Sensitivity Analysis Difficulties Performance of the Model and Results . . .
Contributions to Agricultural Economics
Research . . . . . . . . . .
Further Research . . . . . . . .

COUNTRY COMPOSITION OF THE REGIONS . . . . . DERIVATION OF THE PRODUCT DEMAND EQUATIONS . . . PROCEDURE TO OBTAIN REGIONAL CPIs . . . . . .

PROCESSED ORANGE UTILIZATION . . . . . .

TARIFF DATA . . . . . . . . . . .

PRINCIPAL COMPONENT PROCEDURE AND PROGRAM . . . .


iv


5


. 85 88 92 93


6


. . 222 . . 225


226 226 228 232 235 237
249 253 290
294 296 296 298 299 300 302 303 305 307 309 310 311 312


. . . .
. . . .
. . . .
. . . .









APPENDIX G ESTIMATION PROGRAM . . . . . . . . 313

APPENDIX H EMPIRICAL RESULTS: PRODUCT DEMAND AND CIF PRICE
LINKAGE EQUATIONS STATISTICS . . . . . . 321

APPENDIX I SENSITIVITY ANALYSIS PROGRAM . . . . . . 326

APPENDIX J INDICES OBTAINED FROM THE SENSITIVITY ANALYSIS . . 333 REFERENCES . . . . . . . . . . . . . . 339

BIOGRAPHICAL SKETCH . . . . . . . . . . . . 348


V
















LIST OF TABLES


Table

1.1

1.2 1.3

1.4 1.5 1.6 1.7 1.8 1.9


1.10 1.11


1.12 1.13


1.14 1.15


2.1 2.2


World Orange Production by Region . . . World Fresh Utilization by Region . . . World Processed Production by Region . . World Fresh Orange Exports by Region . . World Processed Orange Exports by Region . World Fresh Orange Imports by Region . . World Processed Orange Imports by Region . World Fresh Orange Export Quantities by Region (Excluding Intraregional Trade) . World Fresh Orange Export Values by Region (Excluding Intraregional Trade) . . . . World Processed Orange Export Quantities by Region (Excluding Intraregional Trade) . World Processed Orange Export Values by Region (Excluding Intraregional Trade) . . . . World Fresh Orange Import Quantities by Region (Excluding Intraregional Trade) . . . . World Fresh Orange Import Values by Region (Excluding Intraregional Trade) . . . . World Processed Orange Import Quantities by Region (Excluding Intraregional Trade) . . World Processed Orange Import Values by Region (Excluding Intraregional Trade) . . . . World Orange Production by Region . . . World Fresh Utilization by Region . . . .


vi


Page

7 8 9

. ..10



. ..12

. ..13


. ..14


. ..15


. . . 16 . . 17 . . 18 . . 19 . . 20 . . 21 . . 30 . . 32









2.3



2.4 2.5


Trade Flow Analysis for Selected Years (1966, 1976 and 1986) by Region in Relation to Partner Regions . . . . . . . . . . .

Trade Flow Analysis for Selected Years (1966, 1976 and 1986) Without Intraregional Trade "Relative Partner Region Exports by Region" . .

Trade Flow Analysis for Selected Years (1966, 1976 and 1986) Without Intraregional Trade "Relative Region Imports from Partner Regions

Trade Flow Analysis for Selected Periods of Five Years (1966-70,1974-78 and 1982-86) . . . .

Trade Flow Analysis for Selected Periods of Five Years (1966-70,1974-78 and 1982-86) Without Intraregional Trade "Relative Partner Region Exports by Region" . . . . . . . .

Trade Flow Analysis for Selected Periods of Five


Years (1966-70,1974-78 and 1982-86) Without Intraregional Trade "Relative Region Imports from Partner Regions" . . . . . . . . . 43

Total Market Demand Equations . . . . . . 108

Export Supply Equations . . . . . . . 109

United States Product Demands . . . . . . 110

Canada Product Demands . . . . . . . . 111

Latin America Product Demands . . . . . . 112

Mediterranean-EC Product Demands . . . . . 113

EC Product Demands . . . . . . . . . 114

Rest of Western Europe Product Demands . . . . 115 Middle East/North Africa Product Demands . . . 116 Rest of Africa Product Demands . . . . . . 117

Far East Product Demands . . . . . . . 118

Oceania Product Demands . . . . . . . . 119

Communist Bloc Product Demands . . . . . . 120


United States CIF Price Linkage Equations vii


121


. . 35 . . 37 . . 38 . . 40 . . 42


2.6 2.7


2.8


5.1 5.2 5.3

5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11 5.12 5.13

5.14









5.15 Canada CIF Price Linkage Equations . . . . . 122

5.16 Latin America CIF Price Linkage Equations . . . 123

5.17 Mediterranean-EC CIF Price Linkage Equations . . 124

5.18 EC CIF Price Linkage Equations . . . . . . 125

5.19 Rest of Western Europe CIF Price Linkage
Equations . . . . . . . . . . . 126

5.20 Middle East/North Africa CIF Price Linkage
Equations . . . . . . . . . . . 127

5.21 Rest of Africa CIF Price Linkage Equations . . . 128

5.22 Far East CIF Price Linkage Equations . . . . 129

5.23 Oceania CIF Price Linkage Equations . . . . . 130

5.24 Communist Bloc CIF Price Linkage Equations . . . 131

5.25 Total Market Demand and Export Supply Equations
Statistics . . . . . . . . . . 193

5.26 Market Demand and Export Supply Equations
Elasticities . . . . . . . . . . 197

5.27 Product Demands Relative Price Elasticities . . . 204

5.28 Product Demands Total Market Demand
Elasticities . . . . . . . . . . 205

5.29 CIF Price Linkage FOB Export Price Elasticities . . 217

5.30 CIF Price Linkage Year Trend Elasticity . . . . 218

5.31 CIF Price Linkage Index Price for Energy
Elasticity . . . . . . . . . . . 219

6.1 World Demand, Imports and Exports Share Per
Region (Cumulative 21 Year Period 1966-1986) . . 238

6.2 Region's Relative Imports Per Partner Region (%
Cumulative 21 Year Period 1966-1986) . . . . 255 H.1 United States, Canada and Latin America Product
Demand and CIF Price Linkage Equations
Statistics . . . . . . . . . . 322


viii









H.2 Mediterranean-EC, EC, Rest of Western Europe
Product Demand and CIF Price Linkage Equations
Statistics . . . . . . . . . . . 323

H.3 Middle East/North Africa, Rest of Africa and Far
East Product Demand and CIF Price Linkage
Equations Statistics . . . . . . . . 324

H.4 Oceania and Communist Bloc Product Demand and
CIF Price Linkage Equations Statistics . . . . 325


ix
















LIST OF FIGURES


Page


Figure

5.1 5.2 5.3


5.4 5.5 5.6 5.7 5.8 5.9


5.10


5.11


Communist Bloc


135 136 137


Total Market Demand for Fresh Oranges in the United States . . . . . . . . .

Total Market Demand for Fresh Oranges in Canada Total Market Demand for Fresh Oranges in Latin America . . . . . . . . . . .

Total Market Demand for Fresh Oranges in the Mediterranean-EC . . . . . . . .

Total Market Demand for Fresh Oranges in the EC Total Market Demand for Fresh Oranges in the Rest of Western Europe . . . . . . .

Total Market Demand for Fresh Oranges in the Middle East/North Africa . . . . . .

Total Market Demand for Fresh Oranges in the Rest of Africa . . . . . . . . .

Total Market Demand for Fresh Oranges in the Far East . . . . . . . . . . .

Total Market Demand for Fresh Oranges in Oceania . . . . . . . . . . .

Total Market Demand for Fresh Oranges in the


Total Export Supply of Fresh Oranges from the United States . . . . . . . . .

Total Export Supply of Fresh Oranges from Canada . . . . . . . . . . .

Total Export Supply of Fresh Oranges from Latin America . . . . . . . . . . .


x


. . 138 . 139 . 140 . 141 . 142 . 143 . 144 . 145


5.12 5.13


5.14


146 147 148


.









Total Export Supply of Fresh Oranges from the


Mediterranean-EC . . . . . . . . . 149

5.16 Total Export Supply of Fresh Oranges from EC . . 150

5.17 Total Export Supply of Fresh Oranges from the
Rest of Western Europe . . . . . . . 151

5.18 Total Export Supply of Fresh Oranges from Middle
East/North Africa . . . . . . . . . 152

5.19 Total Export Supply of Fresh Oranges from the
Rest of Africa . . . . . . . . . 153

5.20 Total Export Supply of Fresh Oranges from the
Far East . . . . . . . . . . . 154

5.21 Total Export Supply of Fresh Oranges from
Oceania . . . . . . . . . . . . 155

5.22 Total Export Supply of Fresh Oranges from the
Communist Bloc . . . . . . . . . . 156

5.23 United States Imports of Fresh Oranges from
Latin America (Product Demand 1_3) . . . . . 163

5.24 United States Imports of Fresh Oranges from the
Middle East/North Africa (Product Demand 1_7) . . 164

5.25 Canada Imports of Fresh Oranges from the United
States (Product Demand 2_1) . . . . . . . 165

5.26 Canada Imports of Fresh Oranges from the Middle
East/North Africa (Product Demand 2_7) . . . . 166

5.27 Canada Imports of Fresh Oranges from the Far
East (Product Demand 2_9) . . . . . . . 167

5.28 Latin America Imports of Fresh Oranges from the
United States (Product Demand 3_1) . . . . . 168

5.29 Latin America Imports of Fresh Oranges from the
EC (Product Demand 3_5) . . . . . . . 169

5.30 Mediterranean-EC Imports of Fresh Oranges from
Latin America (Product Demand 4_3) . . . . . 170

5.31 Mediterranean-EC Imports of Fresh Oranges from
the EC (Product Demand 4_5) . . . . . . . 171

5.32 EC Imports of Fresh Oranges from the
Mediterranean-EC (Product Demand 5_4) . . . . 172 xi


5.15









5.33 EC Imports of Fresh Oranges from the Middle
East/North Africa (Product Demand 5_7) . . . 173

5.34 EC Imports of Fresh Oranges from the Rest of
Africa (Product Demand 5_8) . . . . . . . 174

5.35 Rest of Western Europe Imports of Fresh Oranges
from Mediterranean-EC (Product Demand 6_4) . . . 175

5.36 Rest of Western Europe Imports of Fresh Oranges
from the Middle East/North Africa (Product
Demand 6_7) . . . . . . . . . . 176

5.37 Rest of Western Europe Imports of Fresh Oranges
from the Rest of Africa (Product Demand 6_8) . . 177

5.38 Middle East/North Africa Imports of Fresh
Oranges from Latin America (Product Demand 7_3) . . 178

5.39 Middle East/North Africa Imports of Fresh
Oranges from the Rest of Africa (Product Demand
7_8) . . . . . . . . . . . . 179

5.40 Middle East/North Africa Imports of Fresh
Oranges from the Far East (Product Demand 7_9) . . 180

5.41 Rest of Africa Imports of Fresh Oranges from the
EC (Product Demand 8_5) . . . . . . . . 181

5.42 Rest of Africa Imports of Fresh Oranges from the
Middle East/North Africa (Product Demand 8_7) . . 182

5.43 Far East Imports of Fresh Oranges from the
United States (Product Demand 9_1) . . . . . 183

5.44 Far East Imports of Fresh Oranges from the
Middle East/North Africa (Product Demand 9_7) . . 184

5.45 Far East Imports of Fresh Oranges from Oceania
(Product Demand 9_10) . . . . . . . . 185

5.46 Oceania Imports of Fresh Oranges from the United
States (Product Demand 10_1) . . . . . . 186

5.47 Oceania Imports of Fresh Oranges from the Middle
East/North Africa (Product Demand 10_7) . . . . 187

5.48 Communist Bloc Imports of Fresh Oranges from the
Mediterranean-EC (Product Demand 11_4) . . . . 188

5.49 Communist Bloc Imports of Fresh Oranges from the
Middle East/North Africa (Product Demand 11_7) . . 189


xii









6.1 Total Market Demand Changing Average Market
Price (Major World Consumers) . . . . . . 239

6.2 Total Market Demand Changing Average Market
Price (Major World Importers) . . . . . . 240

6.3 Total Market Demand Changing Income (GDP) (Major
World Consumers) . . . . . . . . . 241

6.4 Total Market Demand Changing Income (GDP) (Major
World Importers) . . . . . . . . . 242

6.5 Export Supply Changing FOB Average Export Price
(Major World Exporters) . . . . . . . . 250

6.6 Export Supply Changing Fresh Production (Major
World Exporters) . . . . . . . . . 251

6.7 United States Imports Changing Import Prices . . 256

6.8 United States Imports Changing Total Market
Demand . . . . . . . . . . . . 257

6.9 Canada Imports Changing Import Prices . . . . 260

6.10 Canada Imports Changing Total Market Demand . . . 261

6.11 Latin America Imports Changing Import Prices . . 264

6.12 Latin America Imports Changing Total Market
Demand . . . . . . . . . . . . 265

6.13 Mediterranean-EC Imports Changing Import Prices . . 267

6.14 Mediterranean-EC Imports Changing Total Market
Demand . . . . . . . . . . . . 268

6.15 EC Imports Changing Import Prices . . . . . 270

6.16 EC Imports Changing Total Market Demand . . . . 272

6.17 Rest of Western Europe Imports Changing Import
Prices . . . . . . . . . . . . 274

6.18 Rest of Western Europe Imports Changing Total
Market Demand . . . . . . . . . . 275

6.19 Middle East/North Africa Imports Changing Import
Prices . . . . . . . . . . . . 277


xiii









6.20


6.21 6.22 6.23

6.24

6.25 6.26 6.27 6.28


Middle East/North Africa Imports Changing Total Market Demand . . . . . . . . .

Rest of Africa Imports Changing Import Prices Rest of Africa Imports Changing Total Market Demand . . . . . . . . . .

Far East Imports Changing Import Prices . . Far East Imports Changing Total Market Demand Oceania Imports Changing Import Prices . . Oceania Imports Changing Total Market Demand Communist Bloc Imports Changing Import Prices Communist Bloc Imports Changing Total Market Demand . . . . . . . . . .


xiv


. . 278 . . 280


. . 281 . . 283 . . 285 . . 287 . . 288 . . 289 . . 291
















Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


MULTIPLE-REGION EQUILIBRIUM WORLD TRADE MODEL: THE ORANGE INDUSTRY

By

Esteban R. Brenes

May, 1992

Chairman: Dr. Ronald W. Ward
Major Department: Food and Resource Economics

A multiple-region equilibrium trade model for the fresh orange industry including 11 regions of the world was developed and estimated.

The model is used to understand the major driving factors affecting fresh orange consumption and trade. The model is a modified spatial equilibrium model that takes into account that products are differentiated by country of origin. Armington developed the demand theory underlying this assumption. The model assumes a constant ratio of elasticity of substitution (CRES) index which makes the model somewhat less restrictive.

The model is estimated using a nonlinear two stage least square procedure. Graphical, statistical and economic analyses of the results

are used to evaluate the performance of the model and the implications for the fresh orange industry. The results indicate that the model performs well. A sensitivity analysis of the system is developed to evaluate the consequences of changes in the main variables of the model.


xv









Total market demand analysis shows that market prices and income are the major drivers for world consumption of fresh oranges in most regions. Major world importers are more sensitive to changes in the average market price than major world consumers with domestic production. Export supply equations show weak FOB export price parameters versus strong fresh production parameters. This indicates that major export decisions are driven mainly by fresh production.

Product demand analysis shows the role of prices as an allocative tool and the importance of the market size to determine consumer preferences when facing several product sources. Market positions and opportunities for all regions were determined. The regions included the

United States, Canada, South America, the Mediterranean-EEC countries, the rest of the EEC, the rest of western Europe, the Middle East and North Africa countries, the rest of Africa, the Far East, Oceania, and the Communist Bloc. The Communist Bloc was defined as it existed prior to the recent political changes of 1991.


xvi
















CHAPTER 1
INTERNATIONAL TRADE AND AGRICULTURAL PRODUCTS: THE ORANGE INDUSTRY



Introduction



Developing countries have long recognized the importance of trade to their national welfare. Exchanging the goods that they produce with their endowments and experience for the goods in which other countries have a

comparative advantage provides the potential for both growth and development. The world is becoming smaller in terms of communication and the ability to trade. International trade has been expanding at an increasing rate, especially in the last three decades. Most countries are dependent to some degree on the foreign currency generated through trade. Some countries consider international trade as one of the most important means for development, especially the ones that have a high foreign debt.

The United States continues to implement important macro-economic and commercial policies to improve its competitive position in the international trade arena.

Less developed countries use international trade as a means for development and subsistence because their domestic economies are poor and small in most cases. Therefore, it is very important to develop new markets for their products as well as to develop new marketable products. Most of these nations possess a high foreign debt that must be paid with


I









2


foreign currency. Imports to these countries are frequently higher than exports, which implies a need for more foreign currency, i.e., they are often net importers.

In the 1980s, many less developed countries changed their

development strategy from an import substitution scheme to export promotion. While the reasons for change differ among countries, the main

reasons include interruption of multilateral and bilateral agreements, the increase of the fiscal deficit due to subsidies, recognition of

inefficiencies due to high import barriers, and the need of foreign currency.

Developed countries such as Japan, Germany, Italy, and France are

highly interdependent on international trade. The United States also recognizes the importance of being competitive and the need of interdependence with potential trade partners. These changes increased

the importance of international trade and market development worldwide with the consequence being an increase in the relative importance of international trade.

In theory, international trade is dependent on comparative advantage. This means that countries will tend to trade goods and services that they produce efficiently for goods and services in which other countries have a comparative advantage. Porter (1990) mentions four broad attributes to shape the environment in which local firms compete in order to achieve international success in a particular industry. First,

factor conditions, which mean the nation's position in factors of production necessary to compete in a given industry; second, demand conditions which consider the natures of domestic demand for the









3

industry's product or service; third, related and supporting industries

which refer to the presence or absence in the nation of supplier and related industries that are internationally competitive; fourth, firm strategy, structure, and rivalry which refer to the conditions of the nation governing how companies are created, organized, and managed, and the nature of rivalry. However, in the real world, international trade is not solely driven on comparative advantage but also on variables including tariffs, quotas, subsidies, international agreements, and domestic

policies.

World real-value trade increased at an average rate of 6.7% a year while real Gross National Product (GNP) increased at an average rate of

4.1% from 1966 to 1986 (International Monetary Fund [IMF] Direction of Trade). During that period the United States economy became increasingly interdependent with world economies. The value of United States total trade as a percentage of GNP increased from 7.5 to 14 for the same period.



International Trade of Agricultural Products



International agricultural trade depends heavily on national and regional policies. For agricultural products, import protection and export subsidies are usual in most countries; especially in cases where a

particular product is socially desirable and support groups are politically strong. For example, grain production is usually protected from imports in most countries. Reasons given to support such policies

have to do with income distribution given the amount of people involved in production, process and distribution of grains; self-sufficiency; and less









4

need of foreign currency. In many cases, the government has to support production through higher prices for these people to continue in business. This implies higher fiscal deficits, given subsidies and underpriced exports to get rid of excess inventories. This promotes inefficiencies and a waste of resources.

A controversial agricultural policy is the Common Agricultural Policy (CAP) of the European Community (EC). The CAP is primarily a market-regulation and price-support policy. It currently covers grains,

rice, poultry and eggs, dairy products, pork, beef and veal, sugar, certain fruits and vegetables, and certain processed agricultural products. The protectionism system includes common customs tariffs for imports and internal regulations designed to protect EC producers. The system gradually eliminates trade barriers among nations within the bloc

but imposes a common external trade barrier. The variable levy system is the major instrument used by the EC to protect domestic markets from foreign competition. It imposes a levy equal to the difference between the world price and the domestic support price. This tends to make EC imports from other countries have a perfectly inelastic demand within a considerable price range, with outside countries the residual suppliers.

Some of the proceeds from the levy are further used to subsidize EC exports (Tweeten, 1979). The CAP regulations in fresh fruits set quality standards for a variety of products and outline a price and intervention system in most cases.

Tariffs and nontariff barriers, along with preferential treatment,

have become increasingly important factors influencing agricultural trade. In the case of fresh fruits, CAP regulations are of particular importance









5

with the recent enlargement of the EC to 12 nations with the inclusion of

Spain and Portugal. Both are major producers and exporters of fresh fruits to the rest of Europe.

Other factors affecting agricultural trade are income, population,

demographic variables within the trading regions, and exchange rates. Income and population are important to determine the level of consumption. As these two variables increase, higher levels of consumption and shifts from one bundle of goods to another are expected. Exchange rates affect

the real terms of trade among countries, especially in cases when they are managed by governments and are not allowed to fluctuate freely in the market. Transportation is an important linkage variable in world trade.

The linkage is between the Free On Board (FOB) export price from any country and the Cost Insurance Freight (CIF) import price at the final market. Substitute product prices should have a positive effect on the consumption of a specific commodity.

Total agricultural trade increased at an annual rate of 2.8% while agricultural production increased by 2.3% a year from 1966 to 1986 (Food and Agricultural Organization [FAO] Trade and Production Yearbooks). The United States agricultural sector represents about 15% of total exports.

Hence, United States agricultural market prices are strongly influenced by supply and demand conditions among major world markets (Statistical Abstract of the United States and FAO Trade Yearbook). These statistics reflect the increasing importance of trade in the world economy and, in particular, the importance of agriculture in world trade.

The focus of the present study will be the fresh orange industry. World trade in the fresh orange industry must be studied from the









6

perspective of two different goods, fresh and processed oranges. To

improve our understanding and ease the following analysis of the fresh orange industry, Tables 1.1 to 1.15 show 11 regions of significant trade.

These regions have been selected by considering similarities in supply or demand among the countries and their importance in the production of and

international trade in fresh oranges. The regions identified are the United States, Canada, Latin America, Mediterranean-European Community countries, the rest of the European Community (EC), the rest of Western Europe, Middle East/North Africa countries, the rest of Africa, the Far

East, Oceania and the Communist Bloc. The Communist Bloc is defined as it existed prior to the recent political changes of 1991. Cuba is included in the Bloc given the existance of trade agreements with Eastern Europe. Appendix A shows the composition of these regions.

As shown in Table 1.1, world orange production increased at a rate

of 3.3% a year from 1966 to 1986. Table 1.2 shows that world fresh utilization increased at a rate of 2.6% a year for the same period. The

processed industry increased faster than fresh utilization in the last decade. From 1978 to 1986, world processed production increased at a rate of 2.7% a year while fresh utilization increased 2.4% a year (see Tables

1.3 and 1.2).

Fresh orange world trade increased by 2.2% a year from 1966 to 1986, and 1.1% from 1978 to 1986 (see Tables 1.4 and 1.6). If intraregional trade or trade between countries of the same region is not considered, international trade in fresh oranges showed an increase of 1.9% a year for the same period (see Tables 1.8 and 1.12). This percentage is higher than 1.1%, meaning that trade among regions increased in the last decade.









Table 1.1 World Orange Production by Region

Annual Annual Annual
Growth Growth Growth
Rate Rate Rate
Region 1966 1976 1978 1986 1966-86 1976-86 1978-86

----- (000) Metric Tons ----- --- Percent of Change --United States 7598 10183 9268 7192 -0.3 -3.4 -3.1

Canada 0 0 0 0 N.A. N.A. N.A.

Latin America 5540 12117 11832 18535 6.2 4.3 5.8

Mediterranean-EC 4208 5472 5267 6840 2.5 2.3 3.3

E.C. 4 31 29 34 11.3 0.9 2.0

Rest of Western Europe 0 00 0 0 N.A. N.A. N.A.

Middle East/North Africa 3067 4664 5364 5794 3.2 2.2 1.0

Rest of Africa 773 1034 1132 1022 1.4 -0.1 -1.3

Far East 4023 6532 6771 8354 3.7 2.5 2.7

Oceania 249 360 410 574 4.3 4.8 4.3

Communist Bloc 192 292 408 728 6.9 9.6 7.5

World Total 25654 40685 40481 49073 3.3 1.9 2.4


Source: FAO Production Yearbook. Various issues.


--j








Table 1.2 World Fresh Utilization by Region

Annual Annual Annual
Growth Growth Growth
Rate Rate Rate
Region 1966 1976 1978 1986 1966-86 1976-86 1978-86

----- (000) Metric Tons ----- --- Percent of Change --United States 2575 2294 2031 2322 -0.5 0.1 1.7

Canada 0 0 0 0 N.A. N.A. N.A.

Latin America 5290 8336 7342 9394 2.9 1.2 3.1

Mediterranean-EC 3689 4807 4600 5787 2.3 1.9 2.9

E.C. 4 31 29 34 11.3 0.9 2.0

Rest of Western Europe 0 0 0 0 N.A. N.A. N.A.

Middle East/North Africa 2862 4273 4943 5122 3.0 1.8 0.4

Rest of Africa 692 894 980 919 1.4 0.3 -0.8

Far East 3801 5634 5934 7619 3.5 3.1 3.2

Oceania 202 196 202 251 1.1 2.5 2.7

Communist Bloc 192 287 398 638 6.2 8.3 6.1

World Total 19307 26752 26459 32086 2.6 1.8 2.4









9


Table 1.3 World Processed Production by Region


Region


United States Canada

Latin America Mediterranean-EC E.C.

Rest of Western Europe Middle East/North Africa Rest of Africa Far East Oceania Communist Bloc World Total


1978 1986

(000) 65 Degree Brix Metric Tons

732.5 481.3

0.0 0.0

406.7 895.3

60.4 103.2

0.0 0.0

0.0 0.0

38.1 65.8

13.8 10.1

75.8 72.0

18.8 31.6

0.9 8.8

1347.1 1668.0


Annual Growth
Rate 1978-86

Percent of Change

-5.1

N.A.

10.4 6.9 N.A.

N.A.

7.1 -3.8

-0.6

6.7 32.9 2.7









Table 1.4 World Fresh Orange Exports by Region

Annual Annual Annual
Growth Growth Growth
Rate Rate Rate
Region 1966 1976 1978 1986 1966-86 1976-86 1978-86
----- (000) Metric Tons ----- --- Percent of Change --United States 258.6 464.1 355.9 413.0 2.4 -1.2 1.9

Canada 0.1 0.1 0 0.3 6.9 16.1 31.5

Latin America 106.6 119.2 164.7 229.3 3.9 6.8 4.2

Mediterranean-EC 1514.6 1937.5 1802.2 2833.9 3.2 3.9 5.8

E.C. 32.8 122.6 121.5 195.0 9.3 4.8 6.1

Rest of Western Europe 2.1 5.9 7.8 2.2 0.2 -9.5 -14.7

Middle East/North Africa 1215.9 1871.9 1893.2 1316.7 0.4 -3.5 -4.4

Rest of Africa 264.4 285.7 274.4 209.0 -1.2 -3.1 -3.3

Far East 65.8 142.3 137.5 113.9 2.8 -2.2 -2.3

Oceania 23.5 11.2 30.1 47.3 3.5 15.5 5.8

Communist Bloc 0.2 60.4 140.9 35.0 30.5 -5.3 -16.0

World Total 3484.5 5020.7 4928.3 5395.6 2.2 0.7 1.1


Source: United Nations Trade Data Tapes.


C)












World Processed Orange Exports by Region


Region


United States Canada

Latin America Mediterranean-EC E.C.

Rest of Western Europe Middle East/North Africa Rest of Africa Far East Oceania Communist Bloc World Total


1978 1986

(000) 65 Degree Brix Metric Tons 148.1 76.4

0.5 2.8

296.1 831.4

29.1 33.5

97.6 207.2

3.3 7.1

95.9 112.3

2.3 1.4

0.9 6.7

0.1 1.5

.0 0.3

673.8 1280.7


Source: United Nations Trade Data Tapes.


Table 1.5


11


Annual Growth
Rate 1978-86

Percent of Change

-7.9

23.1 13.8 1.8 9.9

10.1

2.0 -5.6

29.0

45.3 46.2

8.4








Table 1.6 World Fresh Orange Imports by Region

Annual Annual Annual
Growth Growth Growth
Rate Rate Rate
Region 1966 1976 1978 1986 1966-86 1976-86 1978-86

----- (000) Metric Tons ----- --- Percent of Change --United States 28.3 32.6 52.9 49.4 2.8 4.2 -0.8

Canada 180.5 225.5 180.2 182.1 .0 -2.1 0.1

Latin America 9.7 18.1 14.2 3.8 -4.6 -14.5 -15.2

Mediterranean-EC 0.0 0.9 1.5 8.2 N.A. 24.8 24.2

E.C. 2398.4 2736.2 2655.5 3464.6 1.9 2.4 3.4

Rest of Western Europe 384.6 439.7 436.0 551.2 1.8 2.3 3.0

Middle East/North Africa 22.0 590.4 440.5 279.0 13.5 -7.2 -5.5

Rest of Africa 10.6 10.6 14.5 8.2 -1.3 -2.5 -6.8

Far East 119.3 238.9 251.7 382.8 6.0 4.8 5.4

Oceania 15.3 8.9 15.5 20.4 1.4 8.7 3.5

Communist Bloc 315.6 718.8 865.8 445.8 1.7 -4.7 -8.0

World Total 3484.5 5020.7 4928.3 5395.6 2.2 0.7 1.1


Source: United Nations Trade Data Tapes.


I'3









13


Table 1.7 World Processed Orange Imports by Region


Region


United States Canada

Latin America Mediterranean-EC E.C.

Rest of Western Europe Middle East/North Africa Rest of Africa Far East Oceania Communist Bloc World Total


1978 1986

(000) 65 Degree Brix Metric Tons 136.8 500.6

98.3 89.5

6.6 7.1

4.0 16.7

324.6 568.4

69.8 49.6

14.1 10.0

2.0 1.6

11.7 29.4

1.0 3.7

5.1 4.1

673.8 1280.7


Source: United Nations Trade Data Tapes.


Annual Growth
Rate 1978-86

Percent of Change

17.6

-1.2

0.9 19.6 7.3

-4.2

-4.2

-2.4

12.3 18.1 -2.7

8.4









Table 1.8 World Fresh Orange Export Quantities by Region (Excluding Intraregional Trade)

Annual Annual Annual
Growth Growth Growth
Rate Rate Rate
Region 1966 1976 1978 1986 1966-86 1976-86 1978-86
----- (000) Metric Tons ----- --- Percent of Change --United States 258.6 464.1 355.9 413.0 2.4 -1.2 1.9

Canada 0.1 0.1 .0 0.3 6.9 16.1 31.5

Latin America 103.4 103.1 152.7 226.9 4.0 8.2 5.1

Mediterranean-EC 1514.6 1937.3 1802.2 2833.7 3.2 3.9 5.8

E.C. 2.8 9.6 13.0 21.1 10.5 8.2 6.3

Rest of Western Europe 1.1 0.9 4.9 1.1 .0 2.6 -17.0

Middle East/North Africa 1205.3 1410.9 1524.9 1092.6 -0.5 -2.5 -4.1

Rest of Africa 262.7 280.2 268.6 206.9 -1.2 -3.0 -3.2

Far East 24.3 64.5 47.9 41.9 2.8 -4.2 -1.6

Oceania 13.7 8.7 21.2 36.7 5.1 15.5 7.1

Communist Bloc 0.2 2.0 4.2 12.4 23.9 19.8 14.6

World Total 3386.6 4281.3 4195.5 4886.6 1.9 1.3 1.9


Source: United Nations Trade Data Tapes.


9
-'S









Tab] 1 .9 World lresh Orange Export Value!; by Regioi (Excluiding Itftraregional Trade) Annual Annual Annual
Growth Growth Growth
Rate Rate Rate
Region 1966 1976 1978 1986 1966-86 1976-86 1978-86
----- In Million of U.S. Dollars ------ --- Percent of Change --United States 47.0 118.8 144.0 233.0 8.3 7.0 6.2

Canada .0 .0 .0 0.1 11.5 20.7 26.6

Latin America 6.2 14.7 27.9 61.5 12.1 15.4 10.4

Mediterranean-EC 178.9 406.1 530.0 1025.2 9.1 9.7 8.6

E.C. 0.7 2.8 5.7 12.6 15.2 16.1 10.3

Rest of Western Europe 0.2 0.3 1.3 0.6 5.4 9.2 -8.3

Middle East/North Africa 172.7 322.1 410.7 361.6 3.8 1.2 -1.6

Rest of Africa 35.7 59.1 103.6 83.5 4.3 3.5 -2.7

Far East 4.6 21.1 17.2 17.6 7.0 -1.8 0.2

Oceania 2.5 3.1 6.5 15.2 9.4 17.1 11.2

Communist Bloc .0 0.4 0.8 3.9 28.7 25.0 21.9

World Total 448.7 948.6 1247.7 1814.7 7.2 6.7 4.8


Source: United Nations Trade Data Tapes.









16


Table 1.10 World Processed Orange Export Quantities by Region (Excluding
Intraregional Trade)


Region


United States Canada

Latin America Mediterranean-EC E.C.

Rest of Western Europe Middle East/North Africa Rest of Africa Far East Oceania Communist Bloc World Total


1978 1986

(000) 65 Degree Brix Metric Tons 148.1 76.4

0.5 2.8

296.1 827.8

28.9 33.3

16.4 31.9

2.6 6.1

95.5 112.1

2.3 1.4

0.5 2.2

0.1 0.9

.0 0.3

591.0 1095.2


Source: United Nations Trade Data Tapes.


Annual Growth
Rate 1978-86

Percent of Change

-7.9

22.9 13.7

1.8 8.6

11.5

2.0 -5.7

19.4 34.0 39.4 8.0









17


Table 1.11 World Processed Orange Export Values by Region (Excluding
Intraregional Trade)


Region


United States Canada

Latin America Mediterranean-EC E.C.

Rest of Western Europe Middle East/North Africa Rest of Africa Far East Oceania Communist Bloc World Total


1978 1986

In Millions of
U.S. Dollars

98.0 66.6

0.6 5.1

288.6 671.8

23.4 32.9

16.2 29.3

2.9 4.2

59.8 102.2

4.2 1.2

0.3 1.5

0.1 0.7

.0 0.2

494.1 915.8


Source: United Nations Trade Data Tapes.


Annual Growth
Rate 1978-86

Percent of Change

-4.7

31.7 11.1

4.4

7.6

4.4 6.9

-14.2

23.0 32.5

52.4 8.0









Table 1.12 World Fresh Orange Import Quantities by Region (Excluding Intraregional Trade)

Annual Annual Annual
Growth Growth Growth
Rate Rate Rate
Region 1966 1976 1978 1986 1966-86 1976-86 1978-86

----- (000) Metric Tons ----- --- Percent of Change --United States 28.3 32.6 52.9 49.4 2.8 4.2 -0.8

Canada 180.5 225.5 180.2 182.1 .0 -2.1 0.1

Latin America 6.5 2.0 2.2 1.4 -7.5 -4.0 -5.9

Mediterranean-EC 0.0 0.7 1.5 8.0 N.A. 27.3 23.7

E.C. 2368.5 2623.3 2546.9 3290.7 1.7 2.3 3.3

Rest of Western Europe 383.7 434.6 433.2 550.2 1.8 2.4 3.0

Middle East/North Africa 11.4 129.4 72.2 54.9 8.2 -8.2 -3.4

Rest of Africa 8.8 5.1 8.7 6.1 -1.8 1.8 -4.3

Far East 77.8 161.1 162.0 310.9 7.2 6.8 8.5

Oceania 5.5 6.5 6.7 9.8 3.0 4.3 4.9

Communist Bloc 315.6 660.5 729.1 423.2 1.5 -4.4 -6.6

World Total 3386.6 4281.3 4195.5 4886.6 1.9 1.3 1.9


Source: United Nations Trade Data Tapes.









Table 1.13 World Fresh Orange Import Values by Region (Excluding Intraregional Trade)

Annual Annual Annual
Growth Growth Growth
Rate Rate Rate
Region 1966 1976 1978 1986 1966-86 1976-86 1978-86

--- In Millions of U.S. Dollars --- --- Percent of Change --United States 3.7 8.8 14.2 28.9 10.8 12.6 9.3

Canada 38.6 99.1 76.7 133.1 6.4 3.0 7.1

Latin America 1.0 0.8 2.6 1.0 0.2 2.6 -11.1

Mediterranean-EC .0 0.2 0.8 4.5 N.A. 38.4 24.7

E.C. 405.3 751.0 966.2 1367.7 6.3 6.2 4.4

Rest of Western Europe 68.7 153.1 198.4 333.9 8.2 8.1 6.7

Middle East/North Africa 1.1 44.3 28.0 14.5 13.6 -10.5 -7.8
Rest of Africa 3.0 2.8 4.5 6.2 3.7 8.4 3.9

Far East 19.1 66.0 97.4 241.1 13.5 13.8 12.0

Oceania 1.5 2.2 3.0 6.7 7.6 12.1 10.9

Communist Bloc 43.9 168.8 246.9 203.9 8.0 1.9 -2.4

World Total 586.0 1297.0 1638.6 2341.5 7.2 6.1 4.6


Source: United Nations Trade Data Tapes.









20


Table 1.14 World Processed Orange Import Quantities by Region (Excluding
Intraregional Trade)


Region


United States Canada

Latin America Mediterranean-EC E.C.

Rest of Western Europe Middle East/North Africa Rest of Africa Far East Oceania Communist Bloc World Total


1978 1986

(000) 65 Degree Brix Metric Tons 136.8 500.6

98.3 89.5

6.5 3.4

3.8 16.6

243.4 393.0

69.1 48.6

13.7 9.8

1.9 1.6

11.3 25.0

1.0 3.0

5.2 4.2

591.0 1095.2


Source: United Nations Trade Data Tapes.


Annual Growth
Rate 1978-86

Percent of Change

17.6

-1.2

-7.8

20.1 6.2

-4.3

-4.1

-2.5

10.4 15.4 -2.6

8.0









21


Table 1.15 World Processed Orange Import Values by Region (Excluding
Intraregional Trade)


Region


United States Canada

Latin America Mediterranean-EC E.C.

Rest of Western Europe Middle East/North Africa Rest of Africa Far East Oceania Communist Bloc World Total


1978 1986

In Millions of
U.S. Dollars

150.6 518.9

106.0 104.4

5.9 8.1

3.7 16.7

230.2 414.1

83.0 55.0

12.5 16.1

2.3 2.1

19.5 49.5

1.4 3.5

3.3 3.4

618.4 1191.9


Source: United Nations Trade Data Tapes.


Annual Growth
Rate 1978-86

Percent of Change

16.7

-0.2

3.9 20.6 7.6 -5.0

3.2 -1.3

12.4 12.1 0.5 8.5









22


World trade in the processed industry showed a higher average increase from 1978 to 1986, reaching 8.4% a year (see Tables 1.5 and 1.7). If intraregional trade is not considered, the processed industry grew by 8% a year during the same period.

Tables 1.9, 1.11, 1.13, and 1.15 show world fresh and processed orange exports and imports excluding intraregional trade. Trade is

expressed in value terms measured in United States dollars.

The United States fresh orange production decreased at an average

rate of .3% a year from 1966 to 1986. During the 1970s, production

increased rapidly and later decreased mainly due to unfavorable weather

conditions. Oranges used for fresh consumption decreased at a rate of .5% a year from 1966 to 1986. Total United States trade increased at a rate

of 2.4% a year for the same period. This shows that the United States has actually decreased its real participation in world fresh utilization. It has increased the use of oranges in the processed industry, along with a

slight increase in its role in the international trade arena for fresh oranges. In the processed industry, the United States has decreased its

production participation relative to the rest of the world and passed from a net exporter to a net importer of FCOJ (Frozen Concentrated Orange Juice) (Tables 1.3, 1.5, and 1.7).

As shown in the different tables introduced in this chapter, in the

last two decades trade patterns in the orange industry have changed dramatically. The United States, once the world's major producer of fresh oranges and orange juice, today is no longer the leading producer or exporter. Latin America, mainly Brazil, is the major producer of oranges in the world. Most of Brazil's production is used for FCOJ and is









23

exported. Per capita consumption of fresh oranges within Brazil is quite low. The Mediterranean countries are major producers and exporters of fresh oranges. The rest of Europe as well as Canada have always been net importers of fresh and processed oranges. Other regions, especially the

Middle East/North Africa and the Far East, have increasingly become important producers and traders in the orange industry. Chapter 2 will outline production and trade flows of the fresh orange industry in more detail.



Problem and Objectives



World consumption trends indicate that consumers are interested in

healthy and natural products. Fresh product consumption is increasing and its potential growth is promising. Given the changes in consumption patterns and the improvement in the transportation systems, studying the

fresh markets is of increasing importance. Fresh oranges, in particular, provide consumers with natural flavor and important vitamins and minerals. Fresh oranges and FCOJ are direct substitutes in the supply decision process, but are not considered substitutes in the consumption side. Consumer satisfaction is considered to be different for each good. Most

recent literature has studied mainly processed trade, which has been growing faster than fresh trade in the last two decades. However, the value and quantity of fresh trade are two and four times that for the processed trade, respectively (see Tables 1.8 to 1.15). World fresh

utilization represents 65% of total orange production.









24

The discussion in the previous section described some of the factors affecting trade patterns and market shares of the fresh orange industry during the last 21 years. Even though the market has experienced

important increases, several countries, including the United States, have

experienced pronounced changes in their trade patterns for value and quantity. The dynamics of the marketplace are illustrated with

participation of the Middle East/North Africa countries in the European

market; Latin America's increasing share of the European market; the increasing portion of the United States in the Far East markets; the Middle East/North Africa's increasing participation as consumers of fresh and processed oranges; and the potential growth of China as a supplying and consuming country.

The fresh orange industry is quite important for some regions, especially for the United States, Latin America, Mediterranean-EC, Middle East/North Africa, and Far East, as producers, consumers, and exporters. Producers and exporters need to understand the major driving factors for fresh consumption and their competitive position in foreign markets. It

will allow them to compete with more information, possibly achieve international success, and help to develop new markets. The fresh orange industry is also important for net importers such as Canada, EC, rest of

Western Europe, and the Communist Bloc. These regions will be interested in knowing which are the major driving factors for fresh consumption, and

demand and price linkages between the region and its major trading partners.

Given the changes in the fresh orange market, studying the world trade flows becomes important for the future of the United States orange









25

industry as well as for other partner regions. Modeling these changes is

the major objective of this study. The analysis will provide information to help understand the reasons for changes in market shares among major suppliers and facilitate longer term forecasts and policy analyses.

To accomplish the objectives of the present study, international trade linkages among the major trading regions must be identified. It is

also necessary to recognize the current and emerging problems in the industry. This information will be helpful to study changes in trade patterns arising from changes in supply and demand conditions and from changes in policy variables such as tariff levels and institutional constraints.

Analysis of the demand parameters will show the likely future direction of trade. Using price elasticities, it will be possible to predict responses in the different markets to changes in prices. The role of prices as an allocative tool can be shown. Income and population

elasticities will give an indication of possible adjustments in

consumption and trade patterns. In general, it will be possible to forecast trade patterns among importers and exporters. The system could be used to construct a sensitivity analysis to study the behavior of the

fresh orange trade model given shocks in the different variables including price, market size, income, population, fresh production, tariff and nontariff barriers, and other variables.

The specific objectives of this research are

1. Specify a multiple-region equilibrium world trade model for the

fresh orange industry. Relative and substitute prices,









26

transportation costs, incomes, populations, exchange rates, and

policy variables were considered.

2. Estimate the demand, export supply and price equations that

explain the individual elements of the trade flows. A

simultaneous system was specified and estimated.

3. Analyze the implications contained in the estimated model.

The estimated parameters were used to study the reasons for changes in market shares and to provide information for

specific policy issues.

4. Develop a sensitivity analysis of the model for the major

trading regions under different economic scenarios. To ease forecasts, exogenous changes in the different variables such as import and export price, market size, income, population,

fresh production, tariffs and nontariff barriers, and other

variables were considered.



Scope



The proposed study will develop a world trade model with demand and

export supply equations for selected major trading regions. 'The model will include fresh oranges as defined by the United Nations Standard International Trade Classification (SITC) (1975) code 05711. Fresh

oranges and orange juice are direct substitutes in the supply decision process but are not considered substitutes on the consumption side. To be complete, the model should take into consideration the supply relationship between the two goods. The end product model includes market size, market









27

and relative prices, transportation costs, tariff barriers (national or

regional agricultural policies), price of substitutes, income levels, exchange rates, and population.

International trade data including value and quantity were obtained from United Nations' Commodity Trade Statistic Tapes (1987). These data are gathered by each member country and sent to the statistics office in New York. The data contain import and export value and quantity information for each member, showing the partner country. The price data used in this dissertation are unit prices obtained by dividing value by

quantity for each trade flow. As expected, many errors were found. Most of them were probably related to gathering problems and inconsistencies. Where errors were detected, the data were corrected in what seemed to be

an appropriate way. Tariff barriers were obtained from different sources. The kinds of tariffs differ from country to country, from an ad valorem basis, using CIF import prices or FOB export prices as a base, to fixed

dollar amounts per ton. Tariffs were averaged using different methods to obtain the best possible regional tariff. Nontariff barriers are not considered in the study, given that most of them are seasonal and the model uses annual data. The period of study is 1966 to 1986.



Methodology



The present study develops a fresh orange multiple-region equilibrium world trade model. To ease the estimation and the analysis, world countries are aggregated into 11 regions. The regions have been selected by considering similarities in supply and/or demand among the









28

countries and their importance in production and international trade in fresh oranges.

The model is a nonlinear simultaneous system of equations that contains 440 equations of which 242 were estimated. The equations to

estimate were total market demands (one per region), export supplies (one per region), product demands (one per partner in each region), and price

linkage equations (one per partner in each region). The rest of the

equations in the model were identities.

A nonlinear simultaneous system estimator was used for the

estimation of the model. Model results were analyzed to evaluate the fit of the model and its accuracy for simulation. The final model and its parameters were used to develop a sensitivity analysis to investigate the effects of changes in selected policy variables.



Overview



Chapter 2 discusses world production and trade flows for fresh oranges. Chapter 3 covers the literature review for agricultural trade and fresh orange trade models. Chapter 4 presents the fresh orange trade model to be estimated. Chapter 5 discusses the methods used for the estimation of the model. It also develops a graphical, statistical, and economic analysis for the results of the estimation. Chapter 6 develops the sensitivity analysis.
















CHAPTER 2
FRESH ORANGE WORLD PRODUCTION AND TRADE



Introduction



This chapter discusses world production and trade flows of fresh oranges. The discussion will be based on several tables for 11 specified regions of the world. These regions were selected based on similarities

of supply and demand conditions among the different countries included in each region with regard to the orange industry. The regions are the United States (US), Canada (CAN), Latin America (LA), MediterraneanEuropean Community countries (MED-EC), the rest of the European Community

countries (EC), rest of Western Europe (RWE), Middle East/North Africa (ME/NA), rest of Africa (RAF), Far East (FE), Oceania (OCE), and Communist Bloc (COMMB). The Communist Bloc is defined as it existed before the political changes of 1991. Appendix A shows the countries included in each region.



Production Analysis



Table 2.1 shows the production levels of oranges in the 11 regions

identified for 1966, 1976, and 1986. These years were selected to

illustrate changes through time. World orange production increased at an annual rate of 3.3% in the last 20 years and increased at a rate of 1.9% 29








Table 2.1 World Orange Production By Region Annual Annual
Growth Growth
Rate Rate
Region 1966 1976 1986 1966-86 1976-86

--- (000) Metric Tons --- Percent of Change

United States 7598 10183 7192 -0.3 -3.4

Canada 0 0 0 N.A. N.A.

Latin America 5540 12117 18535 6.2 4.3

Mediterranean-EC 4208 5472 6840 2.5 2.3

E.C. 4 31 34 11.3 0.9

Rest of Western Europe 0 0 0 N.A. N.A.

Middle East/North Africa 3067 4664 5794 3.2 2.2

Rest of Africa 773 1034 1022 1.4 -0.1

Far East 4023 6532 8354 3.7 2.5

Oceania 249 360 574 4.3 4.8

Communist Bloc 192 292 728 6.9 9.6

World Total 25654 40685 49073 3.3 1.9


Source: FAO Production Yearbook. Various issues.


I~~)
0









31

in the last decade. Table 2.2 shows the portion of that production used as fresh product. Production utilization in fresh form decreased from 75.3% in 1966 to 65.4% in 1986 (compare data in Table 2.2 as a percent of the corresponding figures in 2.1).

Table 2.1 also shows that the major world producer was Latin America, with 21.6% in 1966, 29.8% in 1976, and 37.8% in 1986. This

region exhibited one of the faster annual growth rates, 6.2% during the 20 year period. However, as shown in Table 2.2, over 50% of the oranges of this region went to the processed industry, leaving 9.4 million tons for fresh utilization in 1986. This represented 29.3% of total world fresh utilization.

The second largest producer of oranges was the Far East, with 15.7%

in 1966, 16.1% in 1976, and 17% in 1986. These percentages show that the Far East region has not only maintained its participation in the total world production of oranges in the last 20 years, but has also increased it. Table 2.1 shows that the Far East region has doubled in absolute terms its total production in the same period. In addition, 91.2% of total production was used fresh in 1986.

The third largest producer was the United States, with 29.6% in 1966, 25% in 1976, and 14.7% in 1986. Even though the United States is

still a major world producer, its share of total production of oranges has been decreasing, especially in the last decade. The United States used

most of its production in the processed industry. In 1986, 32.3% of total production was used fresh, indicating that the United States was not the third major supplier of oranges to the fresh markets.








Table 2.2 World Fresh Util ization by Region Annual Annual
Growth Growth
Rate Rate
Region 1966 1976 1986 1966-86 1976-86

--- (000) Metric Tons --- Percent of Change

United States 2575 2294 2322 -0.5 0.1

Canada 0 0 0 N.A. N.A.

Latin America 5290 8336 9394 2.9 1.2

Mediterranean-EC 3689 4807 5787 2.3 1.9

E.C. 4 31 34 11.3 0.9

Rest of Western Europe 0 0 0 N.A. N.A.

Middle East/North Africa 2862 4273 5122 3.0 1.8

Rest of Africa 692 894 919 1.4 0.3

Far East 3801 5634 7619 3.5 3.1

Oceania 202 196 251 1.1 2.5

Communist Bloc 192 287 638 6.2 8.3

World Total 19307 26752 32086 2.6 1.8


r'.)









33


In that year, the United States occupied the fifth position in fresh sales worldwide.

The fourth largest producer of oranges was the Mediterranean-EC. This region's share in world production of fresh oranges was 16.4% in 1966, 13.4% in 1976, and 13.9% in 1986. This region's production grew rapidly in the last decade. This growth coincided with the incorporation into the European Community (EC) of all the countries included in this

particular region. The Mediterranean-EC dedicated 15.4% of its total orange production to the processed industry in 1986. Fresh utilization

represented 18% of total world fresh orange production. The region

occupied the third position in the fresh market in 1986.

The fifth major producer of oranges was the Middle East/North Africa, with 12.0% in 1966, 11.5% in 1976, and 11.8% in 1986. In 1986,

88.4% of total production was used fresh, giving the Middle East/North Africa region the fourth position in the world fresh orange market.

The rest of Africa was the sixth largest producer of oranges with 3.0% in 1966, 2.5% in 1976, and 2.1% in 1986. As shown in Table 2.2, out of total orange production, this region dedicated 10.1% to the processed

industry in 1986. In that year, the region occupied the sixth position in fresh sales worldwide.

The rest of world production of oranges was provided by the Communist Bloc, Oceania, and the EC with .7%, 1.0%, and .02% in 1966 and 1.5%, 1.2%, and .1% in 1986, respectively. The Communist Bloc dedicated 99.1% of total production to the fresh orange industry in 1986, while in Oceania 43.7% of total production was used for fresh consumption.









34


Trade Flow Analysis



Table 2.3 shows the quantities traded between the 11 regions for 1966, 1976, and 1986. These years were selected to illustrate changes through time. The first column of this table represents the different years, the second column and the top row of the table represent the region and the partner region names, respectively. Each of the 11 columns

depicts the quantities exported from the partner region to each region. The following two columns show the total product imported by each region,

with the first one including the intraregional trade and the other including interregional trade. Since the first and second regions consist of a single country, both columns display the same values. The last two columns exhibit the percentages associated with the previous two columns in relation to total world imports. Similarly, the last four rows of the

table contain total exports from each partner region. The first row

includes intraregional trade, the second one only interregional trade, and the last two rows show the percentages associated with total world exports.

Tables 2.4 and 2.5 contain the percentages needed to illustrate the allocation of exports, imports, and trade flows in total and among the 11 regions. Table 2.4 shows the percentages from the exporter or partner region position and Table 2.5 from the importer or region perspective. In both tables, intraregional trade was excluded, given that the major interest of the present study has to do with trade among the regions. Intraregional quantities are part of the region's production that is consumed domestically.










Table 2.3 Trade Flow Analysis for Selected Years (1966, 1976 and 1986) by Region in Relation to Partner Regions


Total Total z I W/O
Year Region US CAN LA MED-EC EC RWE ME/NA RAF RE OCE COMMB W/INTR* W/O INTRAb W/INTR INTR

66 US 0 0 23285 222 13 0 4617 0 212 0 0 26349 28349 0 8 0 8


0 11 31094 196
0 19 22334 15808


CAN 140386
190943 123772


0 1513 81
0 231 24
0 5297 18534


LA 6397 0 3191
1932 45 16110
582 3 2455


7
25

20
0
68


0
0

0
0
0


79 0
56 2
217 0


76 86

66 76
86

66 76 86

66 76
86

66
76 86

66 76
86

66 76
86

66 76
86

66 76
86

66 76 86

66 76
86


300 9706


0
0


952 1477


89
0


8536 16101 13805 59
3225 6445 23062 1616
21490 0 10773 2168


0 0 18

0
429 497


0 0
4

0 0
3326


EC 56148 62 69439 1116393 29955 1079 908131 213821
104359 2 39303 1516481 112937 817 760700 197487
9336 232 176180 2217655 173847 637 696538 176595

RWE 9889 0 2989 229717 1464 976 121588 17024
4704 0 2875 177279 3318 5067 223720 20800
1619 0 6076 321349 15663 1073 186964 17922


2
0 16

0
0
0


0 32649
0 49369

0 180501 0 225547
0 182102


0 0 9669
0 0 18145
0 451 3807


0
0
0


0
0
8


0
900
8243


23 3371 19 2398441
36 2085 1983 2736190
56 1825 11665 3464566

422 429 151 384649
7 1839 52 439661
36 276 265 551243


0 0 1024 17 0 10612 0 9799 579
0 105 29447 5060 12 461007 54334 40379 0
0 3243 16333 116 0 224078 0 29574 5650


RAF 54
7 0


FE 38555 9
145976 0
267916 14


OCE 2210 0 1420
6440 0 0
9802 0 0


95 1045 0
747 742 0
490 761 0


9 19
5 0
53 0

0 0
0 0
8 0


7324 1764
3628 5493
4294 2119


2 57
0


300 36
542


13848 15479 41505 8964
9988 1149 77821 3052
5730 8997 71939 26219


1618 23
0


COMB 4940 0 4130 166667 194 2 139601
9701 0 29292 212239 443 0 408836
0 0 12603 241790 1035 314 167406


225
0 0


11
0
0


0 9843
0 2433
5 10588


10
0
0


0 22048
0 590369
1 279010

0 10593
0 10588
0 8211

0 119341
3 238918 0 382829

0 15316
0 8896
0 20403


0 0 315555
0 58318 718829 33 22656 445839


32649 49369

180501 225547
182102

6478 2035 1352

0
717 7984

2368486 2623253 3290719

383673
434594 550170

11436 129362
54932

8829 5095 6092

77836 161097 310890

5473 6463 9815

315555 660511
423181


0.7
0.9

5.2 4.5 3.4

0.3
0.4 0.1

0.0 0.0
0.2


0.8 1.0

5.3 5.3 3.7

0.2 0.0 0.0

0.0 0.0
0.2


68.8 69.9
54.5 91.3
64.2 67.3

11.0 11.3
8.8 10.2
10.2 11.3


0.6 11.8 5.2

0.3
0.2 0.2

3.4 4.8 7.1

0.4 0.2 0.4

9.1
14.3 8.3


0.3 3.0 1.1

0.3 0.1 0.1

2.3 3.8
6.4

0.2
0.2 0.2

9.3
15.4 8.7


MED-EC


0
0 61


0
7
0


0
0
0


0
0
828


0 0 0
183 253 28
259 3167 158


ME/NA


17 25 15


0
0
0


9
178
5

595
0
324


358
924 1637

0 0 0










Table 2.3--continued.


Total Total Z Z W/O
Year Region US CAN LA MED-EC EC RWE ME/NA RAF RE OCE COM4B W/INTR" W/O INTRA W/INTR INTR

66 Total 258596 71 106571 1514557 32796 2076 1215875 264425 65780 23545 170 3484462 3386616 100.0 100.0
76 W/INTR 464094 58 119188 1937490 122551 5926 1871856 285709 142314 11150 60356 5020692 4281323 100.0 100.0
86 413042 268 229345 2833916 194960 2182 1316721 208963 113876 47301 35046 5395620 4886606 100.0 100.0

66 Total 258596 71 103380 1514557 2841 1100 1205263 262661 24275 13702 170 3386616
76 W/O INTR 464094 58 103078 1937307 9614 859 1410849 280216 64493 8717 2038 4281323
86 413042 268 226890 2833657 21113 1109 1092643 206844 41937 36713 12390 4886606

66 2 W/INTR 7.4 .0 3.1 43.5 0.9 0.1 34.9 7.6 1.9 0.7 .0 100.0
76 9.2 .0 2.4 38.6 2.4 0.1 37.3 5.7 2.8 0.2 1.2 100.0
86 7.7 .0 4.3 52.5 3.6 .0 24.4 3.9 2.1 0.9 0.6 100.0

66 % W/O INTR 7.6 .0 3.1 44.7 0.1 .0 35.6 7.8 0.7 0.4 .0 100.0
76 10.8 .0 2.4 45.3 0.2 .0 33.0 6.5 1.5 0.2 .0 100.0
86 8.5 .0 4.6 58.0 0.4 .0 22.4 4.2 0.9 0.8 0.3 100.0


'Total includes intraregional trade. 'Total does not include intraregional trade.


0S















Table 2.4


Trade Flow Analysis for Selected Years (1966, 1976 and 1986) Without Intraregional Trade "Relative Partner Region Exports


by Region"

Year Region US CAN LA MED-EC EC RWE ME/NA RAF FE OCE COMMB

- - Percentages - - -


US 0.0
0.0 0.0


0.0 19.0
7.1


54.3 0.0 41.1 0.0 30.0 0.0


LA 2.5
0.4 0.1

MED-EC 0.0
0.0 0.0


0.0
77.6 1.1

0.0 0.0 0.0


22.5 0.0 30.2 0.0 9.8 0.6


1.5
0.2 2.3

0.0
0.0 0.0

0.0 0.0
0.4


0.0
0.0 0.7

0.0 0.0 0.0

0.0 0.0 0.0


0.5 0.1 0.1

0.7 0.0
0.3

2.8 0.6 1.0


0.0 0.0 0.0

0.0 0.0 0.0

0.0
0.2 0.0


0.0 0.0
2.6 3.3 15.0 14.2


21.7 87.3 67.2 73.7 0.0 22.5 3.4 38.1 78.3 0.0 2.3 86.6 77.6 78.3 0.0


66 76 86

66 76 86

66 76 86

66 76 86

66 76 86

66 76 86

66 76 86

66
76 86

66 76 86

66
76
86

66
76 86


66 76 86


0.0 0.0
0.0


ME/NA 0.0 0.0
0.0 0.0 0.0 0.0


0.0 0.0 0.0


0.0 0.0 0.0


98.1 95.1
57.4


2.9 15.2 51.5 0.0 2.8 9.2 34.5 0.0 2.7 11.3 74.2 0.0


0.0 0.1
1.4

0.0
0.2 0.0


0.1 0.6 0.0 1.5 52.6 1.4 0.6 0.5 0.0


0.0 0.0 0.0


14.9 12.7 0.6 0.0 31.5 0.0 0.0 0.0
64.9 5.2 0.1 0.1


0.9
1.4
2.4


COMM 1.9
2.1 0.0


0.0 0.0 0.0

0.0 0.0 0.0


TOTAL 100.0 100.0
100.0 100.0 100.0 100.0


1.4 0.0 0.0


0.0 0.0 0.0


36.8 0.0 4.9 0.0
3.6 0.0


0.3 0.1 0.3

0.0 0.0 0.0


4.0 11.0 6.8
28.4 11.0 4.6 5.6 8.5 4.9


100.0 100.0 100.0


100.0 100.0 100.0


100.0 100.0 100.0


1.7
0.0 0.0

0.0 0.0 0.0

0.2 0.0
28.3


100.0 100.0 100.0


0.4 0.0 0.9

0.7
0.2 2.0

0.0 0.0 0.0

0.0 0.0 0.0


0.0 0.9 0.0 0.0 0.0 1.5 1.0 0.0 0.0 3.5 0.0 0.0

6.1 56.9 0.4 0.0 2.3 35.8 18.5 0.0 0.0 25.7 5.9 0.0


0.0 0.0
0.0


0.0 0.0 0.0 0.0 0.0 0.0


0.0 0.0 0.0 0.0 0.0 0.0 1.6 0.0 0.0


0.0
0.0 3.6

0.0 0.0 0.1


75.3 81.4 0.1 24.6 11.2
53.9 70.5 0.1 23.9 97.3
63.7 85.4 0.1 5.0 94.1


10.1 6.5 1.7
15.9 7.4 0.0 17.1 8.7 0.1


0.0 0.0 0.0

0.6 0.3
0.4

1.1 0.7 0.5

0.1 0.0 0.0


3.1 88.8
21.1 2.6 0.8 2.1


0.0 40.4 4.2 0.0
19.4 62.6 0.0 0.0 0.0 70.5 15.4 0.0


0.0 0.0 0.0

5.9
0.4 4.3

0.1 0.0 0.0


11.6 0.0 29.0 0.0 15.3 0.0


100.0 100.0 100.0


100.0 100.0 100.0


0.0 0.1 0.0

0.0 0.0 0.0

0.0 0.0 0.0

0.0 0.0 0.0


100.0 100.0 100.0


2.2 0.4 1.5


0.0 0.0 0.0


65.4 0.0 35.0 0.1 71.4 0.0


0.0 0.0 0.0

0.0 0.0 0.1


0.0 0.0 0.0

0.0 0.0 0.0


100.0 100.0 100.0 100.0 100.0 100.0


37


CAN


EC




RWE


3.8 1.0
0.4


RAF FE OCE












38


Table 2.5 Trade Flow Analysis for Selected Years (1966, 1976 and 1986)
Without Intraregional Trade "Relative Region Imports from
Partner Regions"

Year Region US CAN LA MED-EC EC RWE ME/NA RAF FE OCE COMMB Total


- - - Percentages - - -


Us 0.0
0.0 0.0


CAN LA


0.0 82.1 0.8 0.0 0.0 95.2 0.6 0.0 0.0 45.2 32.0 0.1


77.8 0.0 84.7 0.0 68.0 0.0

98.7 0.0
94.9 2.2 43.0 0.2


MED-EC N.A N.A
1.0 0.0 0.0 0.0


EC 2.4
4.0 0.3


66 76 86

66 76
86

66
76 86

66 76 86

66
76 86

66 76
86

66 76 86

66 76 86

66
76 86

66 76 86

66
76 86


0.0 0.0 0.0


2.6 0.0 1.1 0.0 0.3 0.0


ME/NA 0.1 0.0
0.0 0.0 0.0 0.0


RAF


FE


OCE




COMB


0.6 0.0 0.1 0.0 0.0 0.0

49.5 0.0 90.6 0.0 86.2 0.0

40.4 0.0 99.6 0.0 99.9 0.0

1.6 0.0 1.5 0.0 0.0 0.0


0.8 0.0 0.0 0.1 0.0 0.0 2.9 10.2 0.0


0.0 0.0 0.0


0.0 1.2 0.0 2.8 4.5 16.1


N.A N.A N.A 0.0 0.0 35.3 10.4 0.0 39.7


2.9 1.5
5.4


47.1 0.0 57.8 0.0 67.4 0.0


0.8 59.9 0.4 0.7 40.8 0.8 1.1 58.4 2.8

0.0 9.0 0.1 0.1 22.8 3.9 5.9 29.7 0.2

0.1 1.1 11.8 3.5 14.1 9.3 0.1 8.0 12.5


0.8 0.0 0.1


0.5 0.0 0.6 0.0 0.5 0.0


25.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1

1.3 52.8 0.1 4.4 32.1 0.1 3.0 57.1 0.2


0.0 16.3 0.0 0.0 0.9 0.0 0.0 19.7 0.0


0.0 0.0 0.0

0.0 0.1 0.0


0.7
2.9 3.0


0.0 0.0 100.0 0.3 0.0 100.0 0.0 0.0 100.0


4.7 8.9 7.6 0.0 0.0 100.0 1.4 2.9 10.2 0.7 0.0 100.0 11.8 0.0 5.9 1.2 0.0 100.0


0.0 0.0 0.0 0.0 1.3 0.3


0.0 0.0
1.2


N.A N.A N.A N.A 3.9 59.8 0.0 0.0 2.0 6.2 41.7 0.0


0.0 38.3 9.0 0.0 29.0 7.5 0.0 21.2 5.4

0.0 31.7 4.4 0.0 51.5 4.8 0.0 34.0 3.3


0.0 0.0 0.0


0.0 0.0 0.0

0.1 0.0 0.0


0.0 0.0 100.0
0.0 0.0 100.0 0.0 33.4 100.0

N.A N.A N.A 0.0 0.0 100.0 0.0 0.1 100.0

0.1 0.0 100.0 0.1 0.1 100.0 0.1 0.4 100.0

0.1 0.0 100.0 0.4 0.0 100.0 0.1 0.0 100.0


0.0 0.0 85.7 5.1 0.0 100.0 0.0 42.0 31.2 0.0 0.0 100.0 0.0 0.0 53.8 10.3 0.0 100.0


0.0 83.0 0.0 0.0 71.2 0.0 0.0 70.5 0.0


0.0 0.0 0.0

0.0 0.0 0.0

0.0 0.0 0.1


0.0 1.1
0.0


17.8 19.9 0.0 6.2 0.7 0.0 1.8 2.9 0.0


29.6 4.1 0.4 0.0 0.0 0.0

44.2 0.0 61.9 0.0 39.6 0.0


0.0 0.0 0.1

0.0 0.0
0.0


3.4 0.0 100.0 0.7 0.0 100.0
8.9 0.0 100.0

11.5 0.0 100.0 1.9 0.0 100.0 8.4 0.0 100.0

0.0 0.0 100.0 0.0 0.0 100.0 0.0 0.0 100.0

0.0 0.0 100.0
0.0 0.0 100.0 0.0 0.0 100.0


RWE









39

Tables 2.6, 2.7, and 2.8 correspond to the same characteristics described for Tables 2.3, 2.4, and 2.5, respectively. The difference between these two sets of tables is that the former tables present cumulative information for five years instead of yearly information. The periods considered are 1966 to 1970, 1974 to 1978, and 1982 to 1986. The discussion that follows will be based mainly on the first set of tables, because both sets draw similar conclusions. However, given that yearly

information could be biased for uncommon reasons, the results in Table 2.6 to 2.8 are useful to support general conclusions.



Partner Region Perspective



In this section, the discussion will be oriented from the exporters' viewpoint. In all cases, the relative importance of each region is set forth and then a trade flow analysis is developed.

Table 2.3 shows that the world's major fresh orange exporter was the Mediterranean-EC region. With intraregional trade considered, this region's share of total exports was 43.5% in 1966, 38.6% in 1976, and 52.5% in 1986. With intraregional trade not considered, the relative importance of the region in world trade increased to 44.7%, 45.3%, and 58% respectively. These values show the importance of this region in world trade of fresh oranges.

Table 2.4 shows that the major partner of the Mediterranean-EC was the EC region. In 1966, 73.7% of the Mediterranean-EC region's total exports went to the EC region. This percentage increased to 78.3 in 1976

and was the same in 1986. The EC region includes all EC countries except










Table 2.6 Trade Flow Analysis for Selected Periods of Five Years (1966-70, 1974-78 and 1982-86)


Total Total Z 2 W/O
Period Region US CAN LA MED-EC EC RWE ME/NA RAF FE OCE COMMB W/INTR" W/O INTR W/INTR INTR

- - Metric to U.S. - - --


Us


0
0
0


CAN 699870
872603 698262

LA 16653
9149 7866


MED-EC


44 221119 524
50 184606 570
162 177272 23466

0 15199 1489
0 2700 247
0 22155 39563


95 39255 45 75425 7 58938


3
7
9


EC 261380
388239
57895

RWE 30694
26710
8319

ME/NA 55
14826 2108


RAF


106
43 11


FE 227715
653456 1213817

OCE 10170
45305
57437


0
0
0


25
46 200

133 61 70


2 599
0 234
212 1467


83 716 51
16 251 1315
8472 3456 15792


0 35283 0 16201 0 34632


370 2142 0 4263 19 6130


0 36954 68392 73338 0 26027 36787 93671 0 98322 0 76031


0
2
0


4
77
448


812 16 320


0
0 29


1801 0
2202 22
1879 5726


127 287289 5486327 279949 3335 4843666 1074996
61 304034 7307729 570730 3660 3771137 1057439
307 693974 8117851 738464 4352 3443275 775683


2
0 18


0
0
4


452 354 607


0 24337 1033534 10465 6541 785484 103987 554
0 20179 867897 24061 23548 1126693 117572 1163 9 20512 1011829 65269 8745 956474 105954 1136

0 0 2337 276 1166 490985 4643 22659
0 23110 54432 12269 11175 2041502 144372 213855 0 125929 46085 738 1403 1581238 88428 164543


0
0
0


2160 405 4147
2239 3778 3664
506 2959 5110


9 6930 2223 341
1261 561 4411 64
105 1654 6258 349


P1 P2 P3

P1 P2 P3

P1 P2 P3

P1 P2 P3

P1 P2 P3

P1 P2 P3

P1 P2 P3


6627
977
0


24
0
3003


0
0 27


0 24178 13071 45 21629 27635 0 23544 11321


14 146 10


55 81373 41497 293406 5 61734 11860 376488 24 39975 40106 357283


0
0
0


5768 1996
193 1224
418 1207


1
5
243


47 101 685


1307 8106
10741


0
2 25


0
0
0


0
0
0


259554 259554 205837 205837 242566 242566


0 896682 896682
0 1040202 1040202 0 945144 945144


0 57418
2 84875
1173 70055


0
0 197


2658 3890 35983


18163
9450 11117

1942 3639 32527


1.4 0.8 1.0

4.8 4.2 4.0

0.3 0.3 0.3

0.0 0.0
0.2


11516 1334 12250371 11970422 65.2
10791 11940 13426114 12855384 54.1
4764 57678 13894850 13156386 59.1


1154 7224 1511

1136
10744 20389


1345 130 2809


48004 24833
93041

51174 33201 43736


1607 1998357 1991816 10.6 888 2215935 2192387 8.9
690 2180448 2171703 9.3

0 523257 32272 2.8
66 2526351 484849 10.2 4 2030865 449627 8.6


47
1
9


45473 59310 46279


32402 31675 34958


0 701553 408147
13 1134686 758198 27 1752639 1395356


0
0
0


75760 80905 106071


0.2 0.2 0.2

3.7
4.6
7.5


24586 0.4
47704 0.3 62335 0.5


1.5 1.0
1.2


P1l P2 d P3'

P1 P2 P3

P1 P2 P3


5.1
4.9 4.6

0.1 0.0 0.1

0.0 0.0
0.2

67.9 60.3 63.8

11.3 10.3 10.5

0.2 2.3
2.2

0.2 0.1
0.2

2.3 3.6 6.8

0.1
0.2 0.3


0
0
0


0










Table 2.6--continued.


Total Total 2 2 W/O
Period Region US CAN LA MED-EC EC RWE ME/NA RAF FE OCE COMMB W/INTR' W/O INTRb W/INTR INTR

- - Metric to U.S. - - -Pl COMlB 14202 0 30451 846525 1262 18 1089907 247 943 76 230 1983861 1983631 10.6 11.3
P2 36116 0 130615 1154096 1730 340 2368510 0 752 0 331498 4023657 3692159 16.2 17.3
P3 0 0 24227 912980 3138 1071 1171677 41 1 49 100226 2213410 2113184 9.4 10.3

P1 TOTAL1260848 275 633450 7374106 297248 11119 7396211 1309199 393511 115759 3218 18794944 17619617 100.0 100.0
P2 W/INTR2046454 1417 744462 9393411 614174 38852 9435844 1396911 690697 95132 344408 24801762 21321484 100.0 100.0
P3 2045724 590 1133639 10167662 830624 16043 7351754 1028514 606006 177750 160004 23518310 20614903 100.0 100.0

P1 TOTAL 1260848 275 594195 7373390 17299 4578 6905226 1296128 100105 64585 2988 17619617
P2 W/O INTR2046454 1417 669037 9393160 43444 15304 7394342 1369276 314209 61931 12910 21321484
P3 2045724 590 1074701 10164206 92160 7298 5770516 1017193 248723 134014 59778 20614903


P1 X W/INTR 6.7 0.0 3.4 39.2 1.6 0.1 39.4 7.0 2.1 0.6 0.0 100.0
P2 8.3 0.0 3.0 37.9 2.5 0.2 38.0 5.6 2.8 0.4 1.4 100.0
P3 8.7 0.0 4.8 43.2 3.5 0.1 31.3 4.4 2.6 0.8 0.7 100.0


P1 X W/O INTR 7.2 0.0 3.4 41.8 0.1 0.0 39.2 7.4 0.6 0.4 0.0 100.0
P2 9.6 0.0 3.1 44.1 0.2 0.1 34.7 6.4 1.5 0.3 0.1 100.0
P3 9.9 0.0 5.2 49.3 0.4 0.0 28.0 4.0 1.2 0.7 0.3 100.0


'Total includes intraregional trade.
Total does not include intraregional trade.
'Represents period from 1966-70.
'Represents period from 1974-78.
'Represents period from 1982-86.


I-















Table 2.7


Trade Flow Analysis for Selected Periods of Five Years, (196670, 1974-78 and 1982-86) Without Intraregional Trade "Relative


Partner Region Exports by Region"

Period Region US CAN LA MED-EC EC RWE ME/NA RAF FE OCE COMMB

- - Percentages - - -


us 0.0
0.0 0.0


16.0
3.5 27.5


CAN 55.5 0.0
42.6 0.0 34.1 0.0


LA 1.3
0.4 0.4

MED-EC 0.0
0.0 0.0


34.5 3.2
1.2

0.0 0.0 0.0


37.2 0.0 27.6 0.0 16.5 0.2


2.6
0.4 2.1

0.0 0.0 0.0

0.0 0.0 0.8


0.0 0.0
0.4

0.0 0.0 0.0

0.0 0.0 0.0


0.1 0.1
0.2


0.0 0.0 0.0


0.8 0.0 0.1 0.0 0.1 0.0

3.5 0.0 0.5 0.0 1.6 0.0


0.3 3.0
17.1


EC 20.7 46.2 48.3 74.4 0.0
19.0 4.3 45.4 77.8 0.0 2.8 52.0 64.6 79.9 0.0


0.1 0.5 6.1


72.8 23.9 59.6


0.5
0.2 0.6

0.5
0.4 1.7

0.0 0.0 0.0

0.0 0.0 0.0


0.0 0.0 0.0

5.3 2.7 0.0

0.0 0.0 0.0

0.0 0.0
0.6


2.1 1.4 2.5


0.1
0.2 0.5


0.0 0.0 0.0


73.3 2.0 0.0 29.8 13.1 0.0 30.6 8.0 0.0


0.0 0.0 0.0

0.0 0.0 0.0


0.0
0.0 0.0

0.0 0.0 0.0


0.0 0.0
2.0

0.0 0.0 0.3


70.1 82.9 0.5 17.8 44.6 51.0 77.2 0.1 17.4 92.5 59.7 76.3 0.2 3.6 96.5


4.1 3.0 1.9


14.0 60.5 0.0
9.2 55.4 0.0 10.0 70.8 0.0


0.0 0.0
3.5 0.6 11.7 0.5


0.4 0.3 0.0


FE 18.1 3.3 1.2
31.9 89.0 0.1 59.3 17.8 0.2


1.1 0.1 0.0


0.0 0.0 0.0

0.0 0.0 0.1

0.0 0.0 0.0


1.6 25.5
28.2 73.0 0.8 19.2


24.0
8.4 5.5

2.0 0.1
0.4

0.0 0.0 0.0


5.1 11.5 7.3 19.5 12.3 4.0
2.3 9.0 3.4


100.0 100.0 100.0


100.0 100.0 100.0


100.0 100.0
100.0


0.0
0.3 0.0

1.2 0.0 0.3

0.0 0.0 0.0


0.4
2.2 14.7

100.0 100.0 100.0


11.4 8.0 0.6 15.2 8.6 0.4 16.6 10.4 0.5


0.0 0.0 0.0

0.4 0.3
0.4


1.8 53.8 11.7 6.9 1.1 1.2


0.4 22.6 1.8 0.0 10.5 68.1 17.3 0.5 8.7 66.2 15.2 0.0


0.0 0.0 0.0 0.0 0.0 0.0


1.2 3.2 0.0 0.8 0.9 0.0
0.7 3.9 0.0


0.1 0.2 0.0 0.1 0.0 0.1

15.8 0.0 32.0 0.0
20.3 0.0


100.0 100.0 100.0


100.0 100.0 100.0


0.0
0.0 0.1


0.9
0.2 0.0

100.0 100.0 100.0


2.1
0.2 2.1


74.3 0.0
40.1 0.1 69.4 0.0


0.0
0.0 0.0

0.1 0.0 0.0


100.0 100.0 100.0 100.0 100.0 100.0


'Represents period 'Represents period
cRepresents period


from 1966-70. from 1974-78. from 1982-86.


42


P11
P2b
P3'

P1 P2 P3

P1
P2 P3

P1 P2 P3

P1 P2 P3


RWE 2.4 1.3
0.4

ME/NA 0.0
0.7 0.1

RAF 0.0 0.0 0.0


0.0 0.0 1.5

0.0 0.0 0.0

0.0 0.0 0.0


P1 P2 P3

P1 P2 P3

P1 P2 P3

P1
P2
P3

P1
P2 P3

P1 P2 P3

P1 P2 P3


1.6 0.0 0.0


OCE 0.8
2.2 2.8

COMMB 1.1
1.8 0.0

TOTAL 100.0
100.0 100.0


0.0 0.0 0.0


0.0 0.0 0.0

100.0 100.0 100.0


0.0 0.0 0.0

0.0 0.0 0.0












43


Table 2.8


Trade Flow Analysis for Selected Periods of Five Years (196670, 1974-78 and 1982-86) Without Intraregional Trade "Relative


Region Imports from Partner Regions"

PeriodRegion US CAN LA MED-EC EC RWE ME/NA RAF FE OCE COMMB Total

- - Percentages - - -


US 0.0
0.0 0.0


0.0 0.0 0.1


85.2 0.2 89.7 0.3 73.1 9.7


CAN 78.1 0.0 1.7
83.9 0.0 0.3 73.9 0.0 2.3

LA 91.7 0.5 0.0
96.8 0.5 0.0 70.8 0.1 0.0


P1,
P2b
P3c

P1 P2
P3

P1 P2 P3

P1 P2 P3

P1 P2 P3

P1 P2 P3

P1 P2 P3

P1 P2 P3

P1 P2 P3

P1 P2 P3

P1 P2 P3


RAF 0.3
0.1 0.0


0.2 0.0
4.2

0.0 0.0 1.9


0.0 4.3 0.0 0.0 0.4 0.0 0.0 26.0 0.0


0.0 0.0 0.0

0.0 0.0 0.0

0.0 0.0 0.0

0.0 0.0 0.0


2.4 2.4 5.3

1.2 0.9 0.9


0.0
0.0 0.1

0.0 0.0 0.0


0.0 0.0 0.0

0.0 0.0 0.0


3.3 0.0 2.5 0.0 13.2 0.0

2.6 0.2 36.1 2.1 48.6 1.4


45.8 0.0 56.8 0.0 61.7 0.0

51.9 0.5 39.6 1.1 46.6 3.0


0.0 7.2 0.9 4.8 11.2 2.5 28.0 10.2 0.2


6.7 7.1
1.4


FE 55.8 0.0 1.7
86.2 0.2 0.1 87.0 0.0 0.1


0.0 0.0 0.0

0.0 0.0 0.0

3.6 2.3 0.3


1.2 12.8 0.0 11.9 11.6 0.1 8.5 14.6 0.0


0.5 0.6
0.4


OCE 41.4 0.0 27.0 0.1
95.0 0.0 2.0 0.0 92.1 0.0 0.0 4.8


COMB 0.7
1.0 0.0


0.0 1.5 0.0 3.5 0.0 1.1


0.1 0.0 3.0

0.0 0.0 0.0


42.7 0.1 31.3 0.0 43.2 0.1


0.0 0.0 0.0

0.0 0.0 0.0

0.0 0.0 0.1


13.6 0.1 7.9 0.0 14.3 0.0

4.1 7.6 2.5 3.5 10.4 0.0


4.5 0.2 2.9


0.0 0.0 0.3


0.8
2.1 2.5

8.2 9.0 8.0

0.0 0.0
0.2


0.0 0.0 0.3

0.1 0.8 1.1

0.0 0.0
0.2


92.7 0.0 0.0 0.0 60.5 0.6 0.0 0.0 5.8 17.6 0.0 0.0


40.5 9.0 29.3 8.2 26.2 5.9

39.4 5.2 51.4 5.4 44.0 4.9


0.0 0.0 0.0


0.0 0.0 0.0


0.1 0.1 0.0


0.0 0.1 0.1 0.3 0.1 0.1


14.4 70.2 3.5 29.8 44.1 2.2 19.7 36.6 4.5


74.6 0.0 68.3 0.0 67.3 0.0


0.0 4.2 0.5 0.4 0.0 8.0


0.0 0.0 0.0

0.0 0.0 0.0

0.0 0.0 10.6

0.0 0.0 0.6

0.0 0.1
0.4

0.1 0.0 0.0

0.0 0.0 0.0

0.1 0.0 0.0


19.9 10.2 0.0 11.8 0.0 8.1 1.6 0.0 3.3 0.0 2.9 2.9 0.0 6.7 0.0


23.5 8.1 0.4 2.6 0.7 1.9

54.9 0.0 64.1 0.0 55.4 0.0


0.0 0.0 0.0 0.0 0.4 0.0


0.0 0.0 0.0


0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0


100.0 100.0 100.0

100.0 100.0 100.0

100.0 100.0 100.0

100.0 100.0 100.0

100.0 100.0 100.0

100.0 100.0 100.0

100.0 100.0
100.0

100.0 100.0 100.0

100.0 100.0 100.0

100.0 100.0 100.0

100.0 100.0 100.0


period from 1966-70. period from 1974-78. period from 1982-86.


MED-EC 0.2
0.2 0.0

EC 2.2
3.0
0.4

RWE 1.5
1.2 0.4

ME/NA 0.2
3.1 0.5


'Represents 'Represents 'Represents









44

Spain, Italy, Portugal, and Greece. The second largest partner of the

Mediterranean-EC was the rest of Western Europe. Table 2.4 shows that the relative importance of the rest of Western Europe in the MediterraneanEC's total exports decreased from 15.2% in 1966 to 11.3% in 1986. The third major partner of the Mediterranean-EC region was the Communist Bloc. This region accounted for 11.0% of the Mediterranean-EC's total exports in 1966 and 8.5% in 1986. Exports to the rest of the partners were small,

but exports to the United States and Canada have increased in the last few years.

The second major exporter region of the world was the Middle East/North Africa. As opposed to the Mediterranean-EC region, this one has been losing its share of the market in the last 20 years.

Participation in total world exports increased from 34.9% in 1966 to 37.3% in 1976 (see Table 2.3). Nevertheless, the region's share of the export

market decreased to 24.4% in 1986. Examining exports without considering intraregional trade shows that this region was losing its share of the external market faster than its own regional market share. Table 2.3 shows that interregional percentages of the Middle East/North Africa decreased from 35.6 in 1966 to 22.4 in 1986.

The Middle East/North Africa region's major partner was the EC region. In 1966, 75.3% of total interregional exports from the Middle East/North Africa countries went to the EC countries (see Table 2.4). This percentage has since been decreasing, and in 1986 it represented only 63.7. In 1976, the percentage was lower, mainly due to an important shift of exports to the Communist Bloc. The second and third largest partner positions of the Middle East/North Africa region were closely shared by









45

two regions, the rest of Western Europe and the Communist Bloc. Exports

from the Middle East/North Africa region to the rest of Western Europe represented 10.1% in 1966, 15.9% in 1976, and 17.1% in 1986. Exports to the Communist Bloc region represented 11.6%, 29.0% and 15.3% in the same years. Exports from the Middle East/North Africa region to the United States and Canada decreased from 1966 to 1976; however, exports to these countries have been increasing in recent years.

United States exports increased at a rate of 2.25% a year from 1966

to 1986. In relative terms, United States participation in world trade of fresh oranges showed about the same level as 1966. Total United States trade represented 7.4% of total world trade in 1966, increased to 9.2% in 1976 and decreased to 7.7% in 1986 (see Table 2.3). With intraregional trade not considered, the relative importance of the United States trade in the world trade increased. Table 2.3 shows that United States trade represented 7.6% in 1966, 10.8% in 1976, and 8.5% in 1986. In relative

terms, these percentages show the United States to have been the third largest exporter, exceeded by the Mediterranean-EC and the Middle East/North Africa regions. In absolute terms, the Mediterranean-EC and the Middle East/North Africa exports were 6.9 and 2.6 times the United States exports, respectively, in 1986.

The relative importance of the United States partners has been changing through the years. The major United States partner in 1966 was Canada. Exports to Canada accounted for 54.3% of United States fresh exports that year (see Table 2.4). The second largest partner was the EC with 21.7% and the third largest was the Far East with 14.9%. Latin

America, rest of Western Europe, Oceania, and the Communist Bloc absorbed










46

2.5%, 3.8%, .9% and 1.9%, respectively. By 1976, Canada represented 41.1%, the EC region stayed almost the same, and the Far East region absorbed 31.5% of the total United States fresh exports. In 1986, United

States exports to the Far East reached 64.9% of its total volume, representing an important shift of the United States export partners. The second largest partner was Canada, with 30% of the total volume. The EC

was no longer significant for the United States exports, given that it represented only 2.3% in the same year. The rest of the regions also decreased their participation relative to previous years.

Table 2.3 shows that exports from Latin America have doubled in absolute terms in the last two decades. However, exports did not increase from 1966 to 1976, which implies that the increase took place during the last ten years. Total world trade participation of Latin America passed from 3.1% in 1966 to 2.4% in 1976 and 4.3% in 1986. With intraregional trade excluded, its participation in world fresh trade increased to 4.6% in 1986. Note that Brazil generally does not export fresh oranges.

The major export market for Latin American product has been the EC region, which in 1966 absorbed 67.2% of the total product exported (see Table 2.4). This percentage decreased to 38.1 in 1976 and increased to 77.6 in 1986. The United States was the second largest market for Latin America exports in 1966, with 22.5% of the total export level. This

percentage increased slightly in 1976 and decreased to 9.8 in 1986. The third largest market for Latin America was the Communist Bloc, which took most of the reduction shown in the EC region during the 1970s and part of the United States share in the 1980s. Latin American exports to Canada and the Middle East/North Africa have been increasing, especially in the










47

last decade. Some countries of the Middle East/North Africa region utilized the fresh product to produce Frozen Concentrated Orange Juice (FCOJ). The rest of Western Europe is another important market for the Latin American product. An interesting issue about this region is that

its percentage of participation has not changed significantly over the years.

The rest of Africa used to be the third largest exporter of the world, but its share of the market has been decreasing, especially from

1976 to 1986. With intraregional trade considered, this region's share of the world's export market was 7.6% in 1966, 5.7% in 1976, and 3.9% in 1986 (see Table 2.3). Given that most of its trade was external, these percentages increased to 7.8, 6.5, and 4.2, respectively, when only interregional trade is considered. The region's share of the market indicates that it occupied the fifth position relative to the other regions in 1986.

The major export market for the rest of Africa was the EC region. This region represented 81.4%, 70.5% and 85.4% of the total rest of Africa exports in 1966, 1976, and 1986, respectively (see Table 2.4). The second most important partner was the rest of Western Europe, which absorbed 6.5%, 7.4%, and 8.7% of total exports in the same years. The rest of Africa exports to Canada represented 6.1% in 1966 but decreased to 0% in 1986. In that year, the Far East region was the third largest market for the rest of Africa. Exports to that region represented 5.9% in 1966, .4% in 1976, and 4.3% in 1986. During the 1970s, exports from the rest of Africa to the Middle East/North Africa region increased sharply and later decreased.










48

The Far East includes the Asian continent except for the Middle Eastern countries. This region's intraregional trade was more intense than its interregional trade. Its participation in total exports was higher in relative terms with intraregional trade considered (see Table 2.3). The Far East region's share of total world exports was 1.9% in 1966, 2.8% in 1976, and 2.1% in 1986. With intraregional trade not

considered, these percentages decreased to .7, 1.5, and .9, respectively. China is one of the countries with the potential to become an important

exporter in this particular region and worldwide. Given the level of exports of this region, it occupied position six among the 11 regions considered in 1986.

The Far East major partners were Canada and the Middle East/North Africa. In 1966, Canada was the major partner with 56.9% of total Far East exports. In 1986 this percentage decreased to only 25.7. The share of the Middle East/North Africa increased from 40.4% in 1966 to 70.5% in

1986 (see Table 2.4). The United States participation has been increasing slowly through the two decades, passing from .9% in 1966 to 3.5% in 1986.

With these exceptions, the rest of the regions were not major partners of the Far East.

Oceania held the seventh place relative to the rest of the regions

considered. Table 2.3 shows that, with intraregional trade included, the percentages representing this region's participation in world total exports were .7 in 1966, .2 in 1976, and .9 in 1986. Excluding

intraregional trade, these percentages decreased slightly to .4, .2 and .8, respectively. This suggests that Oceania's intraregional trade was relatively more important than its external trade.










49

In 1966, the major partners of Oceania included the Far East and the EC regions with 65.4% and 24.6% of exports, respectively (see Table 2.4).

In 1986, the main partners were the Far East and the Middle East/North Africa regions with 71.4% and 15.4%, respectively. The EC share decreased from 23.9% in 1976 to 5.0% in 1986. Canada's share of Oceania's exports was .4% in 1966, 18.5% in 1976, and 5.9% in 1986. Similarly, the rest of Western Europe sharply increased its participation in the 1970s from 3.1% in 1966 to 21.1% in 1976. In the 1980s, this percentage decreased again to the 1966 level. The rest of Africa was another important partner of Oceania's exports with 2.2% in 1966, .4% in 1976, and 1.5% in 1986.

EC production of oranges is relatively small and mainly concentrated in southern France. Nevertheless, trade data reveal some intraregional trade and a small amount of external trade. Including intraregional

trade, world export participation of the region was .9% in 1966, 2.4% in

1976, and 3.6% in 1986 (see Table 2.3). With intraregional trade

excluded, these percentages decreased to .1, .2, and .4, respectively. This indicates that the EC region occupied position number eight relative to the rest of the regions with regard to world fresh orange export share in 1986.

The main partner of EC exports is the rest of Western Europe, with

51.5% in 1966, 34.5% in 1976, and 74.2% in 1986 (see Table 2.4). Interestingly, in 1966, 36.8% of the EC's total exports were directed to the rest of Africa and, in 1976, 52.6% were sent to the Middle East/North Africa. In both cases, the participation of these regions rapidly decreased to 3.6% and .5%, respectively, in 1986. The rest of the regions were not significant partners to the EC except for the Communist Bloc.










50

This region's participation was 6.8% in 1966, 4.6% in 1976, and 4.9% in 1986.

The Communist Bloc has been increasing its participation in world

total exports from almost zero in 1966 to .6% including intraregional trade and .3% excluding intraregional trade in 1986 (see Table 2.3). With these percentages, the Communist Bloc held position number nine concerning world exports of fresh oranges in 1986.

The Communist Bloc region has two principal partners, the EC and the rest of Western Europe (see Table 2.4). The EC and the rest of Western

Europe absorbed 11.2% and 88.8% in 1966, 97.3% and 2.6% in 1976, and 94.1% and 2.1% in 1986, respectively. The rest of the Communist Bloc exports in 1986 went to the Latin America region.

Finally, Canada and rest of Western Europe are not exporters of fresh oranges. Weather conditions in these regions do not allow them to

produce oranges (see Table 2.3). However, trade data revealed some

exports out of those regions. Most of that trade was related to reexports reported as exports.



Region Perspective



In this section, the discussion will be based on importers'

viewpoint. Once again, the relative importance of each one of the regions will be set forth and then trade flows will be discussed.

As shown in Table 2.3, the leading importer of fresh oranges including intraregional trade was the EC region, with shares of 68.8% in 1966, 54.5% in 1976, and 64.3% in 1986. Considering only interregional










51

trade, the shares for the same years were 69.9%, 61.3%, and 67.3%. These percentages show that EC trade with other regions was more important than its own within-region trade.

Table 2.5 shows that, in 1966, 47.1% of the total EC imports came

from the Mediterranean-EC region, while in 1986 this percentage reached 67.4. The second largest exporter to the EC region was the Middle East/North Africa region with 38.3% in 1966 and 21.2% in 1986. The third largest exporter was the rest of Africa region with 9% in 1966 and 5.4% in 1986. Other important exporters to the EC region included Latin America

and the United States. These two regions' EC market shares were 2.9% and 2.4% in 1966, and 5.4% and .3% in 1986, respectively. The major portion

of the EC market growth has been captured through the years by the Mediterranean-EC region. Latin America was the only other region whose share of the EC market grew in the last two decades. The rest of the regions' exports to the EC were minimal.

The second largest importer of fresh oranges was the rest of Western Europe with 11.0% in 1966, 8.8% in 1976, and 10.2% in 1986 (see Table 2.3). With only interregional trade used, the percentages increased to 11.3, 10.2, and 11.3 for the same years.

As shown in Table 2.5, the leading exporter to the rest of Western Europe region was the Mediterranean-EC region, with 59.9% in 1966, 40.8% in 1976, and 58.4% in 1986. The second largest exporter to this region

was the Middle East/North Africa with 31.7%, 51.5%, and 34.0% for the same years. Another important exporter to the rest of Western Europe was the rest of Africa with 4.4% in 1966, 4.8% in 1976, and 3.3% in 1986. Even

though the EC was not a major producer of fresh oranges, it was the fourth










52

major exporter to the rest of Western Europe region in 1986. Latin

America increased its participation in this market passing, from .8% in 1966 to 1.1% in 1986. The United States, once the fourth major exporter

to the region, was a very small participant in the rest of Western Europe fresh orange trade. The rest of the regions' exports to the rest of Western Europe were relatively small.

The third largest importer was the Communist Bloc with 9.1% in 1966, 14.3% in 1976, and 8.3% in 1986 (see Table 2.3). With only interregional trade considered, the percentages increased to 9.3, 15.4, and 8.7, respectively.

The major supplier of fresh oranges to the Communist Bloc was the Mediterranean-EC with 52.8% in 1966, 32.1% in 1976, and 57.1% in 1986 (see Table 2.5). The second largest exporter to the region was the Middle East/North Africa with 44.2%, 61.9%, and 39.6%, respectively. The third

largest exporter was the Latin America region with 1.3%, 4.4%, and 3.0% in the same years. During the 1970s, the Communist Bloc countries

drastically increased their consumption and the deficit was mainly supplied by the Middle East/North Africa region. During the 1980s,

consumption went back to the original trend. The United States exported 1.6% of the Communist Bloc's total imports in 1966, and 1.5% in 1976. In

1986, the United States did not export fresh oranges to the Communist Bloc region. The rest of the regions were not very important relative to total Communist Bloc's imports of fresh oranges.

The three principal regions mentioned above have been relatively stable in their participation in the world's fresh orange imports in the










53

last two decades. As a whole, they represented 90.5% in 1966, 86.9% in 1976, and 87.3% in 1986 of total world imports (see Table 2.3).

The Far East has been consistently growing as an importing region in the last two decades. It passed from 3.4% in 1966 to 4.8% in 1976, and to 7.1% in 1986 (see Table 2.3). With intraregional trade omitted, these percentages decreased to 2.3, 3.8 and 6.4, respectively. This shows that trade among countries belonging to the Far East region was important relative to the rest of the world's trade with the same region.

The United States was the leading exporter to the Far East region in the period considered. Its exports represented 49.5% in 1966, 90.6% in 1976, and 86.2% in 1986 (see Table 2.5). The second exporter to the Far East was Oceania, with 11.5%, 1.9%, and 8.4%, respectively. The third exporter was the rest of Africa. However, its participation has been decreasing, from 19.9% in 1966 to 2.9% in 1986. The fourth exporter to the Far East was the Middle East/North Africa region, whose participation

has also decreased from 17.8% in 1966 to 1.8% in 1986. The MediterraneanEC region represented the fifth exporter to the Far East region, with a

consistent participation in the market of only .5%. Latin America's share of the market decreased from .8% in 1966 to .1% in 1986. It is clear from these numbers that the United States was the only exporting region whose market share grew in the Far East.

The Middle East/North Africa region was another significant importer of fresh oranges. Its participation grew from .6% in 1966 to 11.8% in 1976, but decreased later to 5.2% in 1986 (see Table 2.3). The table

shows that the percentages excluding intraregional trade were .3 in 1966,










54

3.0 in 1976, and 1.1 in 1986. Therefore, the principal trade of this region was among the countries constituting the region.

The major exporter to the Middle East/North Africa countries was the Far East region, with 85.7% in 1966, 31.2% in 1976, and 53.8% in 1986 (see Table 2.5). The second principal exporter to this region was the Mediterranean-EC with 9.0% in 1966, 22.8% in 1976, and 29.7% in 1986. The third exporter was Oceania with 5.1% in 1966, 0% in 1976, and 10.3% in 1986. Even though exports from Latin America appear insignificant compared to other exporters to the Middle East/North Africa, they have been growing very rapidly in the last few years, passing from 0% in 1966 to 5.9% in 1986.

Canada was an important importer of fresh oranges. During 1966, its imports represented 5.2% of the world's trade. This percentage decreased to 4.5 in 1976 and 3.4 in 1986 (see Table 2.3). With only interregional trade considered, these percentages increased slightly to 5.3 in 1966 and 1976, and 3.7 in 1986.

The major supplier of fresh oranges to Canada was the United States, with 77.8% in 1966, 84.7% in 1976, and 68.0% in 1986 (see Table 2.5). The second largest exporter to Canada was the Middle East/North Africa region, holding 4.7%, 1.4%, and 11.8% for these years. The third major exporter

was the Far East region with 7.6% in 1966, 10.2% in 1976, and 5.9% in 1986. In the last few years, the Mediterranean-EC region, whose share was insignificant during the 1960s and the 1970s, have increased their participation in this market. In 1986, Mediterranean-EC supplied 10.2% of the Canadian market. Latin America increased its share of the market from .8% in 1966 to 2.9% in 1986. Similarly, Oceania increased its










55

participation in recent years, passing from 0% in the 1960s to 1.2% in 1986. The rest of the regions were not very important with regard to exports to the Canadian region.

The rest of the regions represented small percentages of total imports in the world's fresh orange industry (see Table 2.3). The United States import share was .8% in 1966, .7% in 1976, and .9% in 1986. These percentages changed very little if only interregional trade were

considered. Major exporters to the United States were Latin America with 45.2%, Mediterranean-EC with 32.0%, and Middle East/North Africa with 19.7% in 1986. The Mediterranean-EC only had .8% and .6% share of the United States market in 1966 and 1976, respectively, indicating that Mediterranean-EC region's participation in the United States has been growing rapidly in the last decade.

The Oceania portion of total world imports was .4% in 1966, .2% in 1976, and .4% in 1986 (see Table 2.3). These percentages switched to .2 each reported year if only interregional trade were included.

The exporter with the major portion of the Oceania region's market

was the United States, with 40.4% in 1966, 99.6% in 1976, and 99.9% in 1986 (see Table 2.5). Middle East/North Africa and Latin America regions used to have an important share of the Oceania market, reaching 29.6% and 25.9%, respectively in 1966. These regions lost their portion of the market to the United States in the 1970s. The rest of the regions were not major exporters to Oceania.

The rest of Africa's share of total world imports was .3% in 1966,

and .2% in 1976 and 1986 including intraregional trade (see Table 2.3).










56

If intraregional trade were excluded, these percentages changed to .1 for the last two years reported.

The four major suppliers of fresh oranges to the rest of Africa were the Middle East/North Africa with 83.0% in 1966, 71.2% in 1976, and 70.5% in 1986; the EC with 11.8%, 9.3%, and 12.5%, respectively; Oceania with

3.4%, .7%, and 8.9%, respectively; and the Mediterranean-EC with .1%, 14.1% and 8.0%, respectively (see Table 2.5). The United States share of the rest of Africa market was .6% in 1966. However, the United States lost its share totally by 1986.

Latin America's portion of total world imports was .3% in 1966, .4% in 1976, and .1% in 1986 (see Table 2.3). Given that most of its trade was among countries of the region, these percentages decreased to .2 in 1966 and to 0 in 1976 and 1986. Imports in Latin America came from the United States in the 1960s and 1970s (see Table 2.5). In 1986 the United

States share was only 68.0% of total imports. The rest of the product came mainly from the Communist Bloc with 33.4%, the EC region with 16.1%, and the Mediterranean-EC with 4.5%.

The Mediterranean-EC region has only a small share of total world's

fresh orange imports. Imports reached .2% in 1986 with and without considering intraregional trade (see Table 2.3).



Conclusions



In summary, it is possible to describe most of the world production and trade flows of the fresh orange industry with few regions. On the production side, the major producers of oranges were Latin America, Far










57

East, United States, Mediterranean-EC, and Middle East/North Africa. Latin America and United States had high percentages of processed utilization. The Far East had an intense within-region trade. Therefore, as shown above, large orange productions did not necessarily imply high participation in interregional fresh orange trade.

On the supply side, the major exporters were the Mediterranean-EC,

Middle East/North Africa and United States. However, United States share of total fresh exports was small compared to the other two regions. The

Mediterranean-EC region includes Spain, Greece, Italy, and Portugal. The Middle East/North Africa includes the Middle East and the North African countries.

The Middle East/North Africa region has been losing its share of the world market to the Mediterranean-EC in the last few years. It is clear that the leading world exporter was the Mediterranean-EC region. The United States, once a major exporter to the European markets, shifted to the Far East and Oceania markets. United States share declined in most markets, with the exceptions mentioned above. Finally, the Latin America region increased its share of the total market in last two decades.

On the demand side, the major importer was the EC region which includes the EC countries except for Spain, Greece, Italy, and Portugal.

The second largest importer was the rest of Western Europe, which represents the rest of the Western European countries. The third largest

importer was the Communist Bloc, among which the major importers were the Eastern European countries.
















CHAPTER 3
LITERATURE REVIEW



International Agricultural Trade Models



Several models or approaches to study international trade have been

developed in the last two decades. These models were developed mainly due to the need for knowledge and understanding of increasing world trade. Thompson (1981) presented an interesting survey of new developments in international agricultural trade models. In his document, each model was reviewed in three sections: a historical survey, an evaluation, and a summary and implications section. The different modeling approaches were divided into two basic groups determined by the number of regions considered in the model. The two groups were two-region models and multiple-region models of agricultural trade. The latter was further divided into three groups: non-spatial price equilibrium, spatial price equilibrium, and trade-flow and market-share models.

A different classification system for international trade models was developed by Thompson and Abbott (1982). Each modeling approach was

grouped based on the assumptions made about the homogeneity of the commodity traded. The two major categories identified in their research

were single homogeneous commodity models and multiple-product models. The single homogeneous commodity models were divided into three groups: nonspatial price equilibrium, spatial price equilibrium, and two-region 58









59

models. The multiple-product trade models were also divided into three groups: general equilibrium (including agricultural and non-agricultural products), multiple related commodity products (including only

agricultural products), and differentiated product models (differentiated by place of origin). The two-region and the general equilibrium models

were special cases of non-spatial price equilibrium models. Thompson and Abbott's (1982) classification procedure added important insights into the discussion about new developments in international agricultural trade models. The major contribution was their extensive treatment of and emphasis on the characteristics of the products traded and how consumers perceived them.

In the following discussion, Thompson's (1981) approach will be followed. His classification was basically the same as the one presented in Thompson and Abbott's (1982) investigation. The most important

difference between the two studies was the emphasis that the latter researchers gave to product differentiation.

The first type of model covered by Thompson (1981) was the two-region model. The model divided all countries of the world into two groups, the country of interest and the rest of the world. This version was basically a domestic agricultural sector model enlarged with exogenously driven exported or imported quantities. Export equations or excess demand equations were developed for the rest of the world. The model included linkages between the domestic and world prices to reflect

the simultaneous determination of domestic consumption, supply, and prices with the rest of the world. The models did not take into consideration trade flows (destination) but instead accounted for the net trade between









60

the country of interest and the rest of the world. They did not provide information on demand and supply for individual foreign regions or on the share of the market that any particular country has in a specific region.

Without knowledge of the structure of supply and demand in each major trading region, it is impossible to say how the excess demand function will change given an exogenous shock or a change in policy. It is then very difficult under the two-region models to evaluate the impact

of shocks or policies in a given country. Such models do, however,

provide a good framework to analyze domestic farm and trade policies.

According to Thompson (1981), multiple-region world trade models were developed to answer broader questions regarding the impact of exogenous shocks and policy changes for trading regions in the world. They also provide information about the market share of each region by destination. The non-spatial price equilibrium models treat the

interrelations among trading regions by assuming that the world market price is determined simultaneously by the demand-supply balance in all trading regions such that the world market clears. Solution of the model gives the world market prices and the net trade for each region, but it does not provide any information on source or destination of trade flows.

Multiple-region world trade models allow for the introduction of transportation costs, tariffs and nontariff barriers, and other policy variables through the price linkage equations. These models are for many

reasons an improvement over the two-region models, since they endogenously determine the demand and supply in each of the trading regions. However, they usually have an important drawback. The price linkage frequently used is not consistent with the spatial price equilibrium theory. This is









61

so because in some cases a unique world price is assumed and in other cases a base country or region price is used and linked with the rest of the regions by the transportation cost. The model ignores the fact that some regions may not trade at all with the base region. Solutions to these models are obtained by solving an econometric simultaneous system of equations.

The second type is the spatial price equilibrium models. These models differ from the non-spatial and the two-region models in the fact that they consider endogenous trade flows and market shares. Prices are

linked only between those pairs of countries that actually trade with each other. The rest of the characteristics are similar to the ones mentioned for the non-spatial equilibrium models, except for the solution method. They usually follow a quadratic programming procedure for estimation.

None of the models described above can replicate all of the observed trade flows since they are designed to predict trade flows of homogeneous products (Grennes et al., 1977 and 1978; Thompson, 1981; and Thompson and

Abbott, 1982). If products are homogeneous, then price differences between regions are given only by transportation costs and trade barriers. Products may not be perfectly homogeneous and may be differentiated by country of origin. Therefore, prices may vary between regions for reasons other than transportation costs and trade barriers.

A serious formulation of a spatial price equilibrium model will be

to determine trade flows exclusively by minimizing the transportation cost. According to Grennes et al. (1978) "nearly everyone who has employed spatial models concedes that the world does not behave this way". This situation is intuitively appealing, and indeed there is enough









62


empirical evidence that this may be the case for wheat (Grennes et al., 1977 and 1978; Thompson, 1981) and other agricultural products. Spatial

price equilibrium models have few capabilities except for the weak and incomplete explanation of trade flows given the problems mentioned above.

Trade-flow and market-share models are the third type of multiple-region models considered by Thompson (1981). These models were

developed to account for the observed variation in trade flows more adequately than do the spatial equilibrium models.

Taplin (1967) and Johnston (1976) in a partial sense surveyed world

trade models concerned primarily with trade flows. They studied the ones that analyzed the structure of world trade and the short-run trade fluctuations among countries. In his paper, Taplin classified them in two categories: the ones that have separate functions for total exports and total imports but do not attempt to estimate the individual flows between countries; and the ones that look at individual flows directly.

In the first part, Taplin's discussion went from an import-export matrix developed by the League's Network of World Trade (1942), passing by Woolley's (1965) transactions matrices on payments for trade, services, and capital flows, to Beckerman's (1956) input-output approach. These studies provided important insight into the structure of the international economy. However, they did not represent a formal model where hypotheses could be tested, measured or forecasted.

The second part of Taplin's investigation continued with a survey

covering other studies (Tinbergen, 1962; Linnemann, 1966; Waelbroeck, 1962 and 1965) where individual trade flows (from the import-export matrix) between countries were considered to be a function of income, population,









63

trade preference, and distance variables. In these models, prices were normally omitted given that cross-section models were used, with data at the same point of time. Prices were assumed not to change. These models

did not capture shifts or changes of trade which might develop in the long run because of more complicated interrelationships among prices, income, and imports.

Taplin continued his study by reviewing four different transmission models that tried to establish the main relationships between the level of domestic economic activities in the various countries and their

international transactions. The four models surveyed and reported by Taplin were: Metzler (1950) who focused on changes in investment; Neisser

and Modigliani (1953) on income and capital flows; Polak (1954) on autonomous investment and price changes; and Rhomberg (1966) and Rhomberg

and Boissonneault (1964) who focused on income, prices, and capacity. Rhomberg and Boissonneault (1964) developed a trade-flow and market-share model that considered three regions, the United States, Western Europe, and the rest of the World. An aggregated commodity called merchandise, including all commodities traded among the regions, was defined and used to estimate income and price elasticities.

Taplin concluded that a model was needed that incorporates the type

of disaggregation possible with a constant share approach and the flexibility and economic content provided by a transmission model. Taplin also proposed a three stage procedure to accomplish his recommendations: consider the import demand for 10 to 12 regions for six good's classes;

determine what share of the import market the other countries have in supplying the given country's imports; the export-supply schedules should









64

tie into the model. These conclusions provided guidelines for continued research in trade-flow and market-share models during the latter part of the 1960s.

Trade-flow and market-share models are based on the idea that products are differentiated by country of origin. Three alternative

solution approaches exist: mechanical procedures that transform trade flow matrices from one year to the next without regard for price; econometric models designed to explain one or more elements of the trade flow matrix (an example is Ward, 1976); and modified spatial equilibrium models that take into account that products are differentiated by country of origin.

The latter implies that the elasticity of substitution is less than infinite (examples are Hickman and Lau, 1973; Grennes et al. 1978; Johnson et al., 1979; Sarris, 1983 and 1984; Sparks, 1987; Penson and Babula, 1988; Deardorff and Stern, 1986). None of the examples, except for Sparks and Deardorff and Stern, used a simultaneous equation approach to estimate the world trade model. Hence, the results obtained suffer from simultaneity bias (Maddala, 1977, p. 231-251).

The modified spatial equilibrium model approach has been used to estimate a total import demand equation for each importing region and separate market share equations for each region. This approach rests on

the assumption that products have unique characteristics distinguishing them from similar products of other exporters. Most studies mentioned above have proved that consumers view goods of the same kind from different suppliers as imperfect substitutes. This is especially true in agricultural trade, where quality and variety characteristics, national

factors, variations in harvest time, and monopolistic competition are









65

normally present. Therefore, different countries faced different

elasticities that may vary when market shares differ (Grennes et al. 1977).

Armington (1969a, 1969b, 1970a, 1970b, 1973) developed the theory for market share demand studies which considered goods differentiated by place or origin. Most market-share demand studies have used this theory

because of important variations obtained in price and income elasticities among suppliers in the foreign markets (examples are Sirham and Johnson,

1970; Ito et al., 1988; Lin et al., 1988). Later, Rhomberg (1970)

concluded that a complete demand and supply model for a world trade and payments model could also be developed following Armington's approach.

Armington assumed a weakly separable utility function, so that consumers' decision process may be viewed as occurring in two stages (Varian, 1984). Equations can be derived that relate a particular trade

flow between two countries to the importing country's index of total imports and a price ratio or relative price. Each region's market share of a commodity may be affected by changes in the size of the market of destination even if relative prices remain unchanged. The price ratio is between the price of the exporting country and an average of the import

prices of the same type of product from other origins in the importing country. The total quantity of a commodity to be imported is first determined, and then the quantity is allocated among the competing suppliers.

Armington assumed that the total quantity of the product imported is a constant elasticity of substitution (CES) index of the quantities imported from the regions of origin. The assumption was made to simplify








66

the model and reduce the number of parameters to be estimated, especially when the number of trading regions is large.

Under these assumptions, the cross-price elasticities between all pairs of regions need not be estimated, since they can be obtained from the estimated price elasticities and the estimated "constant" elasticity of substitution (Leamer and Stern, 1970). The CES assumption is highly

restrictive. In fact, the model assumes that products are differentiated by country of origin and at the same time assumes that the elasticities of substitution are constant and equal between all pairs of exporting regions in all markets. Arrow et al. (1961) developed the general properties of the CES production function.

Winters (1984) criticized these assumptions on the use of the CES

functional form. Winters accepted the initial assumptions of separability among commodities (e.g., food and machinery), while within each commodity group the domestic and foreigner suppliers were treated as non-separable. However, the adoption of the CES made them homothetic and separable over

all pairs of sources. Winters concluded that "the separability of domestic and foreign supplies essentially slipped in by the back door,..., rather than as a necessary consequence of two-stage budgeting". Winters' empirical results rejected the assumptions of homotheticity and

separability after testing them using the AIDS model (Deaton and Muellbauer, 1980).

Alston et al. (1990) has recently criticized Armington's approach.

His research shows that the assumptions of separability and homotheticity with trade data for the cotton and wheat markets were also empirically








67

rejected using the AIDS model. They also recognized the problems with the AIDS model.

The restrictiveness of the assumptions were recognized earlier by Resnick and Truman (1973). They relaxed several assumptions of Armington's model, especially the one that the elasticities of

substitution need to be constant and identical between all pairs of suppliers to each market. They specified a multi-stage decision process instead of Armington's two-stage procedure. Again, total imports were determined first and then imports from a sequence of successively smaller geographic regions were determined.

Artus and Rhomberg (1973) also recognized the problem with the assumptions and replaced the CES index function. They used the constant

ratios of elasticities of substitution and homogeneous (CRESH) index functions developed by Mukerji (1963) and Hanoch (1971).

Sparks (1987), following Artus and Rhomberg's work, used the constant ratio of elasticity of substitution (CRES) index which makes the model somewhat less restrictive. This assumption implies that the

elasticity of substitution for all the products in a market or region i vary by a constant proportion, but the substitutability between products need not be the same. This assumption increases the flexibility of the model but also increases the computational complexity. The model was applied to a highly aggregated commodity (vegetables). In this case, the

basic assumption of Armington's model, goods distinguished by place of production, seems less applicable given that the aggregated commodity will be composed of several goods. The model explained that trade flows can








68

reflect differences due to commodity composition as well as differences due to country of origin.

Trade-flow and market-share models represent a major improvement over the other models developed to study international trade, since they

can more readily depict observed trade flows. The assumption that

products are differentiated by country of origin and prices may vary between regions for reasons other than transportation costs and trade barriers is intuitively appealing. Furthermore, Armington's

simplification by the introduction of an import quantity index function is, in many cases, a necessary condition to operationalize the model and obtain as much information as possible from the trade flows. As will be shown later, the Armington model provides several practical solutions for dealing with a large number of equations and parameters.



Trade Models: The Orange Industry



The fresh orange industry has been studied many times, usually in

the context of national markets. A few studies have been developed in the international trade of fresh oranges. In addition, none of the research

developed so far considered a complete world trade model for this particular good. Most of the studies have been either partial or descriptive. One of the earliest international trade documents is a descriptive study developed by the U.S. Department of Commerce (1940), which showed citrus world production and trade statistics and trends.

Before the 1950s, little demand estimation for citrus fruits existed. More attention has now been devoted to this economic area by the








69

Florida Agricultural Experiment Station and by the Florida Department of Citrus (FDOC). As reported by Chapman (1963), the first major step in this area was the work on experimental pricing techniques applied to the orange demand analysis developed by Godwin and Powell during the 1950s.

Chapman (1963) and Godwin et al. (1965) developed a study on demand and substitution relationships for California and Florida Indian River and Interior Valencia fresh orange market. Their research was basically concentrated in the U.S. market and focused on questions regarding ownprice elasticities and cross-price elasticities between the three regions' in the Grand Rapids, Michigan market.

Dean and Collins (1967, 1968) studied the effects of the European

Community (EC) tariff policies in a model of world trade for fresh oranges. Their paper included a summary of world production, consumption, and trade of fresh oranges. Projections of orange production and

consumption, estimates of transportation costs, possible future tariffs, and income and price elasticities of demand in the EC for 22 regions were

also included. The price elasticities were estimated at the import demand level, i.e. at the location of consumption, but before retail margins were added to the wholesale price. Transportation costs as well as

tariffs and any special import taxes were included in determining the wholesale price level. Using a transportation model analysis, the impact of possible future tariff policies in the EC was procured on producer and

consumer prices in each of the major countries and on trade flows. Finally, using the results obtained in the different tariff scenarios, the welfare effect on consumers and producers was also captured. The major implication of this document and the ones by Chapman (1963) and Godwin et








70


al. (1965) is that it is possible to argue that consumers actually see products of the same kind coming from different regions as non-perfect substitutes.

Weisenborn et al. (1970) estimated the price-quantity relationships at the processor or packer FOB level in foodstore, institutional, and export market channels for Florida oranges and orange products. The

products included fresh and processed oranges. As reported by Weisenborn et al., virtually no previous demand analysis had been completed for the institutional and export sectors at that time.

Prato (1970) used the concept of separability to separate food from

non-food items. Once the demand equation was defined for only food items, he showed that the correlation between first differences in the prices of orange products and first differences in the prices of each of the other food items were not significantly different from zero. Therefore,

individual demand equations for fresh and processed oranges without the introduction of other food item prices could be defined. As reported by

Prato, research findings appear reasonable when compared with estimates derived using other and more conventional approaches.

Tang (1977) studied the world demand for United States fresh grapefruit in four markets: the United States, Japan, Europe, and Canada.

In his research, Tang identified and measured the effects of the different factors that affect domestic and export demand in order to determine the optimal allocation of United States fresh grapefruit to the domestic and

export markets. The results were used to simulate the grapefruit industry








71

to ascertain its performance to changes in the major factors. The system of equations was estimated using a seemingly unrelated regression (SUR) model.

Nelson and Robinson (1978) developed a model to analyze retail and wholesale fresh navel orange demand under marketing order policy. Two important issues were raised in this study: apples and bananas could be used as substitute products for fresh oranges; and the demands for fresh

and processed oranges are independent. The first issue was raised

previously by Matthews, Womack, and Huang (1974) with encouraging results. Prato (1969) found that demands for fresh oranges and concentrate are independent, at least in the winter season.

Ward (1981) applied time-varying parameters (TVP) to analyze the welfare impact and economic forecast based on a better understanding of the economics of the EC fresh orange industry. This study was especially

important at that time given the plans of enlargement of the EC to include Greece, Spain and Portugal. To support the use of TVP, Ward argued that, given the evolution of the EC and its related regulations, it is possible to hypothesize that some adjustments in the demand parameters are likely

to have occurred over the decades since the early 1960s. He also

estimated the model using Ordinary Least Squares (OLS) and the results were compared. It was clear that the use of TVP performed better that the simple OLS estimation technique.

McCabe (1982) estimated a model to determine the characteristics of

fresh citrus consumers. The major objective was to ascertain how

demographic and household characteristics affect purchase decisions and to determine its relationship with product prices.








72


Wardowski et al. (1986) recently edited a book that includes a descriptive analysis of world production practices and trends and a longterm view of fresh citrus trade. An interesting discussion on trade flow and market share of imports and exports was presented. Global trade for the 1980s was projected, based on the assumption that trade will growth

one third less rapidly than in the previous decade, given special assumptions for each citrus product. Individual country/region

projections were based on historical trends in per capita availability, where such trends were evident or trends on total imports were estimated.

The use of trends in projecting import demand was based on the assumption that future levels of economic factors that determine demand will follow

historic trends. No demand estimation was pursued to determine trade flow and market share in this study.

Lee and Fairchild (1988) used a SUR technique to study the relationship between exchange rates and foreign demand for United States fresh grapefruit. The results showed that exchange rates played a major

role when studying export demand relationships and the United States fresh grapefruit has more than one export market, with markets responding differently to price changes. These results will be used later to define the model to be estimated.

Lee et al. (1990), using the absolute version of the Rotherdam model, studied the Japanese citrus products market. The study used fresh bananas and pineapples as substitutes for fresh citrus products. One of

the major conclusions of the study was that United States fresh grapefruit exports compete against imports of bananas and pineapples for Japanese import dollars. In the case of fresh oranges, the results were not








73

consistent with the expected signs; especially in the case of pineapples, which turned out to be a complement for fresh oranges, an unexplainable result as reported by Lee et al. This article and the one by Nelson and

Robinson (1978) have interesting insights that will be considered later in order to define the best substitute products for fresh oranges.

Even though the present study will not deal directly with the processed orange industry, a few comments on the literature reviewed will be made. Priscott (1969) developed a model to estimate the demand for citrus products (juices) in the European market. One of his major

findings was that there is substantial substitution among products of specified countries, reinforcing once again the need to differentiate products by place of origin. Weisenborn et al. (1970) used the theory of price discrimination to determine the optimal market allocation of Florida orange production for maximum net returns. To solve the problem of price discrimination, quadratic programming and calculus with LaGrangean

multiplier techniques were used. Malick (1980) used a simultaneous equation model of the Florida retail orange-juice marketing system to forecast changes in the FOB price and retail movement of frozen concentrated orange juice (FCOJ). Irias (1981) developed an econometric

model to study international trade of FCOJ among three regions, the United States, Brazil, and Europe. Margoluis (1982) developed a model to estimate implicit prices for juice and drink characteristics using hedonic price functions. Ting (1982) developed a model to test the existence of asymmetric price response in the irreversible demand functions for citrus juice products.








74

Most of the work has been concentrated on the United States domestic market analysis and in specific econometric models designed to explain one or more elements of the international trade flow matrix and markets. The studies are usually related to the United States product behavior in Canada, Europe, and Japan. In most cases, the estimation has been pursued using single-equation estimation and, in a few cases, using SUR. The fresh orange industry has not been studied in a full simultaneous spatial

equilibrium world trade model modified to take into account that products are differentiated by country of origin and therefore are not perfect substitutes. The results presented in many of the articles and books reviewed regarding trade of different commodities, and specifically fresh and processed oranges, strongly support the conclusion that fresh citrus

coming from different countries (or regions) are perceived as different products by consumers. The main objective of the present study will be to develop and estimate a modified spatial equilibrium world trade model for the fresh orange industry. The model will be used to analyze the impact of different trade policies and economic factors affecting the demand for fresh oranges in different regions of the world.















CHAPTER 4
WORLD FRESH ORANGE TRADE MODEL



Introduction



With international trade models it is frequently assumed that goods

of a given kind supplied by different (national) sellers to a single country are perfect substitutes in the final market. With this

assumption, consumers differentiate goods only by kind, and there is no evident difference between products of the same kind supplied from different sellers. It also implies that the elasticities of substitution between suppliers are infinite, and that the corresponding price ratios are constant (Armington, 1969a).

In general, fruits, and in particular fresh oranges, are expected to be differentiated by place of origin. There are several varieties of oranges, and regions have soil and climatic conditions favoring the production of only a few varieties. Production seasons are highly

variable among regions and yield products at different times of the year.

For example, while the Northern Hemisphere countries harvest their oranges from November to June, the Southern Hemisphere countries harvest their fruit from June through October. In addition, product coming from

different regions even at the same time period could be perceived to have distinctive quality features by the final consumer.


75









76

Under these circumstances, the theoretical model defined in this section will be based on Armington's model of international trade (Armington, 1969a). As previously mentioned, it is a modified spatial equilibrium model that takes into account the concept that commodities are differentiated not only by kind but by exporting region. Armington

distinguished commodities from products. For example, the term commodity

refers to a specific good such as fresh oranges, cotton or rice, or an aggregated good such as fruits, meats or vegetables. On the other hand, a product is a commodity exported from one region to another; i.e., fresh oranges coming to France from the U.S. is a different product than fresh oranges coming from Spain to the same country of destination.

The first basic assumption underlying this model is that consumers'

utility is weakly separable; therefore, the decision process may be viewed as occurring in two stages. The first decision stage is to determined the total level of consumption for each commodity known as "market demands". This decision is usually based upon commodity prices, income levels, substitute commodity prices, and other relevant economic variables. The second step is to decide where to buy the product; i.e., given that the

total consumption level for each commodity has been determined, an allocation among the different suppliers has to be made. These are known as "product demands". The distribution among suppliers is based on the commodity's total market demand and relative product prices.

The second basic assumption in Armington's model is that the quantity index function used to represent quantities imported from the regions of origin is linear and homogeneous. This assumption implies that








77

each region's market share of a commodity is influenced by changes in the size of that market, even when relative product prices remain unchanged.

In the present study, 11 regions were defined. The regions were selected consistently with the world orange industry and with particular similarities among the countries included in a region. The regions were

the United States (US), Canada (CAN), Latin America (LA), Mediterranean-EC (MED-EC), EC, rest of Western Europe (RWE), Middle East/North Africa (ME/NA), rest of Africa (RAF), Far East (FE), Oceania (OCE), and Communist Bloc (COMMB).

In the next section, a complete world fresh orange trade model is specified. Demand and supply sides are included with equilibrium conditions and price linkages set forth.



Fresh Orange Trade Model



Demand Side



The model was based on the two assumptions mentioned above. Two stage budgeting is implied. Marginal rates of substitution between two

goods in a commodity group were assumed to be independent of goods in other groups. In the orange industry, the rate at which consumers substitute fresh oranges produced in one country for those produced in another country does not depend on their purchases of other kinds of fruits or other commodities. The first level of the two stage budgeting

is the consumers' decision to allocate their total income among the








78

different commodity groups available in the region. A percentage of that income is allocated to the total market demand for fresh oranges.

In the general case, the utility function for consumers in region i given n commodities is:

(4.1) Ui = U(Xi)

where
Xi (Xli, Xli2, . ,X ,Xi X im) is the total bundle
of commodities for region i, and m the total number of regions
or countries considered.

The first subscript for XkiJ represents the commodity (n

commodities) the second represents the region of destination (m regions), and the last subscript denotes the region of origin (m regions). Given the assumption of weakly separable utility function or independence among commodities in different groups and following Solow (1955-56) and

Armington (1969a), it is possible to write this utility function for region i as follows:

(4.2) U1 U'(Xli.,X2i, ..., Xi.)

where
Xki. pk (Xkil, Xki2, .. Xkim) for k-1, 2, . ,n.

Xki. is the total market demand for commodity k in region i. The

dot represents the sum over all j's or regions of origin including the domestic region i. The pk represents certain quantity index function of the product demands Xkij which represent the demand for commodity k in region i coming from

region j where j-l,. . ,m.

Consumers maximize the utility function (4.2) subject to the budget constraint given by









79

(4.3) INCi 4 X (PRi Xkij) i. Xki.) k-1,2,.. ,n and



where

INCi is total expenditure (or income) for all commodities in region
i,

Pkij is the price for commodity k coming from region j in region i
(or products price),

Pki. is the average price of commodity k in region i for
k-1,2,...,n,

Z is the sum over all k (commodities) for all j (regions).

The resultant "market demand" equation for commodity k in region i is a function of total income or expenditure, commodity k price, other commodities' prices, and other relevant variables: (4.4) Xki. Xki. (INC Pli. P2i., 1ki. P P ni. P ZO where
Zi represents other variables of interest.

An interesting result is that total market demand (Xki.) is a function of only the average import price for this commodity group and the average price of substitutes and not of the individual product prices (Pkij)

Total market demand is then separated by exporting region in the second level of the two stage budgeting process to obtain the "product demand" equations. In this case, consumers minimize the cost of purchasing Xki. (total market demand for commodity k in region i). That is, consumers minimize total expenditure (INC), subject to the following constraint:

(4.5) Xki. pk (Xki, Xki2, Xkim)








80

to obtain the specific product demands. The function ilk is assumed to be a linear and homogenous function of the product demands Xki to ensure that Pki. is independent of Xki.. Pki. is only a function of Pkil, Pki2) -... Pkim The assumption that the quantity index functions are linear and homogeneous is the second restriction (the first being the assumption of independence) that has been placed on U. The product demands generated

under these assumptions are functions of the total market demand level (Xki.) and the individual product prices (Pkii) and is given by (4.6) Xki3 Xkij(Xki. ,Pki1,Pki2, -, Pkim)

This relationship clearly states that the allocation of imports among regions of origin depends on total market demand and relative prices of the products in the market.

The total market demand equation for the world's fresh orange trade

model in any particular region is defined following the theoretical framework developed above. It is possible to write the market demand equation for fresh oranges as independent of other goods consumed in the

same region. The model will be dealing with only one good (fresh oranges) hence the subscript "k" is no longer necessary. Let Xj represent fresh oranges exported from region j to region i.

The market demand equation for fresh oranges is expected to be a function of the average market price, income, population, and the price of substitute products. The average market price should be obtained by taking into consideration the local product price and the price of imports including any tariff or preferential treatment. In the case of substitute products, it has been necessary to define what is really a substitute for fresh oranges. Several alternatives were considered, including an








81

aggregated commodity representing all other fruits, an aggregated

commodity representing all other goods, and an aggregated commodity representing bananas and apples. The latter alternative was selected based on the characteristics of consumption for fresh oranges which makes

bananas and apples better substitutes than the other aggregated goods. Nelson and Robinson (1978) reported that Matthews, Womack, and Huang (1974) used bananas and apples as substitute products for fresh oranges with encouraging results, even though they were in some cases less significant than the ones obtained using other alternatives. In a recent paper, Lee et al. (1990) also used bananas as a substitute product for U.S. citrus in Japan. However, in the latter study pineapples instead of apples were used as the second substitute product.

The general form of the market demand equation for fresh oranges is the following:1

(4.7) X1 f(P2,INCt,POPt,PRSt)

where
f represents some functional relationship between Xi. and the
variables on the right hand side,

Xi. is the total market demand for fresh oranges in region i,

Pi. is the real average market price of fresh oranges in region i,

INCi is the real income level in region i,

POP1 is the population level in region i,

PRS1 is the real average market price for the aggregated commodity
based on bananas and apples or other measure of substitutes in
region i.




'Single letter notation represents endogenous variables while three letters depict exogenous variables. The sign associated with each variable represents the hypothesized behavioral relationship between the exogenous variables and the dependent variable.








82

The second level of two stage budgeting is to allocate total market

demand by supplying region. It requires the definition of a "product demand" equation which represents the demand in region i for fresh oranges coming from region j. The product demand functions consider Armington's demand theory of products differentiated by place of origin. Prices for

products in commodity markets other than fresh oranges have no effect except through the size of the markets. The function pk in equation 4.5 is assumed to be a linear and homogenous quantity index function of the

product demands Xkij to ensure that Pki. is independent of Xki.. The average market price Pki. is a function of Pkit, Pki2, - Pkim. Equation 4.6 defined the product demands to be a function of the market size and all product

prices. Since Pki. is a function of all product prices product demands and market shares can be reduced to depend on relative prices and total market demand or market size. The relative price for each product demand is given by the ratio of the product price to the average fresh orange price in the market. The product demand functions are (4.8) Xj h'(P ,Pi,X:/-)

where
hr represents some functional relationship among variables,

Xi is the demand in region i for fresh oranges coming from region
j,

Pj is the price in region i for fresh oranges coming from region
j.

The actual relationship is given by

(4.9) Xi = h(Pjj/P ,X/~)

where
h represents some functional relationship among variables.








83

Given equations 4.8 and 4.9, it is possible to define each region's market share equation as follows: (4.10) Si -Xij/Xi. = z(Pij/P-.,Xi/-) where
z represents some functional relationship among variables,

Sij is the market share of fresh oranges for region j in region i.



Supply Side



Total orange production (PRD.j) is defined to include oranges supplied to the fresh market (PRDj) and oranges utilized for processing into orange juices (PRD2j). The orange industry requires several years to introduce new trees and new supplies in the market and high levels of investment to built a new processing plant. It is reasonable to assume

that orange production and its utilization levels do not adjust so fast as to be considered part of a simultaneous demand side decision model of international trade. Therefore, total production and, in particular, fresh orange utilization (PRDj) is considered exogenous in this model. The general equations representing this condition are the following:

(4.11) PRDuj PRD.j PRD2J

and

(4.12) PRD2j = ..j*PRD.j

where

PRD. is total orange production in region j,

PRDj is total fresh orange utilization in region j,

PRD2j is total processed orange utilization in region j








84

Ai is the percentage of total orange production utilized in the
processed industry in region j and is assumed to be exogenous.



Export Supply Equations



Exporters will respond to export prices by adjusting their level of exports accordingly. Given changes in total production and fresh utilization, exports will also tend to adjust accordingly.

Export supply equations are consequently assumed to be a function of the average export price from region j (average Free On Board price F ) and total fresh orange utilization (PRD1j) in the region of origin. The export supply equation for fresh oranges is the following: (4.13) X.3 Z Xij = v(F*.,PRD+) for ioj where
The summations represent total exports of fresh oranges from region
j to all other regions,

v represents some functional relationships between variables,

F. represents the average export price of fresh oranges from
region j to all other regions.

The demand equations for local product will follow from the difference between total fresh utilization (PRD3) plus the change in inventories (when applicable) and the export supply from region j. Demand for domestically produced product is:

(4.14) Xjj = PRDiJ + A INV3 X.

Given that fresh oranges can be stored only for short periods of time, it is assumed that inventory levels are zero. Accordingly, the change in inventories will be zero and equation 4.15 will be given by

(4.15) Xii PRDi X




Full Text
92
Ordinary least squares estimated parameters are biased for this
simultaneous equation model. Another estimation problem in the model is
that, even though it is linear in the parameters, it is intrinsically
nonlinear in the variables, given that after transforming equation (4.30)
total market demand will be in the log linear form while it appears
without the log linear form in equation (4.36). As a consequence,
nonlinear two stage least square procedure was used. The specific
estimation steps used and results will be given in the next chapter.
Model Implications
Modeling the changes of world trade flows of the orange industry by
identifying international trade linkages among the major trading regions
and recognizing current and emerging problems in the industry is the major
objective of the present study. Estimation of the world trade model
described above will generate consistent estimates of the parameters of
the market demands, product demands, export supply, and CIF import price
equations for the major trading regions in the industry. Analysis of the
estimated parameters will provide information to help understand the
reasons for changes in market shares and facilitate longer term forecasts
and policy analyses.
The parameters of the market demands measure the strength of the
influence of the average price of fresh oranges in a given region, as well
as the intensity of income and population levels, and substitute commodity
prices. Using price elasticities, it is possible to predict responses in
the different markets to changes in supply prices. Income and population


CHAPTER 2
FRESH ORANGE WORLD PRODUCTION AND TRADE
Introduction
This chapter discusses world production and trade flows of fresh
oranges. The discussion will be based on several tables for 11 specified
regions of the world. These regions were selected based on similarities
of supply and demand conditions among the different countries included in
each region with regard to the orange industry. The regions are the
United States (US), Canada (CAN), Latin America (LA), Mediterranean-
European Community countries (MED-EC), the rest of the European Community
countries (EC), rest of Western Europe (RWE), Middle East/North Africa
(ME/NA), rest of Africa (RAF), Far East (FE), Oceania (OCE), and Communist
Bloc (COMMB). The Communist Bloc is defined as it existed before the
political changes of 1991. Appendix A shows the countries included in
each region.
Production Analysis
Table 2.1 shows the production levels of oranges in the 11 regions
identified for 1966, 1976, and 1986. These years were selected to
illustrate changes through time. World orange production increased at an
annual rate of 3.3% in the last 20 years and increased at a rate of 1.9%
29


Table 5.16 Latin America CIF Price Linkage Equations
Energy 21 Years Average
Partner
Region
Intercept
( + \-)
FOB
Price
( + )
Year
Trend
(-\ + )
Index
Price
( + )
@OBS
@RSQ
@DW
@FST
UTHEIL
Z of
Total
Imports
Z Total
Market
Demand
us
PARAM.VALUE
STD.ERROR
t STATISTIC
29.787
20.140
1.479
2.960
1.321
2.241
-6.093
4.196
-1.452
-0.246
0.267
-0.920
21
0.87
2.22
37.70
0.145829
88.135
0.031
CAN
PARAM.VALUE
STD.ERROR
t STATISTIC
9.739
18.950
0.514
0.689
0.275
2.510
-2.441
4.389
-0.556
0.407
0.209
1.950
21
0.80
2.11
21.99
0.129447
0.344
0.000
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-0.883
11.891
-0.074
0.726
0.338
2.151
0.128
2.759
0.047
0.170
0.183
0.928
21
0.76
2.03
18.38
0.181972
0.796
0.000
EC
PARAM.VALUE
STD. ERROR
t STATISTIC
-3.003
6.741
-0.445
0.975
0.322
3.025
0.775
1.555
0.498
0.004
0.115
0.034
21
0.86
2.58
33.91
0.094185
5.706
0.002
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
1
0.041
0.000
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
-8.814
28.433
-0.310
0.649
0.535
1.214
2.005
6.582
0.305
0.032
0.326
0.098
21
0.39
2.30
3.65
0.248192
2.175
0.001
RAF
PARAM.VALUE
STD.ERROR
t STATISTIC
4
0.090
0.000
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
6
0.262
0.000
OCE
PARAM.VALUE
STD.ERROR
t STATISTIC
2
0.086
0.000
COMffl
PARAM. VALUE
STD.ERROR
t STATISTIC
-7.965
16.448
-0.484
0.408
0.485
0.841
1.640
3.701
0.443
0.186
0.139
1.338
21
0.78
2.20
19.89
0.179599
2.365
0.001
100.000 0.035
Total


METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.25. Canada Imports of Fresh Oranges from the United States (Product Demand 21).
165


298
Chapter 4 develops the fresh orange trade model used in the present study.
The theoretical background and the empirical model to be estimated are
presented. Chapter 5 discusses the methods used for the estimation of the
model. It also develops graphical, statistical, and economic analyses to
study the performance of the model and the implications of the results.
Chapter 6 presents a sensitivity analysis to study the changes in the
total market demands, export supplies, and product demands, given changes
in the right-hand side variables.
Data Limitations
The data required for this model has serious limitations. It is
necessary to have all trade flows, import and export value, and quantity
for every country of the world, showing the partner country. The data are
then aggregated by region. If all countries of the world are included,
the data are not available except from the United Nations trade data
tapes. These data are gathered by each member country and sent to the
statistics office in New York. The price data used in this dissertation
are unit prices obtained by dividing value by quantity for each trade
flow. As expected in trade data, many errors were found. Most of them
were probably related with gathering problems and inconsistencies. Where
errors were detected, the data were corrected in what was believed to be
the most appropriate way.
Tariff barriers for fresh oranges were not available in a single
document for all countries. It was necessary to review many different
sources to obtain the final data presented in Appendix E. Tariffs of the


158
and the years after. Those changes culminated in 1986 with the admission
of Spain and Portugal to the EC.
The rest of Africa consumption pattern has been very irregular over
time and the model has been unsuccessful in reflecting the major turning
points. This region is formed by one large producer and exporter (South
Africa) and many countries that usually consume only what they produce.
Imports in this region are very small as compared to total market demand.
Turning points in this region's total market demand are probably related
to exogenous changes in local production of oranges and therefore are not
predicted by the model.
Export supply
Figures 5.12 to 5.22 show that the model generates the trend of the
export supply equations for every region in the model. The model has the
ability to predict if a region has a growing export supply or if the
export supply is decreasing over time. The results also show that the
model does not capture the turning points as well as it did for total
market demand equations with some exceptions. The model does reflect
most of the turning points for the United States, Mediterranean-EC, and
Middle East/North Africa. These regions represented 88% of total world
exports between 1966 and 1986.
Product demand
The third group of figures show the actual and fitted values for
selected product demand equations. Total trade for selected regions
represented over 90% of total world trade in the 21-year period


265
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.12. Latin America Imports Changing Total Market Demand.


46
2.5%, 3.8%, .9% and 1.9%, respectively. By 1976, Canada represented
41.1%, the EC region stayed almost the same, and the Far East region
absorbed 31.5% of the total United States fresh exports. In 1986, United
States exports to the Far East reached 64.9% of its total volume,
representing an important shift of the United States export partners. The
second largest partner was Canada, with 30% of the total volume. The EC
was no longer significant for the United States exports, given that it
represented only 2.3% in the same year. The rest of the regions also
decreased their participation relative to previous years.
Table 2.3 shows that exports from Latin America have doubled in
absolute terms in the last two decades. However, exports did not increase
from 1966 to 1976, which implies that the increase took place during the
last ten years. Total world trade participation of Latin America passed
from 3.1% in 1966 to 2.4% in 1976 and 4.3% in 1986. With intraregional
trade excluded, its participation in world fresh trade increased to 4.6%
in 1986. Note that Brazil generally does not export fresh oranges.
The major export market for Latin American product has been the EC
region, which in 1966 absorbed 67.2% of the total product exported (see
Table 2.4). This percentage decreased to 38.1 in 1976 and increased to
77.6 in 1986. The United States was the second largest market for Latin
America exports in 1966, with 22.5% of the total export level. This
percentage increased slightly in 1976 and decreased to 9.8 in 1986. The
third largest market for Latin America was the Communist Bloc, which took
most of the reduction shown in the EC region during the 1970s and part of
the United States share in the 1980s. Latin American exports to Canada
and the Middle East/North Africa have been increasing, especially in the


APPENDIX C
PROCEDURE TO OBTAIN REGIONAL CPIs
The procedure developed by Edwards and Ng (1985) to obtain the
regional CPIs (Consumer Price Indices) is the following:
1.- Get the percentage change of the CPIs per country (annual inflation)
2.- Get the exchange rates with respect to the U.S. dollar per country
3.- Get an index of the exchange rate for a base year
4.- Divide the CPIs by the exchange rate index to obtain the CPIs by
country
5.- The individual country's CPIs are weighed using trade levels to obtain
the regional CPIs aggregate values
309


REFERENCES
Allen, R. G. D. "Mathematical Analysis for Economist," St. Martin's
Press, London, 1938.
Amemiya, T., "On the Use of Principal Components of Independent Variables
in Two-Stage Least-Squares Estimation," International Economic
Review 7(1966):283-303.
Amemiya, T., "The Nonlinear Two-Stage Least-Squares Estimator," Journal of
Econometrics 2(1974):105-110.
Armington, P. S., "A Theory of Demand for Products Distinguished by Place
of Production," IMF Staff Papers 16(1969a):159-178.
Armington, P. S., "The Geographic Pattern of Trade and the Effects of
Price Changes," IMF Staff Papers 16(1969b):179-201.
Armington, P. S., "A Many Country Model of Equilibrating Adjustments in
Prices and Spending," IMF Staff Papers 17(1970a):23-26.
Armington, P. S., "Adjustment of Trade Balances: Some Experiments with a
Model of Trade Among Many Countries," IMF Staff Papers
17(1970b):488-517.
Armington, P. S., "A Note of Income-Compensated Price Elasticity of Demand
Used in the Multilateral Exchange Rate Model," IMF Staff Papers
20(1973):488-511.
Arrow, K .J., H.B. Chenery, B.S. Minhas, and R.M. Solow, "Capital-Labor
Substitution and Economic Efficiency," The Review of Economic and
Statistics 43(1961):225-250.
Artus, J. R., and R. R. Rhomberg, "A Multilateral Exchange Rate Model,"
IMF Staff Papers 20(1973):591-611.
Aslton, J. M., C. A. Carter, R. Green, and D. Pick, "Whither Armington
Trade Models?" American Journal of Agricultural 72(1990):455-467.
Baker G. and H. Mori, "Strawmen in Trade Protectionism: The Case of
Citrus Import Quotas," Western Journal of Agricultural Economics
10(1985):338-343.
339


110
Table 5.3 United States Product Demands
Partner
Region
Intercept
( + \-)
Relative
Price
(-)
Total
Market
Demand
<-\+>
OBS
RSQ
DW
FST
UTHEIL
21 Year Average
Z Z Total
Total Market
Imports Demand
CAN
PARAM.VALUE
STD. ERROR
t STATISTIC
75.967
84.489
0.899
1.436
1.118
1.285
-5.097
5.869
-0.868
13
0.17
1.42
1.04
0.39015
0.035
0.001
LA
PARAM.VALUE
STD.ERROR
t STATISTIC
-36.637
40.069
-0.914
1.051
0.370
2.840
3.298
2.785
1.184
21
0.31
1.68
4.06
0.197823
83.619
2.194
MED-EC
PARAM.VALUE
STD.ERROR
t STATISTIC
-388.180
153.100
-2.535
-0.857
1.473
-0.582
27.351
10.626
2.574
21
0.28
1.03
3.48
0.70272
2.790
0.073
EC
PARAM.VALUE
STD.ERROR
t STATISTIC
-7.761
68.958
-0.113
-1.214
0.364
-3.334
0.815
4.778
0.171
19
0.44
2.17
6.20
0.335905
0.036
0.001
RWE
PARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
ME/NA
PARAM.VALUE
STD.ERROR
t STATISTIC
-159.400
104.410
-1.527
-5.564
1.607
-3.464
11.692
7.248
1.613
21
0.43
1.90
6.77
0.265217
11.705
0.307
RAF
PARAM.VALUE
STD.ERROR
t STATISTIC
3
0.040
0.001
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
-8.263
14.097
-0.586
7.510
1.763
4.261
0.578
1.002
0.577
21
0.54
1.29
10.70
0.179367
1.671
0.044
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
-427.920
122.020
-3.507
-1.512
0.726
-2.082
30.060
8.461
3.553
10
0.75
0.78
10.45
0.288411
0.104
0.003
COM
PARAM.VALUE
0
0.000
0.000
SID.ERROR
t STATISTIC
100.000 2.624
Total


APPENDIX F
PRINCIPAL COMPONENT PROCEDURE AND PROGRAM
FREQ A;
SMPL 66,86;
READ (FORMAT-LOTUS FILE-' C: \LOTUS\DATA. WK1') ;
? START OF PROGRAM;
LIST EXOG
POPI GDP1
PRD1 CPU YEAR
PEN BAVAL1
POP2 GDP2
PRD2 CPI2
BAVAL2
POP3 GDP3
PRD3 CPI3
BAVAL3
P0P4 GDP4
PRD4 CPI4
BAVAL4
P0P5 GDP5
PRD5 CPI5
BAVAL5
P0P6 GDP6
PRD6 CPI6
BAVAL6
P0P7 GDP7
PRD7 CPI7
BAVAL7
P0P8 GDP8
PRD8 CPI8
BAVAL8
P0P9 GDP9
PRD9 CPI9
BAVAL9
POPIO GDP10
PRD10 CPI10
BAVAL10
P0P11 GDP11
PRD11 CPI11
BAVAL11;
PRIN (NAME-PC.NCOM-6,FRAC-,98.NOPRINT) EXOG;
PRINT PCI PC2 PC3 PC4 PC5 PC6;
END;
The exogenous variables included in the principal component procedure are
Population (POP), Gross Domestic Product (GDP), Production (PRD), Consumer
Price Index (CPI), and Bananas and Apples Price Index per region (BAVAL).
The Year Trend (YEAR) and the Price Index for Energy (PEN) were also
included. PCI to PC6 refer to the principal components obtained with the
procedure.
312


METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.23. United States Imports of Fresh Oranges from Latin America (Product Demand 13).
163


80
to obtain the specific product demands. The function /k is assumed to be
a linear and homogenous function of the product demands xkij to ensure that
Pki is independent of Xki Pki is only a function of Pkil, Pki2> > Pkim-
The assumption that the quantity index functions are linear and
homogeneous is the second restriction (the first being the assumption of
independence) that has been placed on U. The product demands generated
under these assumptions are functions of the total market demand level
(Xki ) and the individual product prices (PkiJ) and is given by
(4.6) Xkij Xkij(Xki ,Pkil,Pki2 Plcira)
This relationship clearly states that the allocation of imports
among regions of origin depends on total market demand and relative prices
of the products in the market.
The total market demand equation for the world's fresh orange trade
model in any particular region is defined following the theoretical
framework developed above. It is possible to write the market demand
equation for fresh oranges as independent of other goods consumed in the
same region. The model will be dealing with only one good (fresh oranges)
hence the subscript "k" is no longer necessary. Let X^ represent fresh
oranges exported from region j to region i.
The market demand equation for fresh oranges is expected to be a
function of the average market price, income, population, and the price of
substitute products. The average market price should be obtained by
taking into consideration the local product price and the price of imports
including any tariff or preferential treatment. In the case of substitute
products, it has been necessary to define what is really a substitute for
fresh oranges. Several alternatives were considered, including an


CHAPTER 3
LITERATURE REVIEW
International Agricultural Trade Models
Several models or approaches to study international trade have been
developed in the last two decades. These models were developed mainly due
to the need for knowledge and understanding of increasing world trade.
Thompson (1981) presented an interesting survey of new developments in
international agricultural trade models. In his document, each model was
reviewed in three sections: a historical survey, an evaluation, and a
summary and implications section. The different modeling approaches were
divided into two basic groups determined by the number of regions
considered in the model. The two groups were two-region models and
multiple-region models of agricultural trade. The latter was further
divided into three groups: non-spatial price equilibrium, spatial price
equilibrium, and trade-flow and market-share models.
A different classification system for international trade models was
developed by Thompson and Abbott (1982) Each modeling approach was
grouped based on the assumptions made about the homogeneity of the
commodity traded. The two major categories identified in their research
were single homogeneous commodity models and multiple-product models. The
single homogeneous commodity models were divided into three groups: non-
spatial price equilibrium, spatial price equilibrium, and two-region
58


254
demand will depend on consumers' preferences with respect to product
sources, as their total demand changes. For example, if the total market
demand increases in a certain region, the consumers' next step will be to
decide from which region to buy the extra product. The consumers'
decision could be in favor or against any potential source. The analysis
will consider market size increases and decreases. Market size decreases
below the 1986 level are less likely to occur, given the behavior of fresh
orange consumption in the last two decades.
The product demands were selected taking into consideration trade -
flow volumes among the regions. For every region, a group of partners
that accounted for over 90% of total imports were considered.
United States
Table 6.2 shows that Latin America, Middle East/North Africa, and
Mediterranean-EC accounted for 98.1% of total United States imports in the
period considered. Figures 6.7 and 6.8 present United States imports
while changing import prices and total market demand. The sensitivity
analysis presented in Figure 6.7 indicates that United States product
demand behavior differs dramatically, depending on the product source.
Latin America is the major exporter to the United States (based on
reported data). The relationship between the import price index and
product demand is positive in this event. This is an unexpected result,
probably related to the fact that the United States is self-sufficient and
imports exist only when local production is insufficient to fulfill
consumers' demand. If this is the case, then a positive relationship is
possible. As shown in the figure, the demand for Middle East/North Africa


106
equation. It is also known that the expectation of a function is
generally unequal to the function of the expectation. Therefore, the
second alternative is inappropriate (Goldfeld and Quandt, 1968; Goldfeld
and Quandt, 1972). The first method was used to obtain the necessary
instruments for the second stage.
The methodology utilized to estimate the model closely follows the
nonlinear two stage least squares method proposed by Kelejian (1971) and
supported later by Goldfeld and Quandt (1972) and Amemiya (1974).
However, the final procedure used introduced different specifications for
the equations in the first stage.
Time Series Processor (TSP) International PC Version (TSP User's
Guide and Reference Manual, 1983) procedures were used to estimate the
model. The final program used for the estimation of the model is included
in Appendix G.
The data used cover from 1966 to 1986 and have been recorded from
many sources as described in Chapter 4. Trade data were insufficient in
terms of degrees of freedom to perform an adequate estimation for some
equations. In those cases, a special TSP procedure was utilized to select
and estimate only those equations for which trade took place. Equations
with less than six trade observations did not provide enough information
for a reliable estimation. In that event, the equation was not estimated
and its parameters were set to zero. Data were read in from Lotus files.


2.3 Trade Flow Analysis for Selected Years (1966,
1976 and 1986) by Region in Relation to Partner
Regions 35
2.4 Trade Flow Analysis for Selected Years (1966,
1976 and 1986) Without Intraregional Trade
"Relative Partner Region Exports by Region" 37
2.5 Trade Flow Analysis for Selected Years (1966,
1976 and 1986) Without Intraregional Trade
"Relative Region Imports from Partner Regions 38
2.6 Trade Flow Analysis for Selected Periods of Five
Years (1966-70,1974-78 and 1982-86) 40
2.7 Trade Flow Analysis for Selected Periods of Five
Years (1966-70,1974-78 and 1982-86) Without
Intraregional Trade "Relative Partner Region
Exports by Region" 42
2.8 Trade Flow Analysis for Selected Periods of Five
Years (1966-70,1974-78 and 1982-86) Without
Intraregional Trade "Relative Region Imports
from Partner Regions" 43
5.1 Total Market Demand Equations 108
5.2 Export Supply Equations 109
5.3 United States Product Demands 110
5.4 Canada Product Demands Ill
5.5 Latin America Product Demands 112
5.6 Mediterranean-EC Product Demands 113
5.7 EC Product Demands 114
5.8 Rest of Western Europe Product Demands 115
5.9 Middle East/North Africa Product Demands 116
5.10 Rest of Africa Product Demands 117
5.11 Far East Product Demands 118
5.12 Oceania Product Demands 119
5.13 Communist Bloc Product Demands 120
5.14 United States CIF Price Linkage Equations 121
vii


METRIC TONS (Thousands)
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
YEAR
ACTUAL 1 FITTED
Figure 5.39. Middle East/North Africa Imports of Fresh Oranges from the rest of Africa (Product Demand 78).
179


56
If intraregional trade were excluded, these percentages changed to .1 for
the last two years reported.
The four major suppliers of fresh oranges to the rest of Africa were
the Middle East/North Africa with 83.0% in 1966, 71.2% in 1976, and 70.5%
in 1986; the EC with 11.8%, 9.3%, and 12.5%, respectively; Oceania with
3.4%, .7%, and 8.9%, respectively; and the Mediterranean-EC with .1%,
14.1% and 8.0%, respectively (see Table 2.5). The United States share of
the rest of Africa market was .6% in 1966. However, the United States
lost its share totally by 1986.
Latin America's portion of total world imports was .3% in 1966, .4%
in 1976, and .1% in 1986 (see Table 2.3). Given that most of its trade
was among countries of the region, these percentages decreased to .2 in
1966 and to 0 in 1976 and 1986. Imports in Latin America came from the
United States in the 1960s and 1970s (see Table 2.5). In 1986 the United
States share was only 68.0% of total imports. The rest of the product
came mainly from the Communist Bloc with 33.4%, the EC region with 16.1%,
and the Mediterranean-EC with 4.5%.
The Mediterranean-EC region has only a small share of total world's
fresh orange imports. Imports reached .2% in 1986 with and without
considering intraregional trade (see Table 2.3).
Conclusions
In summary, it is possible to describe most of the world production
and trade flows of the fresh orange industry with few regions. On the
production side, the major producers of oranges were Latin America, Far


234
be known and were changed from approximately 30% below to 30% above 1986
levels. It is also possible to determine what the necessary change in the
FOB export price, tariff, tax, or any other factor would be when the final
market import price changes over the proposed range. To perform this
analysis, the respective CIF price linkage equation and some of the model
identities had to be used. These questions are important, but they can
always be addressed at a later step considering the results and the
specific regions of interest. Therefore, the equations selected for the
sensitivity analysis were total market demands, export supplies, and
product demands.
Each equation selected is a function of a different set of
variables. A decision had to be made regarding the set of variables to be
modified for each equation. Total market demands are functions of average
market price, income (GDP), population, and substitute product price.
Changes in the average market price are related to changes in tariffs,
taxes, local prices, and FOB export and CIF import prices. Economic
theory and the empirical results of Chapter 5 indicate that average market
price and income (GDP) are the major driving factors for consumption in
most regions. Therefore, these variables were included in the sensitivity
analysis for total market demand equations.
Export supply equations are functions of the FOB average export
price and domestic fresh production. The empirical results of Chapter 5
indicate that fresh production is probably the major driving factor for
exports. However, economic theory suggests that FOB average export price
should also be a major factor. Given these conditions, both variables
were included in the analysis.


194
The UTHEIL indicates that all export supply equations have a
coefficient far below .5 except for Canada. Overall, the model is
predicting the major turning points of the historical data. Canada is a
net importer and has no production of oranges; therefore, the results will
not have an important impact on the fresh orange trade model.
Product demands and CIF price linkage equations
Tables H.l to H.4 in Appendix H present the statistics for the
product demand and CIF price linkage equations. Detailed region-by-region
discussions about those statistics will not be included here.
Conclusion: statistical analysis
The statistical analysis shows that the model is capturing the major
variations of the different dependent variables for each total market
demand equation. Problems found were usually related to regions that will
not affect the major driving issues of the fresh orange trade model.
The analysis also shows that the model is reflecting the major
variations of the different dependent variables for total market demands
better than it does for export supply equations. However, the results
show that export supply equations for major world exporters are well
captured by the model.
The reported statistics show that the model is predicting the major
variations of the product demands better for which relevant trade took
place. Important trade flows like Canadian imports from the United
States, EC imports from the Mediterranean-EC and the Middle East/North


246
The figures also show that the United States has the largest
response. Its parameter is the most elastic in the group, indicating that
consumers in this market are highly sensitive to changes in the average
market price. If world prices increase, consumers in the United States
will consume proportionally less fresh oranges than other regions in the
world. If world prices decrease, United States consumers will tend to
consume more relative to the rest of the regions considered.
Income (GDP)
Figure 6.3 presents total market demand responses to changes in the
income (GDP) level for major world consumers. The income index is shown
on the bottom axis. The total market demand index is presented on the
left axis. The expected relationship between income and demand is
positive. As the income index increases, it is expected that the demand
index also increases. The figure indicates that Latin America, Far East,
EC, and Middle East/North Africa have positive relationships. The
empirical results indicate that only the parameters for the EC and Middle
East/North Africa are significant. The results also indicate that these
parameters are inelastic (see Table 5.1).
The figure shows that the United States and Mediterranean-EC have
negative responses. The empirical results indicate that the parameter for
Mediterranean-EC is not significant. A negative total market demand index
response produces lower consumption as income increases. This is the case
for the United States, and it is an unexpected result. The negative
relationship shows that fresh oranges are considered an inferior good in
the United States. Consumers are expected to consume more of all normal


METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.44. Far East Imports of Fresh Oranges from the Middle East/North Africa (Product Demand 97).
184


300
The size of the fresh orange trade model developed here, with 440
equations of which 242 were estimated, leaves little space to improve
individual equations by correcting the functional form, the variables
included, or any other alternative solution. Large trade econometric
models like the one developed here are used to provide information about
major trends and shifts of trade flows and market shares through the years
among the different regions. The model provided important information
about the behavior of the fresh orange industry. This information could
be used for policy decisions in the different regions and countries.
If a particular trade flow is of interest and more information is
needed, it is possible to review the particular functional form and obtain
better results. However, if a single-equation estimation procedure is
used, the results suffer from simultaneity bias.
An important limitation of the present study is that it was not
possible to obtain the reduced-form parameters. If the reduced-form
parameters were found, then the whole system of equations could have been
simulated, given changes in the exogenous variables. Given the
simultaneity embodied in the model, this is an important drawback. The
limitation implies that the sensitivity analysis has to be developed on an
equation by equation basis.
Performance of the Model and Results
The graphical and the statistical analyses provided sufficient
information to determine that the model has a good fit, is well specified
and predicted most turning points. The economic analysis shows that the


38
Table 2.5 Trade Flow Analysis for Selected Years (1966, 1976 and 1986)
Without Intraregional Trade "Relative Region Imports from
Partner Regions"
Year
Region
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE COMMB
Total
Percentages -
66
US
0.0
0.0
82.1
0.8
0.0
0.0
16.3
0.0
0.7
0.0
0.0
100.0
76
0.0
0.0
95.2
0.6
0.0
0.0
0.9
0.0
2.9
0.3
0.0
100.0
86
0.0
0.0
45.2
32.0
0.1
0.0
19.7
0.0
3.0
0.0
0.0
100.0
66
CAN
77.8
0.0
0.8
0.0
0.0
0.0
4.7
8.9
7.6
0.0
0.0
100.0
76
84.7
0.0
0.1
0.0
0.0
0.0
1.4
2.9
10.2
0.7
0.0
100.0
86
68.0
0.0
2.9
10.2
0.0
0.0
11.8
0.0
5.9
1.2
0.0
100.0
66
LA
98.7
0.0
0.0
0.0
1.2
0.0
0.0
0.0
0.0
0.0
0.0
100.0
76
94.9
2.2
0.0
0.0
2.8
0.1
0.0
0.0
0.0
0.0
0.0
100.0
86
43.0
0.2
0.0
4.5
16.1
0.0
1.3
0.3
1.2
0.0
33.4
100.0
66
MED-EC
N. A
N.A
N.A
N.A
N.A
N.A
N.A
N.A
N.A
N.A
N.A
N.A
76
1.0
0.0
0.0
0.0
35.3
3.9
59.8
0.0
0.0
0.0
0.0
100.0
86
0.0
0.0
10.4
0.0
39.7
2.0
6.2
41.7
0.0
0.0
0.1
100.0
66
EC
2.4
0.0
2.9
47.1
0.0
0.0
38.3
9.0
0.0
0.1
0.0
100.0
76
4.0
0.0
1.5
57.8
0.0
0.0
29.0
7.5
0.0
0.1
0.1
100.0
86
0.3
0.0
5.4
67.4
0.0
0.0
21.2
5.4
0.0
0.1
0.4
100.0
66
RWE
2.6
0.0
0.8
59.9
0.4
0.0
31.7
A. A
0.1
0.1
0.0
100.0
76
1.1
0.0
0.7
40.8
0.8
0.0
51.5
4.8
0.0
0.4
0.0
100.0
86
0.3
0.0
1.1
58.4
2.8
0.0
34.0
3.3
0.0
0.1
0.0
100.0
66
ME/NA
0.1
0.0
0.0
9.0
0.1
0.0
0.0
0.0
85.7
5.1
0.0
100.0
76
0.0
0.0
0.1
22.8
3.9
0.0
0.0
42.0
31.2
0.0
0.0
100.0
86
0.0
0.0
5.9
29.7
0.2
0.0
0.0
0.0
53.8
10.3
0.0
100.0
66
RAF
0.6
0.0
0.1
1.1
11.8
0.0
83.0
0.0
0.0
3.4
0.0
100.0
76
0.1
0.0
3.5
14.1
9.3
0.0
71.2
0.0
1.1
0.7
0.0
100.0
86
0.0
0.0
0.1
8.0
12.5
0.0
70.5
0.0
0.0
8.9
0.0
100.0
66
FE
49.5
0.0
0.8
0.5
0.0
0.0
17.8
19.9
0.0
11.5
0.0
100.0
76
90.6
0.0
0.0
0.6
0.0
0.0
6.2
0.7
0.0
1.9
0.0
100.0
86
86.2
0.0
0.1
0.5
0.0
0.0
1.8
2.9
0.0
8.4
0.0
100.0
66
0CE
40.4
0.0
25.9
0.0
0.0
0.0
29.6
4.1
0.0
0.0
0.0
100.0
76
99.6
0.0
0.0
0.0
0.0
0.0
0.4
0.0
0.0
0.0
0.0
100.0
86
99.9
0.0
0.0
0.0
0.1
0.0
0.0
0.0
0.1
0.0
0.0
100.0
66
C0f*ffl
1.6
0.0
1.3
52.8
0.1
0.0
44.2
0.0
0.0
0.0
0.0
100.0
76
1.5
0.0
4.4
32.1
0.1
0.0
61.9
0.0
0.0
0.0
0.0
100.0
86
0.0
0.0
3.0
57.1
0.2
0.1
39.6
0.0
0.0
0.0
0.0
100.0


200
Population elasticities are reported in the third row. Six out of
11 regions have the correct positive elasticities. The United States,
Latin America, Mediterranean-EC, rest of Africa, and the Far East have
positive and significant population elasticities. The Communist Bloc has
a positive population elasticity, but it is not significant. The
discussion maintained above regarding Latin America and rest of Africa is
again confirmed with the results obtained. Population, and not price and
income, is the major driver of the marginal demands in these regions.
Price and income elasticities are insignificant for these regions.
The rest of the regions have negative population elasticities but
only the ones from Canada and the EC are apparently significantly
different from zero. This is an unexpected result, which implies that as
population increases, consumption of fresh oranges decreases. Unexpected
signs may be resulting from data or specification errors that are more
likely to occur in large models as used in this study.
The magnitudes of positive population elasticities range from .56
for the rest of Africa to 7.2 for Mediterranean-EC. The rest of them are
between one and 2.12. The results indicate highly elastic population
elasticities in most cases.
Substitute product price elasticities are shown in row four. Seven
out of the 11 regions have the correct positive elasticities, but only
United States and Oceania are significantly different from zero indicating
inelastic substitute product price elasticities. Consumption of fresh
oranges increases less than proportionally to increases in the price index


59
models. The multiple-product trade models were also divided into three
groups: general equilibrium (including agricultural and non-agricultural
products), multiple related commodity products (including only
agricultural products), and differentiated product models (differentiated
by place of origin). The two-region and the general equilibrium models
were special cases of non-spatial price equilibrium models. Thompson and
Abbott's (1982) classification procedure added important insights into the
discussion about new developments in international agricultural trade
models. The major contribution was their extensive treatment of and
emphasis on the characteristics of the products traded and how consumers
perceived them.
In the following discussion, Thompson's (1981) approach will be
followed. His classification was basically the same as the one presented
in Thompson and Abbott's (1982) investigation. The most important
difference between the two studies was the emphasis that the latter
researchers gave to product differentiation.
The first type of model covered by Thompson (1981) was the
two-region model. The model divided all countries of the world into two
groups, the country of interest and the rest of the world. This version
was basically a domestic agricultural sector model enlarged with
exogenously driven exported or imported quantities. Export equations or
excess demand equations were developed for the rest of the world. The
model included linkages between the domestic and world prices to reflect
the simultaneous determination of domestic consumption, supply, and prices
with the rest of the world. The models did not take into consideration
trade flows (destination) but instead accounted for the net trade between


METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.20. Total Export Supply of Fresh Oranges from the Far East.
154


161
Africa, and the rest of Africa products in the rest of Western Europe
respectively. The three product demands show that the model reflects the
trend. The best fit is obtained by the demand for Middle East/North
Africa product for which turning points are predicted by the model. The
demand for Mediterranean-EC product shows that the model generates only
some turning points, and for the rest of Africa product shows that just a
few turning points are captured.
Middle East/North Africa. Middle East/North Africa imports
represented 1.65% of total world imports from 1966 to 1986. The region
has been growing rapidly in terms of total market demand and trade in the
last 15 years. Figures 5.38 to 5.40 display the demand for Latin America,
rest of Africa, and Far East products in the Middle East/North Africa
respectively. The model reflects the trend of the product demands in
every case, but it is not predicting some turning points in each equation.
Rest of Africa. Rest of Africa imports represented 0.16% of total
world imports in the period considered. Figures 5.41 and 5.42 exhibit the
demand for EC and Middle East/North Africa products in the rest of Africa,
respectively. The model generates the trend in both cases. The figures
indicate that several turning points in each product demand are not
captured by the model.
Far East. Far East imports represented 4.3% of total world imports
in the 21-year period studied. This market has been growing fast in the
last two decades. Figures 5.43 to 5.45 show the demand for the United
States, Middle East/North Africa, and Oceania products in the Far East,
respectively. Figure 5.43 shows that the model closely reflects the
demand for the United States product in the Far East. Figures 5.44 and


APPENDIX H
EMPIRICAL RESULTS: PRODUCT DEMAND AND CIF
LINKAGE EQUATIONS STATISTICS


277
PRODUCT DEMAND INDEX (1986=1)
IMPORT PRICE INDEX (1986 = 1)
Figure 6.19. Middle East/North Africa Imports Changing Import Prices.


Table
H. 2
Mediterranean
Statistics
-EEC, EEC,
Rest of
Western Europe
Product
Demand and
CIF Price
Linkage
Equations
Region
= Mediterranean-
-EEC
Product
Demand Equations
CIF Price Linkage Equations
@0BS
@RSQ
@DW
@FST
UTHEIL
@OBS
@RSQ
@DW
@FST
UTHEIL
LA
10
0.27
1.67
1.28
0.736228
21
0.78
1.84
20.03
0.193848
EC
15
0.82
2.25
27.61
0.287940
21
0.57
2.54
7.66
0.190169
RWE
12
0.76
1.18
14.53
0.354966
21
0.84
2.72
29.74
0.092577
ME/NA
16
0.10
1.86
0.73
0.451282
21
0.64
2.52
10.20
0.315843
Region EEC
Product Demand Equations
CIF
Price Linkage Equations
@OBS
@RSQ
@DW
@FST
UTHEIL
80BS
@RSQ
@DW
@FST
UTHEIL
us
21
0.36
2.03
5.12
0.279022
21
0.85
2.10
31.16
0.087405
CAN
14
0.04
1.34
0.25
0.726797
21
0.62
2.62
9.32
0.193148
LA
21
0.19
0.74
2.16
0.189178
21
0.90
1.92
52.34
0.073432
MED-EC
21
0.42
1.45
6.61
0.075885
21
0.94
1.97
184.49
0.042842
RWE
21
0.23
1.44
2.66
0.251156
21
0.74
2.20
15.87
0.090920
ME/NA
21
0.27
0.56
3.30
0.065822
21
0.94
2.10
89.99
0.045453
RAF
21
0.10
0.57
0.95
0.070982
21
0.93
2.77
74.77
0.056549
FE
21
0.15
1.60
1.61
0.626029
21
0.66
2.21
10.97
0.126090
OCE
21
0.27
2.48
3.28
0.286791
21
0.78
1.83
20.58
0.075877
C0M1B
20
0.57
0.71
11.08
0.442780
21
0.88
2.34
40.39
0.058859
Region Rest of Western Europe
Product
Demand Equations
CIF
Price Linkage Equations
SOBS
0RSQ
@DW
8FST
UTHEIL
eoBs
gRSQ
@DW
@FST
UTHEIL
US
21
0.33
2.50
4.44
0.248258
21
0.90
1.36
52.20
0.081064
LA
21
0.13
1.29
1.36
0.143368
21
0.97
1.81
207.61
0.044011
MED-EC
21
0.28
1.24
3.41
0.086744
21
0.98
1.30
242.65
0.040090
EC
21
0.85
2.05
50.81
0.160083
21
0.81
2.81
24.78
0.060969
ME/NA
21
0.73
1.35
23.85
0.047351
21
0.98
1.80
368.62
0.029000
RAF
21
0.15
1.22
1.55
0.062720
21
0.98
2.27
225.44
0.032975
FE
17
0.01
2.32
0.10
0.740192
21
0.68
2.76
11.94
0.149778
OCE
21
0.28
0.89
3.51
0.330177
21
0.91
2.53
28.78
0.050677
COMB
18
0.42
1.98
5.50
0.280677
21
0.94
2.71
86.65
0.060959
323


232
index number for total market demand. The simulated or new total market
demand values are obtained by substituting the modified average market
prices into the original equation while holding all other variable values
fixed at the 1986 levels. The exact index on the left axis will depend on
the simulated values actually obtained, given the step-wise changes in the
average market price.
Rationale for Region. Equation, and Variable Selections
The analysis of the fresh orange industry shows that total
consumption, imports, and exports are concentrated in a few regions. Some
regions are major consumers, others major importers or exporters, and most
regions have a small set of important trade partners. Considering these
conditions, it is reasonable to select a subset of regions on which to
perform the sensitivity analysis in each event. In most cases, a few
regions will represent over 90% of total world consumption or exports, and
a few regions will also account for over 90% of total supply. Applying
the sensitivity analysis to all 262 estimated equations will provide
little additional information since many relationships will have almost no
impact on the major factors affecting the world fresh orange industry.
Total market demand equations represent total consumption or demand
for fresh oranges in a given region. The model developed considers both
total domestic consumption and total trade. Therefore, total demand and
total imports per region relative to the rest of the world should be
considered to select the more important regions. Given that a small group
of regions represented major world consumers and another small but


290
are not a major factor for import decisions in the Communist Bloc. The
figure also indicates that demand for Latin America product has the
strongest response. If import prices for Latin America and the Middle
East/North Africa increase proportionally, consumers will consume
relatively more product from the Middle East/North Africa than from Latin
America. If import prices decrease, then the reverse is true.
Figure 6.28 shows Communist Bloc imports changing total market
demand. The results indicate that the three product demands have a
positive and significant response to changes in total market demand. The
strongest response is for Latin America product, followed by Middle
East/North Africa product. The response for the Mediterranean-EC is small
but significant. The sensitivity analysis implies that, if the Communist
Bloc market size increases, Latin America product will have the strongest
position to penetrate the market. The Middle East/North Africa will have
the second position and Mediterranean-EC the last one.
Mediterranean-EC has an insignificant import price/product demand
relationship. This condition, combined with the analysis on the market
size, gives this region an advantageous position to penetrate the
Communist Bloc market.
Summary
This section of the chapter discussed the sensitivity analysis
results. Regions were analyzed and compared with other regions. Each
major equation of the model was discussed separately. The results in


272
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.16. EC Imports Changing Total Market Demand.


60
the country of interest and the rest of the world. They did not provide
information on demand and supply for individual foreign regions or on the
share of the market that any particular country has in a specific region.
Without knowledge of the structure of supply and demand in each
major trading region, it is impossible to say how the excess demand
function will change given an exogenous shock or a change in policy. It
is then very difficult under the two-region models to evaluate the impact
of shocks or policies in a given country. Such models do, however,
provide a good framework to analyze domestic farm and trade policies.
According to Thompson (1981) multiple-region world trade models
were developed to answer broader questions regarding the impact of
exogenous shocks and policy changes for trading regions in the world.
They also provide information about the market share of each region by
destination. The non-spatial price equilibrium models treat the
interrelations among trading regions by assuming that the world market
price is determined simultaneously by the demand-supply balance in all
trading regions such that the world market clears. Solution of the model
gives the world market prices and the net trade for each region, but it
does not provide any information on source or destination of trade flows.
Multiple-region world trade models allow for the introduction of
transportation costs, tariffs and nontariff barriers, and other policy
variables through the price linkage equations. These models are for many
reasons an improvement over the two-region models, since they endogenously
determine the demand and supply in each of the trading regions. However,
they usually have an important drawback. The price linkage frequently
used is not consistent with the spatial price equilibrium theory. This is


55
participation in recent years, passing from 0% in the 1960s to 1.2% in
1986. The rest of the regions were not very important with regard to
exports to the Canadian region.
The rest of the regions represented small percentages of total
imports in the world's fresh orange industry (see Table 2.3). The United
States import share was .8% in 1966, .7% in 1976, and .9% in 1986. These
percentages changed very little if only interregional trade were
considered. Major exporters to the United States were Latin America with
45.2%, Mediterranean-EC with 32.0%, and Middle East/North Africa with
19.7% in 1986. The Mediterranean-EC only had .8% and .6% share of the
United States market in 1966 and 1976, respectively, indicating that
Mediterranean-EC region's participation in the United States has been
growing rapidly in the last decade.
The Oceania portion of total world imports was .4% in 1966, .2% in
1976, and .4% in 1986 (see Table 2.3). These percentages switched to .2
each reported year if only interregional trade were included.
The exporter with the major portion of the Oceania region's market
was the United States, with 40.4% in 1966, 99.6% in 1976, and 99.9% in
1986 (see Table 2.5). Middle East/North Africa and Latin America regions
used to have an important share of the Oceania market, reaching 29.6% and
25.9%, respectively in 1966. These regions lost their portion of the
market to the United States in the 1970s. The rest of the regions were
not major exporters to Oceania.
The rest of Africa's share of total world imports was .3% in 1966,
and .2% in 1976 and 1986 including intraregional trade (see Table 2.3).


279
positive output for market size increases. Latin America and rest of
Africa will probably have the major relative gains. It is worth
mentioning that this market has been growing rapidly in the last two
decades. This establishes an excellent opportunity for exports,
especially from Latin America and rest of Africa.
Rest of Africa
Table 6.2 indicates that 99.6% of rest of Africa total imports came
from five regions: the Middle East/North Africa, EC, Mediterranean-EC,
Latin America, and Oceania. The sensitivity analysis shown in Figure 6.21
presents rest of Africa imports while changing import prices. The figure
indicates that three out of the five regions have negative relationships.
The correct negative associations correspond to the Mediterranean-EC,
Latin America, and Oceania. The demand elasticity for Mediterranean-EC
product is not significant. The results indicate that, if a similar
increase in the import price of Latin America and Oceania products occur,
consumers will tend to consume more product from Latin America relative to
Oceania. Demand elasticities for Middle East/North Africa and EC products
are positive, but only the former is significant. The wrong direction of
the relationship for the Middle East/North Africa product could be related
to the fact that the rest of Africa is self-sufficient and a net exporter
of fresh oranges. It is possible to argue that imports from its major
supplier occur only when domestic supply is insufficient and prices are
rising.
Figure 6.22 shows rest of Africa imports while changing total market
demand. As shown in the figure, only the demand for Oceania product has


334
Figure 6.6
FOB
Average
Export
Price
OBS
Index
US
CAN
LA
MED-EC
1
0.7
0.72
0.70
0.80
0.79
0.
2
0.8
0.82
0.80
0.87
0.87
0.
3
0.9
0.91
0.90
0.93
0.93
0.
4
1.0
1.00
1.00
1.00
1.00
1.
5
1.1
1.09
1.10
1.06
1.06
1.
6
1.2
1.18
1.20
1.12
1.12
1.
7
Figure
1.3
6.7
Import
Price
1.27
1.30
1.18
1.18
1.
OBS
Index
US
CAN
LA
MED-EC
1
0.7
0.60
0.69
1.36
1.
2
0.8
0.73
0.79
1.21
1.
3
0.9
0.86
0.90
1.09
1.
4
1.0
1.00
1.00
1.00
1.
5
1.1
1.15
1.11
0.92
0.
6
1.2
1.30
1.21
0.86
0.
7
Figure
1.3
6.8
Total
Market
Demand
1.46
1.32
0.80
0.
OBS
Index
US
CAN
LA
MED-EC
1
0.7
6.16
0.31
0.00
0.
2
0.8
3.12
0.48
0.00
0.
3
0.9
1.71
0.71
0.06
0.
4
1.0
1.00
1.00
1.00
1.
5
1.1
0.62
1.37
13.56
1.
6
1.2
0.39
1.82
146.45
1.
7
Figure
1.3
6.9
Import
Price
0.26
2.38
1307.57
1.
OBS
Index
US
CAN
LA
MED-EC
1
0.7
1.22
1.18
0.33
1.
2
0.8
1.13
1.11
0.50
1.
3
0.9
1.06
1.05
0.72
1.
4
1.0
1.00
1.00
1.00
1.
5
1.1
0.95
0.96
1.34
0.
6
1.2
0.90
0.92
1.76
0.
7
Figure
1.3
6.10
Total
Market
Demand
0.86
0.88
2.25
0.
OBS
Index
US
CAN
LA
MED-EC
1
0.7
0.61
752.81
4721.93
0.
2
0.8
0.74
63.05
198.89
0.
3
0.9
0.87
7.08
12.17
0.
4
1.0
1.00
1.00
1.00
1.
5
1.1
1.14
0.17
0.10
3.
6
1.2
1.28
0.03
0.01
11.
7
1.3
1.43
0.01
0.00
32.
RWE
ME/NA
RAF
FE
OCE
COTOffl
0.70
0.65
0.88
0.95
0.90
0.47
0.80
0.77
0.92
0.97
0.93
0.62
0.90
0.88
0.96
0.98
0.97
0.80
1.00
1.00
1.00
1.00
1.00
1.00
1.13
1.12
1.03
1.01
1.03
1.23
1.20
1.24
1.07
1.03
1.06
1.48
1.30
1.37
1.10
1.04
1.08
1.76
RWE
ME/NA
RAF
FE
OCE
COtB
1.00
7.28
1.00
0.07
1.71
1.00
1.00
3.46
1.00
0.19
1.40
1.00
1.00
1.80
1.00
0.45
1.17
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.59
1.00
2.05
0.87
1.00
1.00
0.36
1.00
3.93
0.76
1.00
1.00
0.23
1.00
7.17
0.67
1.00
RWE
ME/NA
RAF
FE
OCE
COMMB
1.00
0.02
1.00
0.81
0.00
1.00
1.00
0.07
1.00
0.88
0.00
1.00
1.00
0.29
1.00
0.94
0.04
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
3.05
1.00
1.06
17.55
1.00
1.00
8.43
1.00
1.11
239.99
1.00
1.00
21.49
1.00
1.16
2661.57
1.00
RWE
ME/NA
RAF
FE
OCE
COMMB
1.00
1.07
0.88
1.39
2.48
1.00
1.00
1.04
0.92
1.23
1.76
1.00
1.00
1.02
0.96
1.10
1.31
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.98
1.04
0.92
0.78
1.00
1.00
0.97
1.07
0.84
0.63
1.00
1.00
0.95
1.10
0.78
0.51
1.00
RWE
ME/NA
RAF
FE
OCE
COMMB
1.00
5.03
5.99
0.32
0.01
1.00
1.00
2.75
3.07
0.49
0.04
1.00
1.00
1.31
1.70
0.71
0.22
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.65
0.62
1.35
3.99
1.00
1.00
0.44
0.40
1.79
14.10
1.00
1.00
0.30
0.27
2.31
45.05
1.00
EC
78
86
93
00
07
13
,20
EC
5*.
31
14
00
89
80
73
EC
,75
83
92
00
08
16
.24
EC
75
42
18
00
86
75
66
EC
01
05
25
00
54
22
42


APPENDIX E
TARIFF DATA
Region
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
COttdB

X OF FOB EXPORT
PRICE
-
USa
0
22.05
22.05
22.05
22.05
22.05
22.05
22.05
22.05
22.05
22.05
CAN
0
0
0
0
0
0
0
0
0
0
0
SA
25
25
25
25
25
25
25
25
25
25
25
MED-EECbc
66
12
12
12
10
2.8
12
6.7
12
12
12
12
67
12
12
12
10
2.8
12
6.7
12
12
12
12
68
12
12
12
10
2.8
12
6.7
12
12
12
12
69
12
12
12
10
2.8
12
6.7
12
12
12
12
70
12
12
12
10
2.8
12
6.7
12
12
12
12
71
12
12
12
10
2.8
12
6.7
12
12
12
12
72
12
12
12
10
2.8
12
6.7
12
12
12
12
73
12
12
12
8.2
2.8
12
6.39
12
12
12
12
74
12
12
12
8.2
2.8
12
6.39
12
12
12
12
75
12
12
12
8.2
2.8
12
6.39
12
12
12
12
76
12
12
12
8.2
2.8
12
6.39
12
12
12
12
77
12
12
12
8.2
2.8
12
6.39
12
12
12
12
78
12
12
12
8.2
2.8
12
6.39
12
12
12
12
79
12
12
12
8.2
2.8
12
6.19
12
12
12
12
80
12
12
12
8.2
2.8
12
6.19
12
12
12
12
81
12
12
12
8.2
2.8
12
6.19
12
12
12
12
82
12
12
12
8.2
2.8
12
6.19
12
12
12
12
83
12
12
12
8.2
2.8
12
6.19
12
12
12
12
84
12
12
12
8.2
2.8
12
6.19
12
12
12
12
85
12
12
12
8.2
2.8
12
6.19
12
12
12
12
86
12
12
12
8.2
2.8
12
6.19
12
12
12
12
EECbc
66
15
15
15
11.43
0
15
6.5
15
15
15
15
67
15
15
15
11.43
0
15
6.5
15
15
15
15
68
15
15
15
11.43
0
15
6.5
15
15
15
15
69
15
15
15
11.43
0
15
6.5
15
15
15
15
70
15
15
15
11.43
0
15
6.5
15
15
15
15
71
15
15
15
11.43
0
15
6.5
15
15
15
15
72
15
15
15
11.43
0
15
6.5
15
15
15
15
73
15
15
15
8.59
0
15
5.98
15
15
15
15
74
15
15
15
8.59
0
15
5.98
15
15
15
15
75
15
15
15
8.59
0
15
5.98
15
15
15
15
76
15
15
15
8.59
0
15
5.98
15
15
15
15
77
15
15
15
8.59
0
15
5.98
15
15
15
15
78
15
15
15
8.59
0
15
5.98
15
15
15
15
79
15
15
15
8.76
0
15
5.65
15
15
15
15
80
15
15
15
8.76
0
15
5.65
15
15
15
15
81
15
15
15
8.76
0
15
5.65
15
15
15
15
82
15
15
15
8.76
0
15
5.65
15
15
15
15
83
15
15
15
8.76
0
15
5.65
15
15
15
15
84
15
15
15
8.76
0
15
5.65
15
15
15
15
85
15
15
15
8.76
0
15
5.65
15
15
15
15
86
15
15
15
8.76
0
15
5.65
15
15
15
15
RWE
0
0
0
0
0
0
0
0
0
0
0
ME/NA
5
5
5
5
5
5
5
5
5
5
5
RAF
5
5
5
5
5
5
5
5
5
5
5
FE
40
40
20
40
40
40
40
20
40
20
40
OCE
0
0
0
0
0
0
0
0
0
0
0
COMMB
10
10
2.5
10
10
10
2.5
10
10
10
0
U.S. dollars per metric ton.
bZ of CIF import price and tariffs vary by year.
cTariffs differ by year.
311


268
PRODUCT DEMAND INDEX (1986=1)
30.0
25.0
20.0
15.0
10.0
5.0
o.offl
0.7
a

EC
|
LA
*
ME/NA
-B-
RWE
0.8 0.9 1 1.1
TOTAL MARKET DEMAND INDEX (1986=1)
1.2
1.3
Figure 6.14. Mediterranean-EC Imports Changing Total Market Demand.


316
FRML EQ6#20 LIQ6D (RH06 + RH16*(LRMP6DH) + RH26*LOG(GDP6/CPI6)
+ RH36*LOG(POP6) + RH46*L0G(BAVAL6/CPI6));
FRML EQ7//20 LIQ7D (RH07 + RH17*(LRMP7DH) + RH27*LOG(GDP7/CPI7)
+ RH37*LOG(POP7) + RH47*L0G(BAVAL7/CPI7)) ;
FRML EQ8//20 LIQ8D (RH08 + RH18*(LRMP8DH) + RH28*LOG(GDP8/CPI8)
+ RH38*LOG(POP8) + RH48*LOG(BAVAL8/CPI8));
FRML EQ9#20 LIQ9D = (RH09 + RH19*(LRMP9DH) + RH29*LOG(GDP9/CPI9)
+ RH39*LOG(POP9) + RH49*LOG(BAVAL9/CPI9));
FRML EQ10#20 LIQ10D = (RHOIO + RH110*(LRMP10DH) + RH210*LOG(GDP10/CPI10)
+ RH310*LOG(POPIO) + RH410*LOG(BAVAL10/CPI10));
FRML EQ11#20 LIQ11D (RHOll + RH111*(LRMP11DH) + RH211*L0G(GDP11/CPI11)
+ RH311*LOG(POPll) + RH411*L0G(BAVAL11/CPI11));
(NOPRINT,
(NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
(NOPRINT,
(NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ
LSQ
LSQ
LSQ
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
EQ1#19;
EQ2#19;
EQ3#19;
EQ4//19;
EQ5//19;
EQ6#19;
EQ7//19;
EQ8#19;
EQ9#19;
EQ10y/19
EQliy/19
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
¡PRINT @RSQ
¡PRINT @RSQ
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW
@DW
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
, @FST;
, @FST;
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
END;
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
EQiy/20;
EQ2#20;
EQ3#20;
EQ4#20;
EQ5#20;
EQ6#20;
EQ7//20;
EQ8#20;
EQ9y/20;
EQ10#20
EQliy/20
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT (3RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
¡PRINT @RSQ,
¡PRINT @RSQ,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
Program y/2
? REGION 1;
FREQ A;
SMPL 66,86;
READ (FORMAT=LOTUS,FILE='C:\LOTUS\DATA.WK1');
? START OF PROGRAM;
LIST EXOG;
POPI GDP1 PRD1 CPU YEAR PEN BAVAL1;
POP2 GDP2 PRD2 CPI2 BAVAL2;
POP3 GDP3 PRD3 CPI3 BAVAL3;
P0P4 GDP4 PRD4 CPI4 BAVAL4;


278
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.20. Middle East/North Africa Imports Changing Total Market
Demand.


Table 1.6 World Fresh Orange Imports by Region
Region
1966
1976
1978
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
Annual
Growth
Rate
1978-86
(000) Metric
Tons
Percent of Changi
e
United States
28.3
32.6
52.9
49.4
2.8
4.2
-0.8
Canada
180.5
225.5
180.2
182.1
.0
-2.1
0.1
Latin America
9.7
18.1
14.2
3.8
-4.6
-14.5
-15.2
Mediterranean-EC
0.0
0.9
1.5
8.2
N. A.
24.8
24.2
E.C.
2398.4
2736.2
2655.5
3464.6
1.9
2.4
3.4
Rest of Western Europe
384.6
439.7
436.0
551.2
1.8
2.3
3.0
Middle East/North Africa
22.0
590.4
440.5
279.0
13.5
-7.2
-5.5
Rest of Africa
10.6
10.6
14.5
8.2
-1.3
-2.5
-6.8
Far East
119.3
238.9
251.7
382.8
6.0
4.8
5.4
Oceania
15.3
8.9
15.5
20.4
1.4
8.7
3.5
Communist Bloc
315.6
718.8
865.8
445.8
1.7
-4.7
-8.0
World Total
3484.5
5020.7
4928.3
5395.6
2.2
0.7
1.1
Source: United Nations Trade Data Tapes.


Table 5.17 Mediterranean-EC CIF Price Linkage Equations
Partner
Region
Intercept
( + \-)
FOB
Price
< + )
Year
Trend
(-\ + )
Energy
Index
Price
( + >
gOBS
gRSQ
§DW
gFST
21 Years
Z of
Total
UTHEIL Imports
Average
Z Total
Market
Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
5
0.085
0.000
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
0.000
0.000
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-26.756
15.945
-1.678
0.090
0.464
0.193
5.846
3.578
1.634
0.109
0.177
0.618
21
0.78
1.84
20.03
0.193848
22.019
0.022
EC
PARAM.VALUE
STD.ERROR
t STATISTIC
-11.072
9.503
-1.165
0.222
0.522
0.426
2.425
2.211
1.097
0.059
0.166
0.357
21
0.57
2.54
7.66
0.190169
47.784
0.048
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
-17.337
13.007
-1.333
1.088
0.338
3.219
4.199
2.987
1.406
-0.289
0.134
-2.155
21
0.84
2.72
29.74
0.092578
1.747
0.002
ME/NA
PARAM. VALUE
STD. ERROR
t STATISTIC
27.404
20.916
1.310
1.192
0.442
2.698
-6.370
4.805
-1.326
0.366
0.249
1.473
21
0.64
2.52
10.20
0.315843
17.668
0.018
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
5
9.973
0.010
FE
PARAM. VALUE
STD. ERROR
t STATISTIC
2
0.310
0.000
OCE
PARAM.VALUE
STD.ERROR
t STATISTIC
1
0.042
0.000
COMMB
PARAM. VALUE
STD.ERROR
t STATISTIC
3
0.372
0.000
Total
100.000
0.100


85
Again, Xjj is the amount of product produced domestically and remains
in the same region. Total market demand is Xj where Xjj is a subset of
XJ.-
Equilibrium Conditions
The equilibrium conditions required to have a closed model for the
fresh orange industry include three basic identities. The total market
demand in region i must equal the supply of products to that region
(4.16) Xi. = Xj Xu
Total market demand in region j (Xj ) must equal fresh utilization
(PRDjj) plus imports (XA iflj X^) minus exports (Xj ij4J Xi) as follows:
(4.17) Xj PRDij + Xt ifij Xji Xj Xld
Finally, total production of oranges must equal the total production
used fresh plus the total production used processed
(4.18) PRD j = PRDu + PRD2j
Price Linkage Equations
Total market and product demand as well as export supply are
functions of different but closely related set of prices. Total market
demand is a function of the average market price for a particular
commodity. This price is associated with the local price and the
individual product prices (Py) Each product demand is a function of its
own product price and indirectly a function of the individual product
prices (Pij) through the average market price (Pi.). Export supply is


5.15 Total Export Supply of Fresh Oranges from the
Mediterranean-EC 149
5.16 Total Export Supply of Fresh Oranges from EC 150
5.17 Total Export Supply of Fresh Oranges from the
Rest of Western Europe 151
5.18 Total Export Supply of Fresh Oranges from Middle
East/North Africa 152
5.19 Total Export Supply of Fresh Oranges from the
Rest of Africa 153
5.20 Total Export Supply of Fresh Oranges from the
Far East 154
5.21 Total Export Supply of Fresh Oranges from
Oceania 155
5.22 Total Export Supply of Fresh Oranges from the
Communist Bloc 156
5.23 United States Imports of Fresh Oranges from
Latin America (Product Demand 1_3) 163
5.24 United States Imports of Fresh Oranges from the
Middle East/North Africa (Product Demand 1_7) 164
5.25 Canada Imports of Fresh Oranges from the United
States (Product Demand 2_1) 165
5.26 Canada Imports of Fresh Oranges from the Middle
East/North Africa (Product Demand 2_7) 166
5.27 Canada Imports of Fresh Oranges from the Far
East (Product Demand 2_9) 167
5.28 Latin America Imports of Fresh Oranges from the
United States (Product Demand 3_1) 168
5.29 Latin America Imports of Fresh Oranges from the
EC (Product Demand 3_5) 169
5.30 Mediterranean-EC Imports of Fresh Oranges from
Latin America (Product Demand 4_3) 170
5.31 Mediterranean-EC Imports of Fresh Oranges from
the EC (Product Demand 4_5) 171
5.32 EC Imports of Fresh Oranges from the
Mediterranean-EC (Product Demand 5_4) 172
xi


CHAPTER 6
ECONOMIC IMPLICATIONS FROM SENSITIVITY ANALYSIS
Introduction
Two basic objectives were laid out in Chapter 1 regarding the
development of the fresh orange trade model. The first objective was to
develop a model to understand the major driving factors affecting world
fresh orange consumption and trade. This was accomplished in the
discussion in Chapter 5. The second objective was to determine what
happens when variables in the model change. In other words, what are the
comparative static implications of the trade model.
This chapter sets forth a sensitivity-analysis procedure to evaluate
the consequences of changes in the main variables of the model. The
results obtained from applying the procedure will complement the
discussion of Chapter 5 and will add new insights into the behavior of the
model.
In this chapter, the more important responses were illustrated by
selecting the major partner regions for each region. The variables that
better explain the model were also selected to be modified by the
sensitivity analysis. The comparative static implications in each case
were assessed. To illustrate the relative responses, scale effects were
removed by indexing the variables to the base year. The chapter also
provides a graphical presentation of the results, which helped to
226


133
Given that the model includes a total of 11 regions, there will be
at least ten product demand equations per region. Each product demand
will represent the region's demand for fresh oranges originating in the
other ten regions. Each table represents one final market or importing
region. The names of the partners or regions of origin are shown in the
first column. The second column explains the values appearing under each
variable. The values are the same presented in Tables 5.1 and 5.2; i.e.,
parameter values, standard errors, and "t" statistics. The following
columns display an intercept and the name of the variables included in the
product demand equations for all regions. The variables are relative
prices and total market demand. The following five columns show the
number of observations and the same statistics used for total market
demand and export supply equations. The last two columns show the
relative importance of each partner region's exports to total imports and
total market demand in the final market region, respectively. Estimated
product demand equations are less than ten in some regions due to
insufficient data points.
Tables 5.14 to 5.24 show the results for the CIF price linkage
equations. Each table refers to one region with a maximum of ten
estimated equations. CIF price linkage equations link every region's
import price with the FOB export price for the rest of the regions in the
model. Therefore, there are unique CIF and FOB prices for every trade
flow. Consequently, as in the case for product demands, each region is
associated with the other ten regions through an equation.
The basic structure of Tables 5.14 to 5.24 is the same as the one
presented for Tables 5.3 to 5.13, and hence it will not be repeated here.


227
visualize the pattern of adjustments to specific variables. It also
facilitated the comparison of the responses among partner regions in each
region. This comparison is not easy to see when looking only at the
coefficients, especially given the size of the model and the number of
parameters estimated. None of the analysis up to this point dealt with
adjustments in the variables. It was mainly a discussion on the
coefficients sign, magnitude, and significance. Certain variables have
important policy implications that could be clarified by using the
information presented here. As an example, given the characteristics of
the fresh orange trade data, a range of 30% above and below the base year
was considered reasonable. This gives an indication of the type of
responses and their limits for the fresh orange trade model (for example,
for relative prices). Much of the information in this chapter is intended
to help the reader to have a better understanding of the full model faced
with such a large number of coefficients or elasticities.
The chapter is divided into four sections. The first section
develops and explains the procedure for the sensitivity analysis. The
second discusses the rationale utilized to select the regions, equations,
and variables to be analyzed. The third develops the sensitivity analysis
for selected regions and equations, including a detailed discussion about
the results. The fourth summarizes the major conclusions and implications
of the chapter.


OLSQ LREPD8, C, PCI, PC2, PC3, PC4, PC5, PC6; LREPD8H@FIT;
OLSQ LREPD9, C, PCI, PC2, PC3, PC4, PC5, PC6; LREPD9H-@FIT;
OLSQ LREPD10, C, PCI, PC2, PC3, PC4, PC5, PC6; LREPD10HH§FIT;
OLSQ LREPD11, C, PCI, PC2 PC3 PC4, PC5, PC6 ; LREPDllH= RMP1D MP1D/CPI1;LRMP1D LOG(RMPID);
RMP2D MP2D/CPI2;LRMP2D LOG(RMP2D);
RMP3D MP3D/CPI3;LRMP3D LOG(RMP3D);
RMP4D = MP4D/CPI4;LRMP4D LOG(RMP4D);
RMP5D MP5D/CPI5;LRMP5D LOG(RMP5D);
RMP6D MP6D/CPI6;LRMP6D LOG(RMP6D);
RMP7D = MP7D/CPI7;LRMP7D = LOG(RMP7D);
RMP8D MP8D/CPI8;LRMP8D LOG(RMP8D);
RMP9D MP9D/CPI9;LRMP9D LOG(RMP9D);
RMPIOD MP10D/CPI10;LRMP10D LOG(RMPIOD);
RMP11D MP11D/CPI11;LRMP11D LOG(RMPllD);
OLSQ LRMP1D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP1DH=@FIT
OLSQ LRMP2D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP2DH=@FIT
OLSQ LRMP3D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP3DH=@FIT
OLSQ LRMP4D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP4DH=@FIT
OLSQ LRMP5D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP5DH=@FIT
OLSQ LRMP6D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP6DH= OLSQ LRMP7D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP7DH=@FIT
OLSQ LRMP8D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP8DHH2FIT
OLSQ LRMP9D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP9DH=@FIT
OLSQ LRMP10D,
c,
PCI,
PC2
PC3
PC4
PC5,
PC6
LRMP10DH=@FIT
OLSQ LRMP11D,
c,
PCI,
PC2
PC3
PC4
PC5,
PC6
LRMP11DH=@FIT
FRML EQiy/19 LEXQD1 (DH01 + DH11*(LREPD1H) + DH21*LOG(PRDl));
FRML EQ2#19 LEXQD2 = (DH02 + DH12*(LREPD2H) + DH22*LOG(PRD2));
FRML EQ3//19 LEXQD3 = (DH03 + DH13*(LREPD3H) + DH23*LOG(PRD3));
FRML EQ4#19 LEXQD4 = (DH04 + DH14*(LREPD4H) + DH24*LOG(PRD4));
FRML EQ5#19 LEXQD5 = (DH05 + DH15*(LREPD5H) + DH25*LOG(PRD5));
FRML EQ6#19 LEXQD6 (DH06 + DH16*(LREPD6H) + DH26*LOG(PRD6));
FRML EQ7#19 LEXQD7 (DH07 + DH17*(LREPD7H) + DH27*LOG(PRD7));
FRML EQ8//19 LEXQD8 (DH08 + DH18*(LREPD8H) + DH28*LOG(PRD8)) ;
FRML EQ9#19 LEXQD9 (DH09 + DH19*(LREPD9H) + DH29*LOG(PRD9));
FRML EQ10//19 LEXQD10 (DHOIO + DH110*(LREPD10H) + DH210*LOG(PRD10));
FRML EQliy/19 LEXQD11 (DH011 + DH111*(LREPD11H) + DH211*LOG(PRDll));
FRML EQ1#20 LIQ1D = (RHOl + RH11*(LRMP1DH) + RH21*L0G(GDP1/CPI1)
+ RH3l*LOG(POPI) + RH41*L0G(BAVAL1/CPI1));
FRML EQ2#20 LIQ2D (RH02 + RH12*(LRMP2DH) + RH22*LOG(GDP2/CPI2)
+ RH32*LOG(POP2) + RH42*LOG(BAVAL2/CPI2));
FRML EQ3#20 LIQ3D (RH03 + RH13*(LRMP3DH) + RH23*LOG(GDP3/CPI3)
+ RH33*LOG(POP3) + RH43*LOG(BAVAL3/CPI3));
FRML EQ4//20 LIQ4D = (RH04 + RH14*(LRMP4DH) + RH24*LOG(GDP4/CPI4)
+ RH34*LOG(POP4) + RH44*L0G(BAVAL4/CPI4));
FRML EQ5#20 LIQ5D (RH05 + RH15*(LRMP5DH) + RH25*LOG(GDP5/CPI5)
+ RH35*LOG(POP5) + RH45*LOG(BAVAL5/CPI5));


263
Latin America
Table 6.2 indicates that four regions accounted for 98.4% of total
imports in Latin America. The regions are the United States, EC,
Communist Bloc, and Middle East/North Africa. The sensitivity analysis
presented in Figure 6.11 shows Latin America imports given changes in
import prices. The EC and Middle East/North Africa have the correct
negative association. The United States and Communist Bloc have positive
relationships. Only the demands for the United States and Middle
East/North Africa products are significant (see Table 5.5). The United
States product demand has an unexpected result. It is probably related to
the fact that Latin America is self-sufficient, and imports exist only
when the market is experiencing a substantial shortage. If this is true,
the direction of the relationship could be positive. On the other hand,
it is expected that demand for Middle East/North Africa product will tend
to decline relative to other products as import prices increase. The
reverse could be expected if prices decrease.
Figure 6.12 presents Latin America imports while changing total
market demand. The figure indicates that the United States and Middle
East/North Africa have negative relationships, while the EC and Communist
Bloc have positive associations. As total market demand increases,
consumption for United States and Middle East/North Africa products will
decline relative to other regions. This is especially true in this case,
since the Communist Bloc and EC import price variables were not
significant, indicating that import prices will probably not affect their
demands. The Communist Bloc has the strongest relationship between


72
Wardowski et al. (1986) recently edited a book that includes a
descriptive analysis of world production practices and trends and a long
term view of fresh citrus trade. An interesting discussion on trade flow
and market share of imports and exports was presented. Global trade for
the 1980s was projected, based on the assumption that trade will growth
one third less rapidly than in the previous decade, given special
assumptions for each citrus product. Individual country/region
projections were based on historical trends in per capita availability,
where such trends were evident or trends on total imports were estimated.
The use of trends in projecting import demand was based on the assumption
that future levels of economic factors that determine demand will follow
historic trends. No demand estimation was pursued to determine trade flow
and market share in this study.
Lee and Fairchild (1988) used a SUR technique to study the
relationship between exchange rates and foreign demand for United States
fresh grapefruit. The results showed that exchange rates played a major
role when studying export demand relationships and the United States fresh
grapefruit has more than one export market, with markets responding
differently to price changes. These results will be used later to define
the model to be estimated.
Lee et al. (1990), using the absolute version of the Rotherdam
model, studied the Japanese citrus products market. The study used fresh
bananas and pineapples as substitutes for fresh citrus products. One of
the major conclusions of the study was that United States fresh grapefruit
exports compete against imports of bananas and pineapples for Japanese
import dollars. In the case of fresh oranges, the results were not


119
Table 5.12 Oceania Product Demands
Partner
Region
Intercept
(+\->
Relative
Price
(-)
Total
Market
Demand
(-\ + )
SOBS
@RSQ
@DW
@FST
UTHEIL
21 Year Average
Z Z Total
Total Market
Imports Demand
us
PARAM. VALUE
STD. ERROR
t STATISTIC
-23.433
18.241
-1.285
-1.235
0.793
-1.558
2.663
1.494
1.782
21
0.21
0.53
2.45
0.268945
85.922
3.342
CAN
PARAM.VALUE
STD. ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM. VALUE
STD. ERROR
t STATISTIC
56.815
27.151
2.093
3.978
2.907
1.368
-4.127
2.208
-1.869
7
0.72
2.09
5.19
0.110958
4.504
0.175
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-69.670
71.041
-0.981
5.089
2.053
2.479
5.972
5.829
1.025
7
0.67
0.52
4.13
0.389675
1.868
0.073
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
5
0.027
0.001
EWE
PARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
197.030
55.499
3.550
-6.590
1.154
-5.710
-15.478
4.520
-3.424
13
0.77
1.25
16.68
0.375281
4.878
0.190
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
6
2.626
0.102
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
-22.286
83.976
-0.265
-1.408
4.445
-0.317
2.104
6.744
0.312
7
0.08
0.11
0.18
0.821817
0.175
0.007
COMMB
PARAM.VALUE
0
0.000
0.000
STD.ERROR
t STATISTIC
100.000 3.890
Total


Table H.3 Middle East/North Africa, Rest of Africa and Far East Product Demand and CIF Price Linkage
Equations Statistics
Region Middle East/Horth Africa
Product Demand Equations CIF Price Linkage Equations
gOBS
gRSQ
gDW
gFST
UTHEIL
gOBS
gRSQ
gDW
gFST
UTHEIL
US
18
0.34
1.20
3.83
0.875418
21
0.21
2.44
1.53
0.789752
LA
13
0.29
2.05
2.03
0.439828
21
0.88
2.34
42.08
0.101174
MED-EC
18
0.32
1.17
3.60
0.610091
21
0.64
2.73
9.89
0.461456
EC
16
0.48
2.79
6.05
0.662988
21
0.71
2.40
13.99
0.198065
RWE
11
0.24
1.48
1.24
0.511070
21
0.61
2.84
8.85
0.308409
RAF
17
0.60
0.72
10.58
0.356633
21
0.90
2.29
52.72
0.070873
FE
21
0.57
0.92
11.83
0.365247
21
0.50
3.11
5.69
0.289771
OCE
20
0.77
1.41
27.91
0.350599
21
0.52
2.34
6.23
0.222936
COMB
7
0.64
1.45
3.55
0.594771
21
0.71
2.05
13.87
0.264828
Region
= Rest of Africa
Product Demand Equations
CIF
Price Linkage Equations
gOBS
gRSQ
gDW
gFST
UTHEIL
gOBS
gRSQ
gDW
gFST
UTHEIL
us
17
0.20
0.75
1.75
0.444632
21
0.67
1.95
11.26
0.193657
LA
21
0.66
1.68
17.09
0.440843
21
0.88
1.77
42.74
0.109284
MED-EC
21
0.48
1.16
8.33
0.354209
21
0.67
2.83
11.74
0.148992
EC
21
0.14
2.12
1.47
0.123770
21
0.96
1.10
149.29
0.032802
ME/NA
21
0.16
2.41
1.72
0.092017
21
0.85
1.75
32.86
0.086437
FE
13
0.66
1.25
9.56
0.367392
21
0.43
2.50
4.31
0.200404
OCE
21
0.56
1.39
11.48
0.507280
21
0.75
2.24
16.85
0.207496
Region
= Far East
Product
Demand Equations
CIF
Price Linkage Equations
gOBS
gRSQ
gDW
gFST
UTHEIL
gOBS
gRSQ
gDW
gFST
UTHEIL
US
21
0.93
2.25
124.47
0.073595
21
0.93
2.51
70.94
0.054251
CAN
9
0.40
1.01
1.97
0.755160
21
0.83
2.13
27.28
0.113137
LA
20
0.46
2.34
7.21
0.599169
21
0.79
2.11
21.80
0.175801
MED-EC
21
0.45
1.71
7.28
0.198234
21
0.94
2.81
95.22
0.041532
EC
20
0.32
1.12
4.09
0.693573
21
0.60
2.28
8.38
0.121249
RWE
13
0.34
0.92
2.58
0.284930
21
0.81
3.49
23.88
0.136008
ME/NA
21
0.46
1.62
7.74
0.145949
21
0.87
2.35
38.49
0.132410
AFR
21
0.15
0.69
1.60
0.397976
21
0.96
2.58
139.78
0.044770
OCE
21
0.16
0.76
1.67
0.226227
21
0.96
2.16
123.62
0.036082
COttffi
9
0.42
2.56
2.16
0.346500
21
0.88
2.19
42.24
0.079807


METRIC TONS (Thousands)
YEAR
ACTUAL t FITTED
Figure 5.45. Far East Imports of Fresh Oranges from Oceania (Product Demand 910).
185


37
Table 2.
4 Trade Flow Analysis for Selected Years (1966, 1976 and 1986)
Without Intraregional Trade "Relative Partner Region Exports
by Region"
Year
Region US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
COMMB
- Percentages -
66
US
0.0
0.0
22.5
0.0
0.5
0.0
0.4
0.0
0.9
0.0
0.0
76
0.0
19.0
30.2
0.0
0.1
0.0
0.0
0.0
1.5
1.0
0.0
86
0.0
7.1
9.8
0.6
0.1
0.0
0.9
0.0
3.5
0.0
0.0
66
CAN
54.3
0.0
1.5
0.0
0.7
0.0
0.7
6.1
56.9
0.4
0.0
76
41.1
0.0
0.2
0.0
0.0
0.0
0.2
2.3
35.8
18.5
0.0
86
30.0
0.0
2.3
0.7
0.3
0.0
2.0
0.0
25.7
5.9
0.0
66
LA
2.5
0.0
0.0
0.0
2.8
0.0
0.0
0.0
0.0
0.0
0.0
76
0.4
77.6
0.0
0.0
0.6
0.2
0.0
0.0
0.0
0.0
0.0
86
0.1
1.1
0.0
0.0
1.0
0.0
0.0
0.0
0.0
0.0
3.6
66
MED-EC
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
76
0.0
0.0
0.0
0.0
2.6
3.3
0.0
0.0
0.0
0.0
0.0
86
0.0
0.0
0.4
0.0
15.0
14.2
0.0
1.6
0.0
0.0
0.1
66
EC
21.7
87.3
67.2
73.7
0.0
98.1
75.3
81.4
0.1
24.6
11.2
76
22.5
3.4
38.1
78.3
0.0
95.1
53.9
70.5
0.1
23.9
97.3
86
2.3
86.6
77.6
78.3
0.0
57.4
63.7
85.4
0.1
5.0
94.1
66
RWE
3.8
0.0
2.9
15.2
51.5
0.0
10.1
6.5
1.7
3.1
88.8
76
1.0
0.0
2.8
9.2
34.5
0.0
15.9
7.4
0.0
21.1
2.6
86
0.4
0.0
2.7
11.3
74.2
0.0
17.1
8.7
0.1
0.8
2.1
66
ME/NA
0.0
0.0
0.0
0.1
0.6
0.0
0.0
0.0
40.4
4.2
0.0
76
0.0
0.0
0.1
1.5
52.6
1.4
0.0
19.4
62.6
0.0
0.0
86
0.0
0.0
1.4
0.6
0.5
0.0
0.0
0.0
70.5
15.4
0.0
66
RAF
0.0
0.0
0.0
0.0
36.8
0.0
0.6
0.0
0.0
2.2
0.0
76
0.0
0.0
0.2
0.0
4.9
0.0
0.3
0.0
0.1
0.4
0.0
86
0.0
0.0
0.0
0.0
3.6
0.0
0.4
0.0
0.0
1.5
0.0
66
FE
14.9
12.7
0.6
0.0
0.3
1.7
1.1
5.9
0.0
65.4
0.0
76
31.5
0.0
0.0
0.0
0.1
0.0
0.7
0.4
0.0
35.0
0.1
86
64.9
5.2
0.1
0.1
0.3
0.0
0.5
4.3
0.0
71.4
0.0
66
OCE
0.9
0.0
1.4
0.0
0.0
0.0
0.1
0.1
0.0
0.0
0.0
76
1.4
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
86
2.4
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
66
COMMB
1.9
0.0
4.0
11.0
6.8
0.2
11.6
0.0
0.0
0.0
0.0
76
2.1
0.0
28.4
11.0
4.6
0.0
29.0
0.0
0.0
0.0
0.0
86
0.0
0.0
5.6
8.5
4.9
28.3
15.3
0.0
0.0
0.1
0.0
66
TOTAL
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
76
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
86
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0


APPENDIX I
SENSITIVITY ANALYSIS PROGRAM
The Program #1 below was used to obtain the indices for the market
demand and export supply equations in the sensitivity analysis. Program
#2 was used to estimate the indices for the United States product demand
equations. Since all regional sensitivity analysis programs are similar
to the one presented for the United States, they will not be included
here. The only differences among the regional programs are the variables
and parameters used.
Program 41
FREQ NONE;
SMPL 1,20;
READ (F0RMAT=L0TUS,FILE**'E:\BRENES\NEWDATA.WK1');
READ (F0RMAT=L0TUS FILE**' E: \BRENES\PARAM. WK1') ;
REPD1 EPD1/CPI1; RMP1D MP1D/CPI1;
REPD2 EPD2/CPI2; RMP2D = MP2D/CPI2;
REPD3 = EPD3/CPI3; RMP3D = MP3D/CPI3;
REPD4 = EPD4/CPI4; RMP4D = MP4D/CPI4;
REPD5 EPD5/CPI5; RMP5D ** MP5D/CPI5;
REPD6 EPD6/CPI6; RMP6D = MP6D/CPI6;
REPD7 = EPD7/CPI7; RMP7D = MP7D/CPI7;
REPD8 = EPD8/CPI8; RMP8D = MP8D/CPI8;
REPD9 = EPD9/CPI9; RMP9D = MP9D/CPI9;
REPD10 = EPD10/CPI10; RMP10D MP10D/CPI10;
REPD11 EPD11/CPI11; RMP11D MP11D/CPI11;
FRML EQiy/19
FRML EQ2#19
FRML EQ3#19
FRML EQ4#19
FRML EQ5#19
FRML EQ6#19
EXPORT1
EXPORT2
EXPORT3
EXP0RT4
EXPORT5
EXPORT6
EXP(DH01
EXP(DH02
EXP(DH03
EXP(DH04
EXP(DH05
EXP(DH06
+ DH11*L0G(REPD1)
+ DH12*LOG(REPD2)
+ DH13*LOG(REPD3)
+ DH14*L0G(REPD4)
+ DH15*L0G(REPD5)
+ DH16*LOG(REPD6)
+ DH21*L0G(PRD1))
+ DH22*LOG(PRD2))
+ DH23*LOG(PRD3))
+ DH24*L0G(PRD4))
+ DH25*LOG(PRD5))
+ DH26*LOG(PRD6))
326


Table 5.23 Oceania CIF Price Linkage Equations
Partner
Region
Intercept
( + \-)
FOB
Price
(+)
Year
Trend
<-\ + >
Energy
Index
Price
< + )
gOBS
gRSQ
gDW
gFST
21 Years
Z of
Total
UTHEIL Imports
Average
Z Total
Market
Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
5.489
5.354
-1.025
0.630
0.176
3.572
1.268
1.218
1.042
-0.011
0.056
-0.206
21
0.89
2.17
48.21
0.060483
85.922
3.342
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-3.085
7.077
-0.436
0.290
0.324
0.896
0.478
1.644
0.291
0.169
0.110
1.540
21
0.68
1.68
12.30
0.101043
4.504
0.175
MED-EC
PARAM.VALUE
STD.ERROR
t STATISTIC
-7.934
11.732
-0.676
0.131
0.442
0.296
1.526
2.616
0.583
0.257
0.149
1.732
21
0.80
2.43
22.42
0.106035
1.868
0.073
EC
PARAM.VALUE
STD.ERROR
t STATISTIC
5
0.027
0.001
RWE
PARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
2.771
12.795
0.217
0.616
0.290
2.125
-0.720
3.091
-0.233
0.158
0.262
0.604
21
0.65
1.95
10.50
0.444433
4.878
0.190
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
6
2.626
0.102
FE
PARAM.VALUE
STD.ERROR
t STATISTIC
0.127
14.163
0.009
0.867
0.468
1.854
-0.005
3.320
-0.001
0.075
0.211
0.356
21
0.46
2.10
4.89
0.311537
0.175
0.007
OCE
PARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
Total
100.000
3.890


81
aggregated commodity representing all other fruits, an aggregated
commodity representing all other goods, and an aggregated commodity
representing bananas and apples. The latter alternative was selected
based on the characteristics of consumption for fresh oranges which makes
bananas and apples better substitutes than the other aggregated goods.
Nelson and Robinson (1978) reported that Matthews, Womack, and Huang
(1974) used bananas and apples as substitute products for fresh oranges
with encouraging results, even though they were in some cases less
significant than the ones obtained using other alternatives. In a recent
paper, Lee et al. (1990) also used bananas as a substitute product for
U.S. citrus in Japan. However, in the latter study pineapples instead of
apples were used as the second substitute product.
The general form of the market demand equation for fresh oranges is
the following:1
(4.7) Xi. f(Pi-,INC,POP,PRS)
where
f represents some functional relationship between XA and the
variables on the right hand side,
Xt is the total market demand for fresh oranges in region i,
P is the real average market price of fresh oranges in region i,
INCi is the real income level in region i,
POPi is the population level in region i,
PRSi is the real average market price for the aggregated commodity
based on bananas and apples or other measure of substitutes in
region i.
Single letter notation represents endogenous variables while three letters depict exogenous
variables. The sign associated with each variable represents the hypothesized behavioral
relationship between the exogenous variables and the dependent variable.


332
SMPL 1,7;
IQ1D=XIQ1D*I;
GENR EQiy/50;GENR EQ1#22;GENR EQ1#24;GENR EQ1#26;
GENR EQiy/30;GENR EQ1#34;GENR EQ1//36;
GENR EQ1#21;GENR EQ1#23;GENR EQ1#25;GENR EQ1#27;
GENR EQ1#31;GENR EQ1#35;GENR EQ1#37;
WRITE (FORMAT-LOTUS,FILE-'C:\LOTUS\SIM1//2.WK1' )
I IQ1D
IQ1_2 IQ1_3 IQ1_4 IQ1_5 IQ1_6 IQ1_7 IQ1_8 IQ1_9
IQ1_10 IQ1_11
END;


210
negative elasticities. The magnitudes for the rest of Africa and the
United States indicate an inelastic relative price elasticity for the rest
of Africa and an elastic one for the United States in the EC.
The other three regions have positive elasticities but only the one
from the Middle East/North Africa is significant. EC's major partner has
been Mediterranean-EC. This partnership has been growing fast and trade
has been shifting from the Middle East/North Africa to the Mediterranean-
EC through the years. It is possible to argue that EC imports from the
Middle East/North Africa are marginal in the sense that they are needed
only to complement fruit purchases from Mediterranean-EC. This suggests
that the fruit is imported when prices are going up due to the lack of
sufficient fruit in the market. This conclusion could partially explain
the positive sign, but is truly a conjecture not based on actual data.
Total market demand elasticities for three of EC's five major
partners are elastic and significant. United States product elasticity is
-3.94; Latin America is 2.63; and Mediterranean-EC is 1.85. Middle
East/North Africa and rest of Africa have insignificant parameters. The
magnitudes of significant elasticities indicate that demand for the
product of the United States, Latin America, and Mediterranean-EC in the
EC is very sensitive to changes in the size of the market. However, as in
the Canadian case, the direction and magnitude of change are different for
each partner region. For example, if EC's market size grows, consumers
will shift from United States product to the Mediterranean-EC or Latin
America products.
Rest of Western Europe. Rest of Western Europe imports represented
10.6% of total world imports in the period studied. Major partners are


Table 5.20 Middle East/North Africa CIF Price Linkage Equations
Partner
Region
Intercept
c+\->
FOB
Price
( + )
Year
Trend
(*\+)
Energy-
Index
Price
< + )
@OBS
@RSQ
@DW
@FST
21 Years
Z of
Total
UTHEIL Imports
Average
Z Total
Market
Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
0.181
27.620
0.007
0.983
0.488
2.014
0.033
6.534
0.005
-0.001
0.412
-0.003
21
0.21
2.44
1.53
0.789752
1.493
0.033
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM. VALUE
STD. ERROR
t STATISTIC
-2.352
10.625
-0.221
0.739
0.366
2.019
0.507
2.334
0.217
0.048
0.112
0.426
21
0.88
2.34
42.08
0.101175
14.926
0.334
MED-EC
PARAM.VALUE
STD.ERROR
t STATISTIC
12.745
13.267
0.961
0.789
0.408
1.935
-3.046
3.130
-0.973
0.251
0.256
0.980
21
0.64
2.73
9.89
0.461457
8.371
0.187
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-10.257
13.997
-0.733
0.872
0.300
2.905
2.450
3.240
0.756
-0.088
0.160
-0.551
21
0.71
2.40
13.99
0.198066
1.122
0.025
RWE
PARAM.VALUE
STD.ERROR
t STATISTIC
15.652
21.853
0.716
1.167
0.483
2.414
-3.583
5.008
-0.716
0.205
0.217
0.943
21
0.61
2.84
8.85
0.308409
1.849
0.041
RAF
PARAM.VALUE
STD.ERROR
t STATISTIC
-4.077
4.985
-0.818
0.433
0.177
2.450
0.746
1.155
0.646
0.157
0.081
1.940
21
0.90
2.29
52.72
0.070873
28.481
0.637
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
10.152
15.306
0.663
1.243
0.365
3.407
-2.223
3.667
-0.606
0.164
0.244
0.675
21
0.50
3.11
5.69
0.289771
39.474
0.883
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
15.765
12.024
1.311
1.004
0.467
2.147
-3.679
2.807
-1.311
0.259
0.184
1.406
21
0.52
2.34
6.23
0.222937
4.230
0.095
cot*
PARAM. VALUE
STD.ERROR
t STATISTIC
9.986
22.097
0.452
1.407
0.599
2.347
-2.049
4.981
-0.411
-0.073
0.200
-0.365
21
0.71
2.05
13.87
0.264828
0.054
0.001
100.000 2.237
Total


Figure
METRIC TONS (Thousands)
YEAR
ACTUAL + FITTED
5.48. Communist Bloc Imports of Fresh Oranges from the Mediterranean-EC (Product Demand 114).
188


220
results have similar interpretations and differ very little from the
expected signs and values.
Table 5.29 presents a total of 82 estimated FOB export price
elasticities. All of them are positive as expected. Seventy one are
significant, 30 of them elastic and the rest inelastic. Most FOB export
price elasticities are close to unity. This indicates that, for a given
change in the FOB export price in the exporting region, a similar change
will occur in the CIF import price in the importing region. This result
was expected.
Table 5.30 shows the results for the year trend variable. Signs are
mixed, as expected, and 31 of the 82 parameters are significant. All
regions have at least one CIF price linkage equation with a significant
parameter. It is interesting to notice that only elastic (greater than
one in absolute value) positive or negative elasticities are significant.
This implies that, for those CIF price linkage equations with significant
parameters, the CIF prices are changing faster than the year trend.
Industry and transportation-system structural change seems to be
unimportant for those trading regions with insignificant parameters. This
implies either that there had not been a structural change or that the
change had been negligible.
Table 5.31 presents the elasticities for the index price of energy.
The values obtained indicate that 49 have the correct positive sign and 33
have negative signs. However, only six of the negative elasticities are
significantly different from zero. The six significant elasticities are
spread among a few regions. They are rest of Western Europe product
demanded in Mediterranean-EC; Middle East/North Africa product demanded in


292
each section indicated that domestic and trade-policy decisions can be
enhanced using the information generated in the sensitivity analysis.
Total market demand analysis concluded that if world prices
increase, major importers will consume proportionally less than regions
with low import levels and high local production. If world prices
decrease, importers will consume relatively more than regions with low
import levels. The analysis also shows that the largest response
coincides with the United States, implying that consumers in this market
are highly sensitive to changes in the average market price. If world
prices increase, the percentage adjustment in demand by U.S. consumers
will be greater than the percentage downward adjustment seen in the other
regions. If world prices decrease, United States consumers will consume
more relative to the rest of the regions considered. If a region is
interested in increasing exports to another region, the information
provided in this section of the chapter could be used for price policy
decisions and price discrimination.
Export supply equations show weak FOB average export price
parameters versus very strong and significant fresh production parameters.
This indicates that major export decisions are driven mainly by fresh
production (see Sparks results for vegetables). Since fresh orange
production implies a long-term decision, the results are appealing.
Middle East/North Africa exports have the correct relationship. This is
probably related to the fact that its industry is completely open, while
the Mediterranean-EC has important trade agreements with major partners,
and the United States with Canada. The results indicate that the Middle
East/North Africa is the region with the most flexible relationship


Table 1.4 World Fresh Orange Exports by Region
Region
1966
1976
1978
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
Annual
Growth
Rate
1978-86
(000) Metric Tons
Percent of Change
United States
258.6
464.1
355.9
413.0
2.4
-1.2
1.9
Canada
0.1
0.1
0
0.3
6.9
16.1
31.5
Latin America
106.6
119.2
164.7
229.3
3.9
6.8
4.2
Mediterranean-EC
1514.6
1937.5
1802.2
2833.9
3.2
3.9
5.8
E.C.
32.8
122.6
121.5
195.0
9.3
4.8
6.1
Rest of Western Europe
2.1
5.9
7.8
2.2
0.2
-9.5
-14.7
Middle East/North Africa
1215.9
1871.9
1893.2
1316.7
0.4
-3.5
-4.4
Rest of Africa
264.4
285.7
274.4
209.0
-1.2
-3.1
-3.3
Far East
65.8
142.3
137.5
113.9
2.8
-2.2
-2.3
Oceania
23.5
11.2
30.1
47.3
3.5
15.5
5.8
Communist Bloc
0.2
60.4
140.9
35.0
30.5
-5.3
-16.0
World Total
3484.5
5020.7
4928.3
5395.6
2.2
0.7
1.1
Source: United Nations Trade Data Tapes.
o


291
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.28. Communist Bloc Imports Changing Total Market Demand.


39
Tables 2.6, 2.7, and 2.8 correspond to the same characteristics
described for Tables 2.3, 2.4, and 2.5, respectively. The difference
between these two sets of tables is that the former tables present
cumulative information for five years instead of yearly information. The
periods considered are 1966 to 1970, 1974 to 1978, and 1982 to 1986. The
discussion that follows will be based mainly on the first set of tables,
because both sets draw similar conclusions. However, given that yearly
information could be biased for uncommon reasons, the results in Table 2.6
to 2.8 are useful to support general conclusions.
Partner Region Perspective
In this section, the discussion will be oriented from the exporters'
viewpoint. In all cases, the relative importance of each region is set
forth and then a trade flow analysis is developed.
Table 2.3 shows that the world's major fresh orange exporter was the
Mediterranean-EC region. With intraregional trade considered, this
region's share of total exports was 43.5% in 1966, 38.6% in 1976, and
52.5% in 1986. With intraregional trade not considered, the relative
importance of the region in world trade increased to 44.7%, 45.3%, and 58%
respectively. These values show the importance of this region in world
trade of fresh oranges.
Table 2.4 shows that the major partner of the Mediterranean-EC was
the EC region. In 1966, 73.7% of the Mediterranean-EC region's total
exports went to the EC region. This percentage increased to 78.3 in 1976
and was the same in 1986. The EC region includes all EC countries except


269
Latin America. Since rest of Western Europe is not a producer and EC
production is small, then it is apparent that consumers' first choice will
be Latin America and then the Middle East/North Africa. Given that Latin
America product demand is not affected by import price changes, its
position in the market is even stronger. As explained in Chapter 5,
United Nations trade data tapes included exports from rest of Western
Europe which are probably related to reexports.
EC
The EC is the world largest importer of fresh oranges. Table 6.2
indicates that five regions accounted for 99.6% of total imports during
the period considered. The regions are the Mediterranean-EC, Middle
East/North Africa, rest of Africa, Latin America, and the United States.
As shown in Figure 6.15, two regions have the correct negative association
between import price and product demand indices. The product demands for
the other three regions have positive relationships, but two of them are
not significant (see Table 5.7). Demand relationships for rest of Africa
and United States products are negative and significant. However, the
United States parameter is more elastic. As the import price index
increases, consumers in the EC will shift their consumption from United
States product to rest of Africa product. The demand for Middle
East/North Africa product is positive and significant. This result might
be related to the fact that the EC will buy product from the Middle
East/North Africa only when market prices are increasing, due to


243
this case, the response index goes from below one up to approximately 2.2.
In other words, it ranges from below the 1986 level up to approximately
2.2 times that level. The figure indicates that major world consumers
have different responses to changes in the average market price. Since
the response index depends on the magnitude and sign of the original
parameter, a negative relationship is expected. As the average market
price index increases, the total market demand index should decrease and
vice versa.
The figure shows that responses are negative, with the exception of
the Far East, and the magnitude of the responses differs from region to
region. For example, a 10% increase in the average market price is
represented on the bottom axis as 1.1. The values corresponding to the
total market demand index, starting from the lowest, are .812 for United
States; .893 for EC; .897 for Middle East/North Africa; .928 for
Mediterranean-EC; .995 for Latin America; and 1.017 for the Far East (see
Appendix J). The index numbers are less than one, except for the Far
East. This result indicates that total market demands will be lower than
the 1986 level in all regions after a 10% increase in the average market
price. The only exception is the Far East. Recall from Chapter 5 that
unique problems with the Far East equations were discussed.
The rest of the analysis will be based on the graphical results.
The specific index numbers utilized to construct the figures are shown in
Appendix J. The relative importance for each region is captured by the
order in which the figures present the different regions. For example,
Latin America is the largest consumer of fresh oranges in the world,
therefore, Figure 6.1 lists Latin America in the first position.


329
? SIMULATION if3 EXPORT SUPPLY VARYING FOB EXPORT PRICE;
SMPL 1,1;
I-.5;
SMPL 2,20; I-I(-l)+.l;
SMPL 1,20;
ID-1;
REPD1XREPD1*I;REPD2=XREPD2*I;REPD3-XREPD3*I;REPD4-XREPD4*I;
REPD5-XREPD5*I;REPD6-XREPD6*I;REPD7-XREPD7*I;REPD8=XREPD8*I;
REPD9-XREPD9*I;REPD10=XREPD10*I;REPD11=XREPD11*I;
GENR EQ1#19;GENR EQ2#19;GENR EQ3#19;GENR EQ4#19;GENR EQ5#19;
GENR EQ6#19;GENR EQ7#19;GENR EQ8#19;GENR EQ9//19;GENR EQ10//19;
GENR EQliy/19;
WRITE (FORMAT-LOTUS,FILE-'C:\LOTUS\SIMy/3.WK1')
ID I REPD1 REPD2 REPD3 REPD4 REPD5 REPD6 REPD7 REPD8 REPD9
REPD10 REPD11
EXPORT1 EXPORT2 EXPORT3 EXP0RT4 EXPORT5 EXPORT
EXPORT7 EXPORT8 EXP0RT9 EXPORTIO EXP0RT11;
REPD1-XREPD1;REPD2-XREPD2;REPD3-XREPD3;REPD4-XREPD4;REPD5-XREPD5;
REPD6-XREPD6;REPD7-XREPD7;REPD8-XREPD8;REPD9-XREPD9;REPD10-XREPD10;
REPD11-XREPD11;
? SIMULATION #4 EXPORT SUPPLY VARYING FRESH PRODUCTION;
SMPL 1,1;
I-.5;
SMPL 2,20; I-I(-l)+.1;
SMPL 1,20;
ID-1;
PRD1XPRD1*I;PRD2=XPRD2*I;PRD3-XPRD3*I;PRD4-XPRD4*I;PRD5-XPRD5*I;
PRD6=XPRD6*I;PRD7=XPRD7*I;PRD8=XPRD8*I;PRD9-XPRD9*I;PRD10=XPRD10*I;
PRD11-XPRD11*I;
GENR EQiy/19; GENR EQ2#19;GENR EQ3#19;GENR EQ4#19;GENR EQ5#19;
GENR EQ6#19;GENR EQ7#19;GENR EQ8#19;GENR EQ9#19;GENR EQ10#19;
GENR EQliy/19;
WRITE (FORMAT-LOTUS,FILE-'C:\LOTUS\SIM#4.WK1')
ID I PRD1 PRD2 PRD3 PRD4 PRD5 PRD6 PRD7 PRD8 PRD9
PRD10 PRD11
EXPORT1 EXP0RT2 EXPORT3 EXP0RT4 EXPORT5 EXPORT
EXPORT7 EXPORT8 EXP0RT9 EXPORTIO EXP0RT11;
PRD1-XPRD1;PRD2-XPRD2;PRD3-XPRD3;PRD4-XPRD4;PRD5-XPRD5;
PRD6-XPRD6;PRD7-XPRD7;PRD8-XPRD8;PRD9-XPRD9;PRD10-XPRD10;
PRD11XPRD11;
END;


42
Table 2.7 Trade Flow Analysis for Selected Periods of Five Years, (1966-
70, 1974-78 and 1982-86) Without Intraregional Trade "Relative
Partner Region Exports by Region"
Period Region US CAN LA MED-EC EC RWE ME/NA RAF FE OCE COMMB
Percentages
PI
US
0.
.0
16
.0
37
.2
0.
.0
0
.1
0
.0
0
.5
0
.0
2
. 1
0
.1
0
.0
PE"
0.
.0
3
.5
27
.6
0,
.0
0
.1
0.
.0
0
.2
0
.0
1
. 4
0
.2
0
.0
P3*
0.
.0
27
.5
16
.5
0,
.2
0.
.2
0.
.0
0
.6
0
.0
2
.5
0
.5
0
.0
PI
CAN
55.
.5
0.
.0
2
.6
0.
.0
0
.8
0.
.0
0
.5
5
.3
73
.3
2
.0
0
.0
P2
42.
.6
0.
.0
0.
.4
0,
.0
0
.1
0.
.0
0
.4
2
.7
29
.8
13.
.1
0
.0
P3
34.
. 1
0.
.0
2.
.1
0.
.4
0
.1
0.
.0
1
.7
0
.0
30
.6
8
.0
0
.0
PI
LA
1.
.3
34
.5
0
.0
0,
.0
3
.5
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
P2
0.
.4
3
.2
0
.0
0,
.0
0
.5
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
P3
0.
.4
1
.2
0
.0
0,
,0
1
.6
0
.0
0
.0
0
.0
0
.0
0
.0
2
.0
PI
MED-EC
0
.0
0
.0
0
.0
0,
.0
0
.3
0
. 1
0
.0
0
.0
0
.0
0
.0
0
.0
P2
0.
.0
0
.0
0
.0
0,
,0
3
.0
0.
.5
0
.0
0
.0
0
.0
0
.0
0
.0
P3
0.
,0
0
.0
0.
.8
0,
,0
17
. 1
6.
. 1
0
.0
0
.6
0
.0
0.
.0
0
.3
PI
EC
20.
.7
46.
.2
48.
.3
74.
.4
0
.0
72.
.8
70
. 1
82
.9
0
.5
17.
.8
44
.6
P2
19.
.0
4.
.3
45.
.4
77,
,8
0
.0
23.
.9
51
.0
77
.2
0
.1
17
,4
92
.5
P3
2.
.8
52.
.0
64.
.6
79.
.9
0.
.0
59.
.6
59
. 7
76
.3
0.
.2
3.
.6
96
.5
PI
RWE
2.
.4
0.
.0
4,
.1
14.
,0
60
.5
0,
.0
11.
.4
8
.0
0.
.6
1.
.8
53
.8
P2
1.
.3
0.
.0
3.
.0
9.
.2
55.
.4
0.
.0
15.
.2
8
.6
0.
.4
11.
,7
6
.9
P3
0,
.4
1
.5
1.
.9
10.
.0
70
.8
0.
.0
16
.6
10.
.4
0.
.5
1.
.1
1
.2
PI
ME/NA
0.
.0
0.
.0
0.
.0
0.
,0
1.
.6
25.
.5
0
.0
0,
.4
22.
.6
1.
.8
0
.0
P2
0.
,7
0.
.0
3.
.5
0.
.6
28
.2
73,
.0
0
.0
10.
.5
68.
. 1
17.
.3
0
.5
P3
0.
.1
0.
.0
11.
.7
0.
.5
0.
.8
19.
.2
0
.0
8.
.7
66.
.2
15.
.2
0
.0
PI
RAF
0.
0
0.
.0
0.
.4
0.
.0
24,
,0
0,
,0
0
. 4
0,
.0
0.
.0
2.
, 1
1
.6
P2
0.
.0
0.
.0
0.
,3
0.
.0
8.
.4
0.
,3
0,
.3
0,
.0
0.
.0
0.
2
0
.0
P3
0.
0
0.
.0
0.
.0
0.
0
5.
.5
0.
.0
0.
.4
0,
,0
0.
.0
2.
.1
0.
,0
PI
FE
18.
.1
3,
.3
1.
,2
0.
0
2,
,0
1.
2
1.
.2
3.
.2
0.
.0
74.
3
0.
.0
P2
31.
9
89.
.0
0.
.1
0.
0
0,
,1
0,
,0
0,
,8
0,
.9
0.
.0
40.
1
0.
,1
P3
59.
3
17.
.8
0.
2
0.
1
0.
,4
0.
,3
0,
,7
3,
.9
0.
.0
69.
.4
0,
,0
PI
OCE
0.
8
0.
.0
1.
.1
0.
0
0,
.0
0.
.0
0.
.1
0.
.2
0.
.0
0.
0
0.
.0
P2
2.
2
0.
.0
0.
.1
0.
0
0.
,0
0.
.0
0.
.0
0.
,1
0.
.0
0.
0
0,
,0
P3
2.
8
0.
.0
0.
.0
0.
0
0.
.0
0.
.0
0,
.0
0.
.1
0.
1
0.
0
0.
.0
PI
COMiB
1.
1
0.
,0
5.
,1
11.
5
7.
.3
0.
.4
15.
.8
0.
.0
0.
9
0.
1
0.
.0
P2
1.
8
0.
0
19.
.5
12.
3
4.
.0
2.
,2
32.
,0
0.
.0
0.
2
0.
0
0.
.0
P3
0.
0
0.
.0
2.
.3
9.
0
3.
.4
14.
.7
20.
,3
0.
.0
0.
0
0.
0
0.
.0
PI
TOTAL
100.
0
100.
,0
100.
.0
100.
0
100.
,0
100.
,0
100.
.0
100.
.0
100.
0
100.
0
100.
.0
P2
100.
0
100.
0
100.
.0
100.
0
100.
.0
100.
,0
100.
.0
100.
.0
100.
0
100.
0
100.
0
P3
100.
0
100.
0
100.
.0
100.
0
100.
,0
100.
,0
100.
,0
100.
.0
100.
0
100.
0
100.
0
Represents period from 1966-70.
Represents period from 1974-78.
Represents period from 1982-86.


276
Middle East/North Africa
As indicated in Table 6.2, 96.9% of total Middle East/North Africa
imports came from the Far East, rest of Africa, Latin America,
Mediterranean-EC, and Oceania. The sensitivity analysis shown in Figure
6.19 presents the Middle East/North Africa product demand index while
changing the import price index. The figure indicates that three of the
five major product demands have negative relationships. Far East and
Oceania parameters are negative and significant (see Table 5.9). Demand
for the rest of Africa product has also a negative association, but it is
not significant. Latin America and Mediterranean-EC have positive but
insignificant relationships. The figure indicates that Far East and
Oceania responses are different to changes in the import price index.
Oceania has a stronger response than the Far East. Demand for Oceania
product in the Middle East/North Africa is more sensitive to changes in
import prices. If Oceania and Far East import prices increase in the same
proportion, consumers in the Middle East will consume more product from
the Far East relative to Oceania product. The reverse happens when import
prices decrease.
Figure 6.20 presents Middle East/North Africa imports while changing
total market demand. The sensitivity analysis indicates that all
responses are positive and strong. The demand for Latin America product
has the strongest response. The second strongest corresponds to the rest
of Africa product. The third, fourth, and fifth places correspond to
Oceania, Mediterranean-EC, and Far East, respectively. The results imply
that, if market size increases, Latin America product becomes the
consumers' first choice. The rest of the regions will also have a


5
with the recent enlargement of the EC to 12 nations with the inclusion of
Spain and Portugal. Both are major producers and exporters of fresh
fruits to the rest of Europe.
Other factors affecting agricultural trade are income, population,
demographic variables within the trading regions, and exchange rates.
Income and population are important to determine the level of consumption.
As these two variables increase, higher levels of consumption and shifts
from one bundle of goods to another are expected. Exchange rates affect
the real terms of trade among countries, especially in cases when they are
managed by governments and are not allowed to fluctuate freely in the
market. Transportation is an important linkage variable in world trade.
The linkage is between the Free On Board (FOB) export price from any
country and the Cost Insurance Freight (CIF) import price at the final
market. Substitute product prices should have a positive effect on the
consumption of a specific commodity.
Total agricultural trade increased at an annual rate of 2.8% while
agricultural production increased by 2.3% a year from 1966 to 1986 (Food
and Agricultural Organization [FAO] Trade and Production Yearbooks). The
United States agricultural sector represents about 15% of total exports.
Hence, United States agricultural market prices are strongly influenced by
supply and demand conditions among major world markets (Statistical
Abstract of the United States and FAO Trade Yearbook). These statistics
reflect the increasing importance of trade in the world economy and, in
particular, the importance of agriculture in world trade.
The focus of the present study will be the fresh orange industry.
World trade in the fresh orange industry must be studied from the


TABLE OF CONTENTS
page
ACKNOWLEDGEMENTS ii
LIST OF TABLES vi
LIST OF FIGURES x
ABSTRACT xv
CHAPTER
1 INTERNATIONAL TRADE AND AGRICULTURAL PRODUCTS:
THE ORANGE INDUSTRY 1
Introduction 1
International Trade of Agricultural Products ... 3
Problem and Objectives 23
Scope 26
Methodology 27
Overview 28
2 FRESH ORANGE WORLD PRODUCTION AND TRADE 29
Introduction 29
Production Analysis 29
Trade Flow Analysis 34
Partner Region Perspective 39
Region Perspective 50
Conclusions 56
3 LITERATURE REVIEW 58
International Agricultural Trade Models 58
Trade Models: The Orange Industry 68
4 WORLD FRESH ORANGE TRADE MODEL 75
Introduction 75
Fresh Orange Trade Model 77
Demand Side 77
Supply Side 83
Export Supply Equations 84
Equilibrium Conditions 85
iii


228
Sensitivity Analysis Procedure
The fresh orange trade model developed is a nonlinear simultaneous
system of equations. If the reduced form of the model can be obtained,
the model can be simulated as a whole for changes in the different
exogenous variables. The reduced form of a simultaneous system of
equations is obtained when all equations are expressed with only exogenous
variables in the right-hand side. This approach implies that, for a given
change in any exogenous variable, it is possible to assess the impact in
all 561 endogenous variables. A change in any exogenous variable produces
changes in all equations. The impact comes first from the exogenous
variable itself, and then from all the endogenous variables that will be
affected through the different equations. Given the size and complexity
of the model, the reduced-form parameters are difficult to obtain; and it
is not assured that they can be found. For the fresh orange trade model
presented in this study, it was not possible to solve for the reduced form
parameters. However, it was possible to perform a comparative static
analysis equation by equation.
Sensitivity analysis can be conducted to investigate the effects of
changes in the different variables of the model. It is possible to assess
the impact on any dependent variable, using the estimated parameters and
introducing changes in selected variables. This approach implies that the
analysis will not take into consideration the rest of the model when a
variable is changed and the impact on a given equation evaluated.
However, the estimation procedure does take into consideration the rest of
the model and its nonlinear and simultaneous characteristics.


205
Table 5.28 Product Demands Total Market Demand Elasticities3
Region ib
REGION jb
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
COMMB
US
1.372c
1.223d
-1.060
-1.490
-3.938
-1.234
-10.850
-2.645
1.551
0.816
-3.544
-1.267
2.666
10.485
2.663
1.782
0.484
0.200
CAN
-5.097
-0.868
0.634
0.230
-3.856
-0.496
3.662
1.182
LA
3.298
1.184
-18.571
-2.496
8.218
1.062
2.625
1.972
1.200
1.091
6.749
1.764
0.185
0.067
3.155
1.564
-4.127
-1.869
3.375
3.606
MED-EC
27.351
2.574
-23.719
-2.269
5.098
2.258
1.846
2.976
1.612
2.078
3.655
2.448
6.938
3.019
2.012
2.415
5.972
1.025
0.280
1.371
EC
0.815
0.171
13.259
1.532
1.490
1.067
11.146
3.349
2.660
2.454
2.576
2.072
1.081
1.667
4.084
2.674
0.925
1.674
RWE
12.805
5.388
-5.828
-2.101
3.596
1.466
-2.612
-2.198
4.159
2.691
ME/NA
11.692
1.613
-4.531
-1.215
-2.791
-0.646
2.365
0.788
0.144
0.247
0.212
0.486
0.251
0.490
-0.026
-0.058
-15.478
-3.424
1.237
4.177
RAF
-5.021
-1.694
-0.042
-0.068
0.782
1.582
5.605
4.036
2.688
0.935
FE
0.578
0.577
3.187
4.017
23.361
1.795
-4.150
-0.417
3.294
3.998
-6.961
-2.426
2.104
0.312
7.847
1.611
OCE
30.060
3.553
14.513
3.746
10.671
2.420
7.456
2.262
4.743
6.678
-1.179
-0.590
1.129
1.690
COMffl
13.613
2.228
12.889
2.191
-1.672
-0.640
8.263
1.537
6.273
1.781
aProduct demand equals Xj and total market demand equals X.
^Region i across the top is the region importing from region j down the column.
cThe first line in each region represents parameter values.
dThe second line in each region represents the t Statistic.


45
two regions, the rest of Western Europe and the Communist Bloc. Exports
from the Middle East/North Africa region to the rest of Western Europe
represented 10.1% in 1966, 15.9% in 1976, and 17.1% in 1986. Exports to
the Communist Bloc region represented 11.6%, 29.0% and 15.3% in the same
years. Exports from the Middle East/North Africa region to the United
States and Canada decreased from 1966 to 1976; however, exports to these
countries have been increasing in recent years.
United States exports increased at a rate of 2.25% a year from 1966
to 1986. In relative terms, United States participation in world trade of
fresh oranges showed about the same level as 1966. Total United States
trade represented 7.4% of total world trade in 1966, increased to 9.2% in
1976 and decreased to 7.7% in 1986 (see Table 2.3). With intraregional
trade not considered, the relative importance of the United States trade
in the world trade increased. Table 2.3 shows that United States trade
represented 7.6% in 1966, 10.8% in 1976, and 8.5% in 1986. In relative
terms, these percentages show the United States to have been the third
largest exporter, exceeded by the Mediterranean-EC and the Middle
East/North Africa regions. In absolute terms, the Mediterranean-EC and
the Middle East/North Africa exports were 6.9 and 2.6 times the United
States exports, respectively, in 1986.
The relative importance of the United States partners has been
changing through the years. The major United States partner in 1966 was
Canada. Exports to Canada accounted for 54.3% of United States fresh
exports that year (see Table 2.4). The second largest partner was the EC
with 21.7% and the third largest was the Far East with 14.9%. Latin
America, rest of Western Europe, Oceania, and the Communist Bloc absorbed


115
Table 5.8 Rest of Western Europe Product Demands
Partner
Region
Intercept
(+\->
Relative
Price
(-)
Total
Market
Demand
(*\ + )
@OBS
@RSQ
@DW
@FST
UTHEIL
21 Year Average
% Z Total
Total Market
Imports Demand
us
FARAM. VALUE
STD.ERROR
t STATISTIC
151.270
53.762
2.814
-8.876
3.255
-2.727
-10.850
4.102
-2.645
21
0.33
2.50
4.44
0.248259
0.941
0.941
CAN
FARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM.VALUE
STD.ERROR
t STATISTIC
-7.246
14.239
-0.509
-0.653
0.513
-1.274
1.200
1.100
1.091
21
0.13
1.29
1.36
0.143369
1.022
1.022
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-8.745
10.036
-0.871
-2.700
1.061
-2.544
1.612
0.776
2.078
21
0.28
1.24
3.41
0.086744
43.412
43.412
EC
FARAM. VALUE
STD.ERROR
t STATISTIC
-24.543
14.097
-1.741
-2.654
0.333
-7.974
2.660
1.084
2.454
21
0.85
2.05
50.81
0.160084
1.501
1.501
ME/NA
PARAM.VALUE
STD.ERROR
t STATISTIC
9.311
5.635
1.652
-2.666
0.504
-5.289
0.212
0.437
0.486
21
0.73
1.35
23.85
0.047351
47.547
47.547
RAF
PARAM.VALUE
STD. ERROR
t STATISTIC
-0.171
6.415
-0.027
0.622
0.513
1.212
0.782
0.494
1.582
21
0.15
1.22
1.55
0.062721
5.327
5.327
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
58.197
129.350
0.450
-0.754
1.827
-0.413
-4.150
9.953
-0.417
17
0.01
2.32
0.10
0.740192
0.039
0.039
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
-91.473
42.710
-2.142
1.792
1.181
1.518
7.456
3.296
2.262
21
0.28
0.89
3.51
0.330177
0.163
0.163
COMMB
PARAM.VALUE
STD.ERROR
t STATISTIC
27.034
33.816
0.799
5.388
1.662
3.243
-1.672
2.613
-0.640
18
0.42
1.98
5.50
0.280677
0.048
0.048
Total 100.000100.000


67
rejected using the AIDS model. They also recognized the problems with the
AIDS model.
The restrictiveness of the assumptions were recognized earlier by
Resnick and Truman (1973). They relaxed several assumptions of
Armington's model, especially the one that the elasticities of
substitution need to be constant and identical between all pairs of
suppliers to each market. They specified a multi-stage decision process
instead of Armington's two-stage procedure. Again, total imports were
determined first and then imports from a sequence of successively smaller
geographic regions were determined.
Artus and Rhomberg (1973) also recognized the problem with the
assumptions and replaced the CES index function. They used the constant
ratios of elasticities of substitution and homogeneous (CRESH) index
functions developed by Mukerji (1963) and Hanoch (1971).
Sparks (1987), following Artus and Rhomberg's work, used the
constant ratio of elasticity of substitution (CRES) index which makes the
model somewhat less restrictive. This assumption implies that the
elasticity of substitution for all the products in a market or region i
vary by a constant proportion, but the substitutability between products
need not be the same. This assumption increases the flexibility of the
model but also increases the computational complexity. The model was
applied to a highly aggregated commodity (vegetables). In this case, the
basic assumption of Armington's model, goods distinguished by place of
production, seems less applicable given that the aggregated commodity will
be composed of several goods. The model explained that trade flows can


286
associations between import price and product demand indices presented in
the sensitivity analysis are negative and significant for United States
and Middle East/North Africa products (see Figure 6.25). Latin America
has an incorrect positive and significant relationship (see Table 5.12).
The direction of this association could be the result of a relatively
small trade between the regions. The strongest negative relationship
corresponds to the Middle East/North Africa. That is, if the United
States and Middle East/North Africa import prices increase proportionally,
consumers will consume relatively more product from the United States. If
import prices decrease, consumers will tend to consume relatively more
from the Middle East/North Africa.
Figure 6.26 and the empirical results indicate that the United
States has an important advantage in Oceania. It is the only region with
a positive and significant relationship between total market demand and
product demand indices. The other regions have significant and highly
negative relationships. If Oceania market size increases, consumers will
import most product from the United States.
Communist Bloc
As indicated by Table 6.2, three regions accounted for 99.4% of the
Communist Bloc imports in the period considered. The regions are the
Middle East/North Africa, Mediterranean-EC, and Latin America. Figure
6.27 presents Communist Bloc imports while changing import prices. The
figure shows that two of the three regions have the correct negative
relationship. The Mediterranean-EC has a positive but insignificant
relationship (see Table 5.13). That is, import prices from this region


260
PRODUCT DEMAND INDEX (1986=1)
IMPORT PRICE INDEX (1986 = 1)
Figure 6.9. Canada Imports Changing Import Prices.


114
Table 5.7 EC Product Demands
Partner
Region
Intercept
( + \-)
Relative
Pries
(-)
Total
Market
Demand
(-\ + )
@OBS
@RSQ
@DW
@FST
UTHEIL
21 Year Average
Z Z Total
Total Market
Imports Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
70.236
47.110
1.491
-4.896
1.594
-3.072
-3.938
3.191
-1.234
21
0.36
2.03
5.12
0.279023
1.821
1.812
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
60.184
114.670
0.525
-1.616
2.825
-0.572
-3.856
7.778
-0.496
14
0.04
1.34
0.25
0.726798
0.002
0.002
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-27.456
19.627
-1.399
0.199
0.699
0.285
2.625
1.331
1.972
21
0.19
0.74
2.16
0.189179
3.278
3.265
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-13.094
9.152
-1.431
0.755
1.191
0.634
1.846
0.620
2.976
21
0.42
1.45
6.61
0.075885
54.192
53.955
RWE
PARAM. VALUE
STD. ERROR
t STATISTIC
93.332
41.206
2.265
-2.261
1.084
-2.085
-5.828
2.774
-2.101
21
0.23
1.44
2.66
0.251156
0.030
0.029
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
11.630
8.549
1.361
1.868
0.783
2.387
0.144
0.582
0.247
21
0.27
0.56
3.30
0.065822
32.690
32.548
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
12.939
9.060
1.428
-0.712
0.537
-1.326
-0.042
0.616
-0.068
21
0.10
0.57
0.95
0.070983
7.691
7.656
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
-342.120
193.120
-1.772
4.105
2.517
1.631
23.361
13.018
1.795
21
0.15
1.60
1.61
0.626029
0.007
0.007
OCE
PARAM.VALUE
STD.ERROR
t STATISTIC
-151.980
65.557
-2.318
5.085
2.766
1.838
10.671
4.410
2.420
21
0.27
2.48
3.28
0.286791
0.108
0.106
COMMB
PARAM. VALUE
STD.ERROR
t STATISTIC
-183.660
86.698
-2.118
-9.158
2.519
-3.636
12.889
5.882
2.191
20
0.57
0.71
11.08
0.44278
0.181
0.180
Total 100.000 99.560


53
last two decades. As a whole, they represented 90.5% in 1966, 86.9% in
1976, and 87.3% in 1986 of total world imports (see Table 2.3).
The Far East has been consistently growing as an importing region in
the last two decades. It passed from 3.4% in 1966 to 4.8% in 1976, and to
7.1% in 1986 (see Table 2.3). With intraregional trade omitted, these
percentages decreased to 2.3, 3.8 and 6.4, respectively. This shows that
trade among countries belonging to the Far East region was important
relative to the rest of the world's trade with the same region.
The United States was the leading exporter to the Far East region in
the period considered. Its exports represented 49.5% in 1966, 90.6% in
1976, and 86.2% in 1986 (see Table 2.5). The second exporter to the Far
East was Oceania, with 11.5%, 1.9%, and 8.4%, respectively. The third
exporter was the rest of Africa. However, its participation has been
decreasing, from 19.9% in 1966 to 2.9% in 1986. The fourth exporter to
the Far East was the Middle East/North Africa region, whose participation
has also decreased from 17.8% in 1966 to 1.8% in 1986. The Mediterranean-
EC region represented the fifth exporter to the Far East region, with a
consistent participation in the market of only .5%. Latin America's share
of the market decreased from .8% in 1966 to .1% in 1986. It is clear from
these numbers that the United States was the only exporting region whose
market share grew in the Far East.
The Middle East/North Africa region was another significant importer
of fresh oranges. Its participation grew from .6% in 1966 to 11.8% in
1976, but decreased later to 5.2% in 1986 (see Table 2.3). The table
shows that the percentages excluding intraregional trade were .3 in 1966,


77
each region's market share of a commodity is influenced by changes in the
size of that market, even when relative product prices remain unchanged.
In the present study, 11 regions were defined. The regions were
selected consistently with the world orange industry and with particular
similarities among the countries included in a region. The regions were
the United States (US), Canada (CAN), Latin America (LA), Mediterranean-EC
(MED-EC), EC, rest of Western Europe (RWE), Middle East/North Africa
(ME/NA), rest of Africa (RAF), Far East (FE), Oceania (OCE), and Communist
Bloc (COMMB).
In the next section, a complete world fresh orange trade model is
specified. Demand and supply sides are included with equilibrium
conditions and price linkages set forth.
Fresh Orange Trade Model
Demand Side
The model was based on the two assumptions mentioned above. Two
stage budgeting is implied. Marginal rates of substitution between two
goods in a commodity group were assumed to be independent of goods in
other groups. In the orange industry, the rate at which consumers
substitute fresh oranges produced in one country for those produced in
another country does not depend on their purchases of other kinds of
fruits or other commodities. The first level of the two stage budgeting
is the consumers' decision to allocate their total income among the


62
empirical evidence that this may be the case for wheat (Grennes et al.,
1977 and 1978; Thompson, 1981) and other agricultural products. Spatial
price equilibrium models have few capabilities except for the weak and
incomplete explanation of trade flows given the problems mentioned above.
Trade-flow and market-share models are the third type of
multiple-region models considered by Thompson (1981). These models were
developed to account for the observed variation in trade flows more
adequately than do the spatial equilibrium models.
Taplin (1967) and Johnston (1976) in a partial sense surveyed world
trade models concerned primarily with trade flows. They studied the ones
that analyzed the structure of world trade and the short-run trade
fluctuations among countries. In his paper, Taplin classified them in two
categories: the ones that have separate functions for total exports and
total imports but do not attempt to estimate the individual flows between
countries; and the ones that look at individual flows directly.
In the first part, Taplin's discussion went from an import-export
matrix developed by the League's Network of World Trade (1942), passing by
Woolley's (1965) transactions matrices on payments for trade, services,
and capital flows, to Beckerman's (1956) input-output approach. These
studies provided important insight into the structure of the international
economy. However, they did not represent a formal model where hypotheses
could be tested, measured or forecasted.
The second part of Taplin's investigation continued with a survey
covering other studies (Tinbergen, 1962; Linnemann, 1966; Waelbroeck, 1962
and 1965) where individual trade flows (from the import-export matrix)
between countries were considered to be a function of income, population,


251
EXPORT SUPPLY INDEX (1986=1)
FOB AVERAGE EXPORT PRICE INDEX (1986=1)
Figure 6.6. Export Supply Changing Fresh Production (Major World
Exporters).


METRIC TONS (Thousands)
YEAR
ACTUAL FITTED
Figure 5.34. EC Imports of Fresh Oranges from the rest of Africa (Product Demand 58).
174


Table H.l United States, Canada and South America Product Demand and CIF Price Linkage Equations Statistics
Region = United States Region = United States
Product Demand Equations CIF Price Linkage Equations
SOBS
6RSQ
SDW
SFST
UTHEIL
SOBS
6RSQ
SDW
SFST
UTHEIL
CAN
13
0.17
1.42
1.04
0.390149
21
0.68
2.20
11.92
0.191764
LA
21
0.31
1.68
4.06
0.197823
21
0.77
2.13
19.20
0.079479
MED-EC
21
0.28
1.03
3.48
0.702719
21
0.93
2.36
72.38
0.048339
EC
19
0.44
2.17
6.20
0.335904
21
0.68
1.37
11.84
0.134116
ME/NA
21
0.43
1.90
6.77
0.265216
21
0.83
2.34
27.36
0.100821
FE
21
0.54
1.29
10.70
0.179367
21
0.94
1.61
87.50
0.051424
OCE
10
0.75
0.78
10.45
0.288410
21
0.59
2.47
8.22
0.334228
Region = Canada
Product Demand Equations
Region = Canada
CIF Price Linkage Equations
SOBS
SRSQ
SDW
SFST
UTHEIL
SOBS
SRSQ
SDW
SFST
UTHEIL
US
21
0.30
2.06
3.80
0.057893
21
0.81
1.82
24.55
0.096759
LA
21
0.33
1.15
4.38
0.38328A
21
0.87
1.95
39.33
0.123417
MED-EC
21
0.32
1.00
A. 30
0.785845
21
0.82
2.54
26.13
0.154420
EC
9
0.35
0.65
1.60
0.383800
21
0.87
2.12
37.59
0.215998
ME/NA
21
0.12
0.65
1.21
0.286251
21
0.86
2.73
34.64
0.141847
RAF
12
0.46
3.31
3.87
0.086752
21
0.88
2.16
42.52
0.123417
FE
21
0.50
1.74
8.89
0.064944
21
0.89
1.18
44.40
0.093370
OCE
21
0.52
0.87
9.69
0.228376
21
0.82
1.56
26.01
0.089568
Region = South America
Product Demand Equations
Region South America
CIF Price Linkage Equations
SOBS
SRSQ
SDW
SFST
UTHEIL
SOBS
SRSQ
SDW
SFST
UTHEIL
us
21
0.21
0.75
2.41
0.275534
21
0.87
2.22
37.70
0.145828
CAN
7
0.48
0.09
1.87
0.417780
21
0.80
2.11
21.99
0.129447
MED-EC
8
0.80
1.07
9.93
0.402183
21
0.76
2.03
18.38
0.181971
EC
21
0.12
1.01
1.19
0.352092
21
0.86
2.58
33.91
0.094185
ME/NA
11
0.14
1.67
0.65
0.786243
21
0.39
2.30
3.65
0.248191
COM4B
8
0.51
0.28
2.60
0.479401
21
0.78
2.20
19.89
0.179598
322


102
could affect all the equations in the model. Maddala (1971) supports this
assumption for large econometric models based on two essential points
where conventional methods pose problems. One is when the unrestricted
reduced form is not estimable because the number of predetermined
variables in the system is larger than the number of observations. The
second is when one uses system methods where the covariance matrix of the
residuals can not be computed again because of too few degrees of freedom.
The fresh orange trade model developed here fits partially Maddala's
classification, given that the number of equations to estimate is 242 and
the number of observations available is 21.
There are several additional benefits in using NL2SLS over the full
information methods in this particular trade model. Specification errors
are common in large econometric models, and data problems are also
expected, given that the model deals with trade data. If there is any
specification error in one of the equations of the model, the use of
NL2SLS will prevent the error from affecting the rest of the estimated
results. On the other hand, full information methods are sensitive to
small changes in specification and/or data (Goldstein and Khan, 1976).
Using NL2SLS clearly simplifies the estimation procedure, given that it
can be applied to an individual equation without directly taking into
consideration any other equation(s) in the system. In addition, full
information methods require for practical implementation sharpness of
identification of the whole model, otherwise it will interfere with the
estimation (Klein, 1969) Finally, research has been inconclusive about
the performance of the full information methods when compared to the


302
Contributions to Agricultural Economics Research
The dissertation represents the first multiple-region world trade
model for the fresh-orange industry. The study provides a conceptual
framework and model which can be used for international trade research on
other individual agricultural products. The model is a modified spatial
equilibrium model that follows Armington's demand theory that products are
differentiated by place of origin. The model is a revised version of the
Armington model, which is more flexible and capable of predicting most
trade flows and market shares accurately. There has been only one other
study that used a similar model (Sparks, 1987); however, in that case the
model was used to study a highly aggregated commodity, fresh vegetables.
This is the first time that the model has been applied to an individual
good, which is more appealing, given that aggregated goods are difficult
to differentiate.
Exchange rates are explicitly included and uniquely introduced in
the present study for this type of model. The use of the United States
CPI, instead of the regional CPIs, implies the assumption of purchasing
power parity in all regions. The model utilized regional CPIs to obtain
real prices and income.
The model was estimated using a simultaneous system of equations.
There have been only two other studies which estimated this type of model
in a simultaneous system (Deardorff and Stern, 1986; Sparks, 1987).
Finally, the estimation procedure uniquely introduced different nonlinear
relationships for the first step of the principal components procedure.


METRIC TONS (Thousands)
YEAR
ACTUAL FITTED
Figure 5.4. Total Market Demand for Fresh Oranges in the Mediterranean-EC.
138


Table 2.3--continued.
Year
Region
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
RE
OCE
COMMB
Total
W/INTR1
Total
W/0 INTRAb
z
W/INTR
Z W/0
INTR
66
Total
258596
71
106571
1514557
32796
2076
1215875
264425
65780
23545
170
3484462
3386616
100.0
100.0
76
W/INTR
464094
58
119188
1937490
122551
5926
1871856
285709
142314
11150
60356
5020692
4281323
100.0
100.0
86
413042
268
229345
2833916
194960
2182
1316721
208963
113876
47301
35046
5395620
4886606
100.0
100.0
66
Total
258596
71
103380
1514557
2841
1100
1205263
262661
24275
13702
170
3386616
76
W/0 INTR
464094
58
103078
1937307
9614
859
1410849
280216
64493
8717
2038
4281323
86
413042
268
226890
2833657
21113
1109
1092643
206844
41937
36713
12390
4886606
66
Z W/INTR
7.4
.0
3.1
43.5
0.9
0.1
34.9
7.6
1.9
0.7
.0
100.0
76
9.2
.0
2.4
38.6
2.4
0.1
37.3
5.7
2.8
0.2
1.2
100.0
86
7.7
.0
4.3
52.5
3.6
.0
24.4
3.9
2.1
0.9
0.6
100.0
66 Z
W/0 INTR
7.6
.0
3.1
44.7
0.1
.0
35.6
7.8
0.7
0.4
.0
100.0
76
10.8
.0
2.4
45.3
0.2
.0
33.0
6.5
1.5
0.2
.0
100.0
86
8.5
.0
A .6
58.0
0.4
.0
22.4
4.2
0.9
0.8
0.3
100.0
'Total includes intraregional trade.
Total does not Include intraregional trade.


LIST OF TABLES
Table Page
1.1 World Orange Production by Region 7
1.2 World Fresh Utilization by Region 8
1.3 World Processed Production by Region 9
1.4 World Fresh Orange Exports by Region 10
1.5 World Processed Orange Exports by Region 11
1.6 World Fresh Orange Imports by Region 12
1.7 World Processed Orange Imports by Region 13
1.8 World Fresh Orange Export Quantities
by Region (Excluding Intraregional Trade) 14
1.9 World Fresh Orange Export Values by Region
(Excluding Intraregional Trade) 15
1.10 World Processed Orange Export Quantities
by Region (Excluding Intraregional Trade) 16
1.11 World Processed Orange Export Values by Region
(Excluding Intraregional Trade) 17
1.12 World Fresh Orange Import Quantities by Region
(Excluding Intraregional Trade) 18
1.13 World Fresh Orange Import Values by Region
(Excluding Intraregional Trade) 19
1.14 World Processed Orange Import Quantities by
Region (Excluding Intraregional Trade) 20
1.15 World Processed Orange Import Values by Region
(Excluding Intraregional Trade) 21
2.1 World Orange Production by Region 30
2.2 World Fresh Utilization by Region 32
vi


LIST OF FIGURES
Figure Page
5.1 Total Market Demand for Fresh Oranges in the
United States 135
5.2 Total Market Demand for Fresh Oranges in Canada .... 136
5.3 Total Market Demand for Fresh Oranges in Latin
America 137
5.4 Total Market Demand for Fresh Oranges in the
Mediterranean-EC 138
5.5 Total Market Demand for Fresh Oranges in the EC ... 139
5.6 Total Market Demand for Fresh Oranges in the
Rest of Western Europe 140
5.7 Total Market Demand for Fresh Oranges in the
Middle East/North Africa 141
5.8 Total Market Demand for Fresh Oranges in the
Rest of Africa 142
5.9 Total Market Demand for Fresh Oranges in the Far
East 143
5.10 Total Market Demand for Fresh Oranges in
Oceania 144
5.11 Total Market Demand for Fresh Oranges in the
Communist Bloc 145
5.12 Total Export Supply of Fresh Oranges from the
United States 146
5.13 Total Export Supply of Fresh Oranges from
Canada 147
5.14 Total Export Supply of Fresh Oranges from Latin
America 148
x


258
product has the correct relationship and is highly elastic. A small
change in the import price index produces a large change in the demand for
this product in the United States. The relationship for the
Mediterranean-EC is also negative, but the empirical results indicate that
it is not significant (see Table 5.3). Consumption of the Mediterranean-
EC product in the United States is not affected by changes in the import
price. A similar change in the import price index for the Middle
East/North Africa and Mediterranean-EC shows that the demand for the
latter region's product is more stable. This opens an interesting
opportunity for the Middle East/North Africa to increase its market
participation in the United States. A small decrease in their import
price in the United States would causes consumption of relatively more of
their product than that from the Mediterranean-EC and other sources.
Figure 6.8 shows that United States product demands are sensitive to
changes in the size of the market. Product demand indices for the three
regions are positive and highly elastic to changes in total market demand
indices. The figure and the empirical results show that Latin America has
the smallest parameter or elasticity. Parameters for the Middle
East/North Africa and Mediterranean-EC are over four and nine times the
parameter for Latin America, respectively. Under these conditions, if the
United States total market demand increases, consumers will prefer to buy
the extra fruit first from the Mediterranean-EC, second from the Middle
East/North Africa, and finally from Latin America. In all cases, these
volumes of imports are still very small relative to total U.S. consumption
of oranges.


274
PRODUCT DEMAND INDEX (1986=1)
IMPORT PRICE INDEX (1986 = 1)
Figure 6.17. Rest of Western Europe Imports Changing Import Prices.


224
East/North Africa will increase their exports by 46.14%, 19.85%, and 7.27%
per year, respectively. In terms of the United States market share, the
model would point to a decrease in the United States share since the
market share elasticity in this equation is -.516 (see equations 4.25 and
4.26). Thus, for this specific example, the United States loses shares in
this importing region relative to other suppliers. Clearly, other
regions' shares of this market must be increasing. The Far East, Latin
America, and the Middle East/North Africa will increase their market share
in the same period.
Given that the United States share actually declines, the model can
also be used to show how much of a price adjustment would be needed to
offset the decreasing market share. These results clearly indicate that
the United States must be more price competitive in this example in order
to prevent an erosion in their share of this market.
Now, suppose that the population of the United States grows by 1% a
year for the next five years. Table 5.1 indicates that fresh orange total
market demand will increase by 2.12% a year in the same period. Table 5.3
shows the product demands' total market demand elasticities for the United
States. The model predicts that Oceania, Mediterranean-EC, Middle
East/North Africa, and Latin America will benefit from increases in United
States total market demand holding relative prices fixed. However,
Oceania and Mediterranean-EC are predicted to have the major benefits.
Suppose that world prices increase in the same proportion for all
regions. What will happen with the product demands in the Communist Bloc?
Table 5.1 shows that total market demand in the Communist Bloc will
decrease. Table 5.13 indicates that United STates product demand in that


CHAPTER 7
SUMMARY AND CONCLUSIONS
Introduction
The present study developed a fresh orange trade model to study the
major factors affecting exporter and consumer decisions in 11 regions of
the world. The specific objectives were: to specify a multiple-region
equilibrium world trade model for the fresh orange industry; to analyze
the implications contained in the estimated model; to use the estimated
parameters to study analytically the reasons for changes in market shares;
and to develop a sensitivity analysis under different economic scenarios
to make contributions to specific policy issues.
Even though the fresh orange market has experienced important
growth, several countries, including the United States, have experienced
pronounced changes in their trade patterns. The fresh orange industry is
of enormous importance for some regions, especially for the United States,
South America, Mediterranean-EEC, Middle East/North Africa, and Far East,
as producers, consumers, and exporters. Producers and exporters need to
understand the major driving factors for fresh consumption and their
competitive position in foreign markets. Such information will allow them
to compete with its benefits, possibly achieve international success, and
help to develop new markets. This industry is also important for net
importers such as Canada, EEC, rest of Western Europe, and the Communist
296


44
Spain, Italy, Portugal, and Greece. The second largest partner of the
Mediterranean-EC was the rest of Western Europe. Table 2.4 shows that the
relative importance of the rest of Western Europe in the Mediterranean-
EC' s total exports decreased from 15.2% in 1966 to 11.3% in 1986. The
third major partner of the Mediterranean-EC region was the Communist Bloc.
This region accounted for 11.0% of the Mediterranean-EC's total exports in
1966 and 8.5% in 1986. Exports to the rest of the partners were small,
but exports to the United States and Canada have increased in the last few
years.
The second major exporter region of the world was the Middle
East/North Africa. As opposed to the Mediterranean-EC region, this one
has been losing its share of the market in the last 20 years.
Participation in total world exports increased from 34.9% in 1966 to 37.3%
in 1976 (see Table 2.3). Nevertheless, the region's share of the export
market decreased to 24.4% in 1986. Examining exports without considering
intraregional trade shows that this region was losing its share of the
external market faster than its own regional market share. Table 2.3
shows that interregional percentages of the Middle East/North Africa
decreased from 35.6 in 1966 to 22.4 in 1986.
The Middle East/North Africa region's major partner was the EC
region. In 1966, 75.3% of total interregional exports from the Middle
East/North Africa countries went to the EC countries (see Table 2.4).
This percentage has since been decreasing, and in 1986 it represented only
63.7. In 1976, the percentage was lower, mainly due to an important shift
of exports to the Communist Bloc. The second and third largest partner
positions of the Middle East/North Africa region were closely shared by


337
Figure 6.21
Import
Price
OBS
Index
US
CAN
LA
MED-EC
1
0.7
1.91
1.00
3.70
1.14
0
2
0.8
1.50
1.00
2.27
1.09
0
3
0.9
1.21
1.00
1.47
1.04
0
4
1.0
1.00
1.00
1.00
1.00
1
5
1.1
0.84
1.00
0.70
0.97
1
6
1.2
0.72
1.00
0.51
0.94
1
7
Figure
1.3
6.22
Total
Market
Demand
0.62
1.00
0.38
0.91
1
OBS
Index
US
CAN
LA
MED-EC
1
0.7
3.54
1.00
0.94
0.08
0
2
0.8
2.20
1.00
0.96
0.21
0
3
0.9
1.45
1.00
0.98
0.48
0
4
1.0
1.00
1.00
1.00
1.00
1
5
1.1
0.71
1.00
1.02
1.94
1
6
1.2
0.52
1.00
1.03
3.54
1
7
Figure
1.3
6.23
Import
Price
0.39
1.00
1.05
6.17
1
OBS
Index
US
CAN
LA
MED-EC
1
0.7
0.82
2.85
3.51
0.86
2
2
0.8
0.88
1.93
2.19
0.91
1
3
0.9
0.94
1.36
1.45
0.95
1
4
1.0
1.00
1.00
1.00
1.00
1
5
1.1
1.06
0.76
0.72
1.04
0
6
1.2
1.11
0.59
0.53
1.08
0
7
Figure
1.3
6.24
Total
Market
Demand
1.16
0.46
0.40
1.12
0
OBS
Index
US
CAN
LA
MED-EC
1
0.7
0.39
0.27
0.32
0.49
0
2
0.8
0.55
0.44
0.49
0.64
0
3
0.9
0.76
0.68
0.72
0.81
0
4
1.0
1.00
1.00
1.00
1.00
1
5
1.1
1.29
1.42
1.35
1.21
1
6
1.2
1.63
1.95
1.78
1.44
2
7
Figure
1.3
6.25
Import
Price
2.01
2.61
2.29
1.70
2
OBS
Index
US
CAN
LA
MED-EC
1
0.7
1.55
1.00
0.24
0.16
1
2
0.8
1.32
1.00
0.41
0.35
1
3
0.9
1.14
1.00
0.66
0.58
1
4
1.0
1.00
1.00
1.00
1.00
1
5
1.1
0.89
1.00
1.46
1.62
1
6
1.2
0.80
1.00
2.07
2.53
1
7
1.3
0.72
1.00
2.84
3.80
1
RWE
ME/NA RAF
FE
OCE
COftffl
1.00
0.79
2.44
7.68
1.00
1.00
0.86
1.75
3.58
1.00
1.00
0.93
1.30
1.83
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.06
0.79
0.58
1.00
1.00
1.13
0.63
0.35
1.00
1.00
1.19
0.52
0.22
1.00
RWE
ME/NA RAF
FE
OCE
COM1B
1.00
0.91
11.
.97
1.52
1.00
1.00
0.95
4.
.73
1.30
1.00
1.00
0.97
2.
.08
1.13
1.00
1.00
1.00
1.
.00
1.00
1.00
1.00
1.02
0.
.52
0.89
1.00
1.00
1.05
0.
.28
0.81
1.00
1.00
1.07
0.
,16
0.73
1.00
RWE
ME/NA
RAF
FE
OCE
COWffi
0.59
1.32
2.74
1.05
1.19
0.72
1.19
1.88
1.03
1.12
0.86
1.09
1.35
1.01
1.05
1.00
1.00
1.00
1.00
1.00
1.15
0.93
0.76
0.99
0.95
1.31
0.87
0.60
0.98
0.91
1.47
0.81
0.48
0.97
0.88
RWE
ME/NA
RAF
FE
OCE
COttffi
2.54
1.01
0.38
0.67
0.11
1.79
1.01
0.55
0.78
0.25
1.32
1.00
0.75
0.89
0.52
1.00
1.00
1.00
1.00
1.00
0.78
1.00
1.29
1.11
1.82
0.62
1.00
1.63
1.23
3.14
0.50
0.99
2.02
1.34
5.19
RWE
ME/NA
RAF
FE
OCE COMffi
1.00
10.49
1.00
1.65
1.00
1.00
4.35
1.00
1.37
1.00
1.00
2.00
1.00
1.16
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.53
1.00
0.87
1.00
1.00
0.30
1.00
0.77
1.00
1.00
0.18
1.00
0.69
1.00
EC
94
96
98
00
02
03
04
EC
68
79
89
00
11
22
33
EC
93
96
37
00
75
58
45
EC
23
40
65
00
48
11
92
EC
00
00
00
00
00
00
00


CHAPTER 5
ECONOMETRIC PROCEDURE AND EMPIRICAL RESULTS
Introduction
This chapter will be divided into two main sections. The first
covers the estimation procedure and the associated econometric issues.
The second discusses the empirical results and its major implications in
terms of the fresh orange trade model developed. A general conclusion is
given at the end of the chapter.
Econometric and Estimation Procedure
The fresh orange trade model under study is based on 13
relationships for each one of the 11 regions considered. Nine of those
relationships are identities and therefore do not need to be estimated.
The rest are behavioral relationships that must be estimated. The
equations to be estimated are total market demands, export supplies,
product demands, and CIF price linkage equations. Each region has one
total market demand equation, one export supply equation, and ten product
demand and CIF price linkage equations, one for each partner region. The
total number of equations in the model including identities added to 440
and the number of equations to estimate totals 242.
98


202
discussion will cover general issues about the results; subsequently, the
major exporting regions will be addressed separately.
Economic expectations about the sign and magnitude of the different
elasticities vary depending on the variable analyzed. Based on economic
theory, the sign for the FOB export price elasticities is expected to be
positive. If the FOB export price for fresh oranges in a given region
increases, it is expected that exports from that region increase. Fresh
production elasticities are expected to be positive. If fresh production
goes up, it is expected that exports go up.
Two positive FOB export price elasticities were obtained, one for
the Middle East/North Africa, which is a major net exporter, and one for
Canada, which is a net importer. The rest of the regions have negative
FOB export price elasticities. The elasticity obtained for the Middle
East/North Africa is 1.42. This indicates that FOB export price for this
region is highly elastic. A change in the FOB export price will generate
a larger-than-proportional change in fresh export supply.
Out of the nine regions with negative elasticities, there are five
with significant parameters, three of them with strong ones. However, the
regions with strong negative signs are not major exporters. Rest of
Western Europe and Far East are net importers. Oceania exports
represented only .5% of total world exports. The results indicate that
the FOB export price is not a major factor for world fresh supply.
Fresh production is the major driver of exports. Two regions,
Canada and rest of Western Europe, have zero parameters indicating that
these regions have zero local fresh production. The rest of the regions
have correct positive fresh production elasticities. Seven out of nine


P0P5 GDP5 PRD5 CPI5 BAVAL5;
P0P6 GDP6 PRD6 CPI6 BAVAL6;
P0P7 GDP7 PRD7 CPI7 BAVAL7;
P0P8 GDP8 PRD8 CPI8 BAVAL8;
P0P9 GDP9 PRD9 CPI9 BAVAL9;
POPIO GDPIO PRD10 CPIIO BAVALIO;
POP11 GDP11 PRD11 CPI11 BAVAL11;
PRIN (NAME-PC,NCOM-6,FRAC-.98,NOPRINT) EXOG;
? Alternative #1
PC1=.96**PC1;
PC2-.96**PC2;
PC3-.96**PC3;
PC4-.96**PC4;
? Alternative #2
PCI-PCI;
PC2-PC2;
PC3-PC3;
PC4=PC1*PC1;
PC5-PC2*PC2;
PC6=PC3*PC3;
? Alternative #3
PCI-.96**PC1;
PC2-.96**PC2;
PC3-.96**PC3;
PC4-.96**PC4;
PC5-.96**PC5;
PC6-.96**PC6;
LIQ1D = LOG(IQ1D)
LIQ1_2 =
LIQ1_3
LIQ1_4 -
LIQ1_5 =
LIQ1_6 =
LIQl_7 -
LIQ1_8 -
LIQ1_9 -
LIQ1_10
LIQ1_11
LOG(IQl_:
LOG(IQl_
LOG(IQl_
LOG (IQ1_:
LOG (IQ1_J
LOG(IQl
LOG(IQl_
L0G(IQ1J
= LOG(IQ1
= LOG(IQ1
LIP1_2 = LOG(IPl_2);
2)
3)
A)
5)
6)
7)
8)
9)
LIP1_3
LIP1_4
LIP1_5
LIP1_6
LIP1_7
LIP1_8
LIP1_9
_10); LIP1
11); LIPl"
- LOG(IPl_3)
- L0G(IP1_4)
= LOG(IPl_5)
= LOG(IPl_6)
- LOG(IPl_7)
- LOG(IPl_8)
= LOG(IPl_9)
10 LOG(IP1_10);
11 L0G(IP1 11);
PARAM
TH012 1 TH112 -1 TH212 1 LH012 1 LH112 1 LH212 1 LH312 1
TH013 1 TH113 -1 TH213 1 LH013 1 LH113 1 LH213 1 LH313 1
TH014 1 TH114 -1 TH214 1 LH014 1 LH114 1 LH214 1 LH314 1
TH015 1 TH115 -1 TH215 1 LH015 1 LH115 1 LH215 1 LH315 1
TH016 .1 TH116 .1 TH216 .1 LH016 1 LH116 1 LH216 1 LH316 1
TH017 1 TH117 -1 TH217 1 LH017 1 LH117 1 LH217 1 LH317 1
TH018 1 TH118 -1 TH218 1 LH018 1 LH118 1 LH218 1 LH318 1
TH019 1 TH119 -1 TH219 1 LH019 1 LH119 1 LH219 1 LH319 1
TH0110 1 TH1110 -1 TH2110 1 LH0110 1 LH1110 1 LH2110 1
LH3110 1
TH0111 .1 TH1111 .1 TH2111 .1 LH0111 1 LH1111 1 LH2111 1
LH3111 1;
LEP2_1 LOG(EP2_l);LEP3_1 LOG(EP3_l);
LEP4_1 L0G(EP4_1);LEP5_1 = LOG(EP5_l);
LEP6_1 LOG(EP6_l);LEP7_1 = LOG(EP7_l);
LEP8_1 = LOG(EP8_l);LEP9_1 = LOG(EP9_l);
LEP10_1 = LOG(EP10_1);LEP11_1 = L0G(EP11_1);
OLSQ LEP2_1, C, PCI, PC2, PC3, PC4, PC5, PC6; LEP2_1H=@FIT;
OLSQ LEP3_1, C, PCI, PC2, PC3, PC4, PC5, PC6; LEP3_1H=@FIT;


61
so because in some cases a unique world price is assumed and in other
cases a base country or region price is used and linked with the rest of
the regions by the transportation cost. The model ignores the fact that
some regions may not trade at all with the base region. Solutions to
these models are obtained by solving an econometric simultaneous system of
equations.
The second type is the spatial price equilibrium models. These
models differ from the non-spatial and the two-region models in the fact
that they consider endogenous trade flows and market shares. Prices are
linked only between those pairs of countries that actually trade with each
other. The rest of the characteristics are similar to the ones mentioned
for the non-spatial equilibrium models, except for the solution method.
They usually follow a quadratic programming procedure for estimation.
None of the models described above can replicate all of the observed
trade flows since they are designed to predict trade flows of homogeneous
products (Grennes et al., 1977 and 1978; Thompson, 1981; and Thompson and
Abbott, 1982). If products are homogeneous, then price differences
between regions are given only by transportation costs and trade barriers.
Products may not be perfectly homogeneous and may be differentiated by
country of origin. Therefore, prices may vary between regions for reasons
other than transportation costs and trade barriers.
A serious formulation of a spatial price equilibrium model will be
to determine trade flows exclusively by minimizing the transportation
cost. According to Grennes et al. (1978) "nearly everyone who has
employed spatial models concedes that the world does not behave this way".
This situation is intuitively appealing, and indeed there is enough


54
3.0 in 1976, and 1.1 in 1986. Therefore, the principal trade of this
region was among the countries constituting the region.
The major exporter to the Middle East/North Africa countries was the
Far East region, with 85.7% in 1966, 31.2% in 1976, and 53.8% in 1986 (see
Table 2.5). The second principal exporter to this region was the
Mediterranean-EC with 9.0% in 1966, 22.8% in 1976, and 29.7% in 1986. The
third exporter was Oceania with 5.1% in 1966, 0% in 1976, and 10.3% in
1986. Even though exports from Latin America appear insignificant
compared to other exporters to the Middle East/North Africa, they have
been growing very rapidly in the last few years, passing from 0% in 1966
to 5.9% in 1986.
Canada was an important importer of fresh oranges. During 1966, its
imports represented 5.2% of the world's trade. This percentage decreased
to 4.5 in 1976 and 3.4 in 1986 (see Table 2.3). With only interregional
trade considered, these percentages increased slightly to 5.3 in 1966 and
1976, and 3.7 in 1986.
The major supplier of fresh oranges to Canada was the United States,
with 77.8% in 1966, 84.7% in 1976, and 68.0% in 1986 (see Table 2.5). The
second largest exporter to Canada was the Middle East/North Africa region,
holding 4.7%, 1.4%, and 11.8% for these years. The third major exporter
was the Far East region with 7.6% in 1966, 10.2% in 1976, and 5.9% in
1986. In the last few years, the Mediterranean-EC region, whose share was
insignificant during the 1960s and the 1970s, have increased their
participation in this market. In 1986, Mediterranean-EC supplied 10.2% of
the Canadian market. Latin America increased its share of the market from
.8% in 1966 to 2.9% in 1986.
Similarly, Oceania increased its


METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.31. Mediterranean-EC Imports of Fresh Oranges from the EC (Product Demand 45).
171


Program #2
REGION #1
FREQ NONE;
SMPL 1,7;
READ (FORMAT-LOTUS FILE-' C: \LOTUS\NEWDATA. WK1') ;
READ (FORMAT-LOTUS FILE-' C: \LOTUS\PARAM. WK1') ;
PM1_2=MP1_2/MP1D;
PM1_3=MP1_3/MP1D;PM1_4-MP1_4/MP1D;
PM1_5-MP1_5/MP1D;PM1_6=MP1_6/MP1D;
PM1_7MP1_7/MP1D;PM1_8-MP1_8/MP1D;
PM1_9-MP1_9/MP1D;PM1_10-MP1_10/MP1D;
PM1_11MP1_11/MP1D;
FRML EQ1#50 IQ1_2 = EXP(TH012 +
FRML EQiy/22 IQ1_3 = EXP(TH013 +
FRML EQiy/24 IQ1_4 EXP(TH014 +
FRML EQiy/26 IQ1_5 = EXP(TH015 +
FRML EQiy/28 IQ1_6 = EXP(TH016 +
FRML EQ1#30 IQ1_7 = EXP(TH017 +
FRML EQiy/32 IQ1_8 EXP(TH018 +
FRML EQiy/34 IQ1_9 = EXP(TH019 +
FRML EQiy/36 IQ1_10 EXP(TH0110
TH2110*LOG(IQ1D));
FRML EQ1#38 IQ1_11 EXP(TH0111
TH2111*L0G(IQ1D));
TH112 *LOG(PM1_2)
TH113*L0G(PM1_3)
TH114*L0G(PM1_4)
TH115*LOG(PM1_5)
TH116*LOG(PM1_6)
TH117 *LOG(PM1_7)
TH118*LOG(PM1_8)
TH119*L0G(PM1_9)
+ TH1110*LOG(PM1
+ TH1111*L0G(PM1
+ TH212*L0G(IQ1D))
+ TH213*L0G(IQ1D))
+ TH214*L0G(IQ1D))
+ TH215*L0G(IQ1D))
+ TH216*L0G(IQ1D))
+ TH217*L0G(IQ1D))
+ TH218*L0G(IQ1D))
+ TH219*L0G(IQ1D))
.10) +
11) +
FRML EQiy/21 IP1_2
LH312*LOG(PEN))
FRML EQiy/23 IP1_3
LH313*LOG(PEN))
FRML EQiy/25 IP1_4
LH314*LOG(PEN))
FRML EQiy/27 IP1_5
LH315*LOG(PEN))
FRML EQiy/29 I PI
LH316*LOG(PEN))
FRML EQiy/31 IP1_7
LH317*LOG(PEN))
FRML EQiy/33 IP1
EXP(LH012 + LH112*LOG(EP2_l) + LH212*LOG(YEAR) +
EXP(LH013 + LH113*LOG(EP3_l) + LH213*LOG(YEAR) +
EXP(LH014 + LH114*LOG(EP4_l) + LH214*LOG(YEAR) +
EXP(LH015 + LH115*LOG(EP5_l) + LH215*LOG(YEAR) +
6 EXP(LH016 + LH116*LOG(EP6_l) + LH216*LOG(YEAR) +
EXP(LH017 + LH117*LOG(EP7_l) + LH217*LOG(YEAR) +
8 = EXP(LH018 + LH118*LOG(EP8_l) + LH218*LOG(YEAR) +
LH318*LOG(PEN))
FRML EQiy/35 IP1_9 EXP(LH019 + LH119*LOG(EP9_l) + LH219*LOG(YEAR) +
LH319*LOG(PEN))
FRML EQiy/37 IP1_10 = EXP(LH0110 + LH1110*LOG(EP10_1) +
LH2110*LOG(YEAR)+LH3110*LOG(PEN));
FRML EQiy/39 IP1_11 = EXP(LH0111 + LH1111*L0G(EP11_1) +
LH2111*LOG(YEAR)+LH3111*LOG(PEN));
XPM1_2PM1_2
XPM1_3=PM1_3
XPM1 4-PM1 4
XEP2_1EP2_1
XEP3_1-EP3_1
XEP4 1-EP4 1


49
In 1966, the major partners of Oceania included the Far East and the
EC regions with 65.4% and 24.6% of exports, respectively (see Table 2.4).
In 1986, the main partners were the Far East and the Middle East/North
Africa regions with 71.4% and 15.4%, respectively. The EC share decreased
from 23.9% in 1976 to 5.0% in 1986. Canada's share of Oceania's exports
was .4% in 1966, 18.5% in 1976, and 5.9% in 1986. Similarly, the rest of
Western Europe sharply increased its participation in the 1970s from 3.1%
in 1966 to 21.1% in 1976. In the 1980s, this percentage decreased again
to the 1966 level. The rest of Africa was another important partner of
Oceania's exports with 2.2% in 1966, .4% in 1976, and 1.5% in 1986.
EC production of oranges is relatively small and mainly concentrated
in southern France. Nevertheless, trade data reveal some intraregional
trade and a small amount of external trade. Including intraregional
trade, world export participation of the region was .9% in 1966, 2.4% in
1976, and 3.6% in 1986 (see Table 2.3). With intraregional trade
excluded, these percentages decreased to .1, .2, and .4, respectively.
This indicates that the EC region occupied position number eight relative
to the rest of the regions with regard to world fresh orange export share
in 1986.
The main partner of EC exports is the rest of Western Europe, with
51.5% in 1966, 34.5% in 1976, and 74.2% in 1986 (see Table 2.4).
Interestingly, in 1966, 36.8% of the EC's total exports were directed to
the rest of Africa and, in 1976, 52.6% were sent to the Middle East/North
Africa. In both cases, the participation of these regions rapidly
decreased to 3.6% and .5%, respectively, in 1986. The rest of the regions
were not significant partners to the EC except for the Communist Bloc.


31
in the last decade. Table 2.2 shows the portion of that production used
as fresh product. Production utilization in fresh form decreased from
75.3% in 1966 to 65.4% in 1986 (compare data in Table 2.2 as a percent of
the corresponding figures in 2.1).
Table 2.1 also shows that the major world producer was Latin
America, with 21.6% in 1966, 29.8% in 1976, and 37.8% in 1986. This
region exhibited one of the faster annual growth rates, 6.2% during the 20
year period. However, as shown in Table 2.2, over 50% of the oranges of
this region went to the processed industry, leaving 9.4 million tons for
fresh utilization in 1986. This represented 29.3% of total world fresh
utilization.
The second largest producer of oranges was the Far East, with 15.7%
in 1966, 16.1% in 1976, and 17% in 1986. These percentages show that the
Far East region has not only maintained its participation in the total
world production of oranges in the last 20 years, but has also increased
it. Table 2.1 shows that the Far East region has doubled in absolute
terms its total production in the same period. In addition, 91.2% of
total production was used fresh in 1986.
The third largest producer was the United States, with 29.6% in
1966, 25% in 1976, and 14.7% in 1986. Even though the United States is
still a major world producer, its share of total production of oranges has
been decreasing, especially in the last decade. The United States used
most of its production in the processed industry. In 1986, 32.3% of total
production was used fresh, indicating that the United States was not the
third major supplier of oranges to the fresh markets.


METRIC TONS (Thousands)
YEAR
ACTUAL -+- FITTED
Figure 5.19. Total Export Supply of Fresh Oranges from the rest of Africa.
153


Table 5.19 Rest of Western Europe CIF Price Linkage Equations
Energy 21 Years Average
Partner
Region
Intercept
( + \-)
FOB
Price
( + )
Year
Trend
(-\ + )
Index
Price
( + )
@OBS
6RSQ
@DW
@FST
UTHEIL
Z of
Total
Imports
Z Total
Market
Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
-0.318
6.950
-0.046
0.632
0.267
2.365
0.014
1.561
0.009
0.123
0.069
1.788
21
0.90
1.36
52.20
0.081064
0.941
0.941
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-17.386
7.065
-2.461
0.040
0.215
0.185
3.696
1.544
2.394
0.154
0.046
3.377
21
0.97
1.81
207.61
0.044011
1.022
1.022
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-4.745
2.502
-1.897
0.863
0.122
7.095
1.132
0.569
1.991
-0.004
0.044
-0.103
21
0.98
1.30
242.65
0.040091
43.412
43.412
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
8.083
4.781
1.691
1.301
0.296
4.395
-1.738
1.112
-1.563
-0.012
0.092
-0.129
21
0.81
2.81
24.78
0.060969
1.501
1.501
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
-4.677
1.952
-2.395
1.074
0.132
8.156
1.217
0.432
2.819
-0.047
0.035
-1.334
21
0.98
1.80
368.62
0.029
47.547
47.547
RAF
PARAM.VALUE
STD.ERROR
t STATISTIC
-2.233
2.433
-0.918
0.650
0.135
4.807
0.429
0.559
0.768
0.122
0.049
2.458
21
0.98
2.27
225.44
0.032975
5.327
5.327
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
-6.418
10.699
-0.600
0.831
0.291
2.854
1.539
2.536
0.607
-0.024
0.188
-0.126
21
0.68
2.76
11.94
0.149779
0.039
0.039
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
7 759
4.459
1.740
0.644
0.245
2.627
-1.860
1.029
-1.808
0.238
0.087
2.724
21
0.91
2.53
58.78
0.050678
0.163
0.163
COMMB
PARAM.VALUE
STD.ERROR
t STATISTIC
-3.146
5.382
-0.584
0.606
0.159
3.820
0.627
1.218
0.515
0.089
0.051
1.732
21
0.94
2.71
86.65
0.06096
0.048
0.048
Total 100.000100.000


METRIC TONS (Thousands)
YEAR
ACTUAL -+ FITTED
Figure 5.35. Rest of Western Europe Imports of Fresh Oranges from Mediterranean-EC (Product Demand 64).
-vj
Ln


Table 1.8 World Fresh Orange Export Quantities by Region (Excluding Intraregional Trade)
Region
1966
1976
1978
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
Annual
Growth
Rate
1978-86
(000) Metric Tons
Percent of Change
United States
258.6
464.1
355.9
413.0
2.4
-1.2
1.9
Canada
0.1
0.1
.0
0.3
6.9
16.1
31.5
Latin America
103.4
103.1
152.7
226.9
4.0
8.2
5.1
Mediterranean-EC
1514.6
1937.3
1802.2
2833.7
3.2
3.9
5.8
E.C.
2.8
9.6
13.0
21.1
10.5
8.2
6.3
Rest of Western Europe
1.1
0.9
4.9
1.1
.0
2.6
-17.0
Middle East/North Africa
1205.3
1410.9
1524.9
1092.6
-0.5
-2.5
-4.1
Rest of Africa
262.7
280.2
268.6
206.9
-1.2
-3.0
-3.2
Far East
24.3
64.5
47.9
41.9
2.8
-4.2
-1.6
Oceania
13.7
8.7
21.2
36.7
5.1
15.5
7.1
Communist Bloc
0.2
2.0
4.2
12.4
23.9
19.8
14.6
World Total
3386.6
4281.3
4195.5
4886.6
1.9
1.3
1.9
Source: United Nations Trade Data Tapes.


242
TOTAL MARKET DEMAND INDEX (1986=1)
1.6
1.4
0.6
0.4
EC
~\ COMMB RWE
-0- CAN -X- FE -£>- ME/NA 1
0.7 0.8 0.9 1 1.1 1.2 1.3
INCOME INDEX (1986=1)
Figure 6.4. Total Market Demand Changing Income (GDP) (Major World
Importers).


74
Most of the work has been concentrated on the United States domestic
market analysis and in specific econometric models designed to explain one
or more elements of the international trade flow matrix and markets. The
studies are usually related to the United States product behavior in
Canada, Europe, and Japan. In most cases, the estimation has been pursued
using single-equation estimation and, in a few cases, using SUR. The
fresh orange industry has not been studied in a full simultaneous spatial
equilibrium world trade model modified to take into account that products
are differentiated by country of origin and therefore are not perfect
substitutes. The results presented in many of the articles and books
reviewed regarding trade of different commodities, and specifically fresh
and processed oranges, strongly support the conclusion that fresh citrus
coming from different countries (or regions) are perceived as different
products by consumers. The main objective of the present study will be to
develop and estimate a modified spatial equilibrium world trade model for
the fresh orange industry. The model will be used to analyze the impact
of different trade policies and economic factors affecting the demand for
fresh oranges in different regions of the world.


160
The demand for United States product is replicated quite well. The model
is not predicting some turning points in the case of the EC product
demand.
Mediterranean-EC. Mediterranean-EC represented 0.07% of total world
imports from 1966 to 1986. Figures 5.30 and 5.31 show the demand for
Latin America and EC products in the Mediterranean-EC, respectively. The
model generates the trend in both cases. Major turning points for the EC
product demand are also captured by the model. Even though the EC is not
a major producer of oranges, it does have some production and trade with
other regions of the world. Some Latin America product demand's turning
points are not reflected by the model. However, trade between Latin
America and the Mediterranean-EC was negligible until 1980. This could be
a partial explanation for the failure of the model in replicating the data
in this particular case.
EC. EC represented 63.4% of total world imports in the period
considered. Figures 5.32 to 5.34 show the demand for Mediterranean-EC,
Middle East/North Africa, and the rest of Africa products in the EC,
respectively. The model predicts the trend in every case. Figures 5.32
indicate that it also generates most turning points for the case of the
Mediterranean-EC. The demand for the Mediterranean-EC product in the EC
represented 35% of total world trade and 55% of EC's total imports.
Figures 5.33 and 5.34 show a good general fit, but some turning points are
not captured by the model.
Rest of Western Europe. Rest of Western Europe imports represented
10.6% of total world imports in the 21-year period considered. Figures
5.35 to 5.37 present the demand for Mediterranean-EC, Middle East/North


261
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.10. Canada Imports Changing Total Market Demand.


65
normally present. Therefore, different countries faced different
elasticities that may vary when market shares differ (Grennes et al. ,
1977) .
Armington (1969a, 1969b, 1970a, 1970b, 1973) developed the theory
for market share demand studies which considered goods differentiated by
place or origin. Most market-share demand studies have used this theory
because of important variations obtained in price and income elasticities
among suppliers in the foreign markets (examples are Sirham and Johnson,
1970; Ito et al. 1988; Lin et al. 1988). Later, Rhomberg (1970)
concluded that a complete demand and supply model for a world trade and
payments model could also be developed following Armington's approach.
Armington assumed a weakly separable utility function, so that
consumers' decision process may be viewed as occurring in two stages
(Varian, 1984). Equations can be derived that relate a particular trade
flow between two countries to the importing country's index of total
imports and a price ratio or relative price. Each region's market share
of a commodity may be affected by changes in the size of the market of
destination even if relative prices remain unchanged. The price ratio is
between the price of the exporting country and an average of the import
prices of the same type of product from other origins in the importing
country. The total quantity of a commodity to be imported is first
determined, and then the quantity is allocated among the competing
suppliers.
Armington assumed that the total quantity of the product imported is
a constant elasticity of substitution (CES) index of the quantities
imported from the regions of origin. The assumption was made to simplify


20
Table 1.14 World Processed Orange Import Quantities by Region (Excluding
Intraregional Trade)
Region
1978
1986
Annual
Growth
Rate
1978-86
(000) 65 Degree
Brix Metric Tons
Percent
of Change
United States
136.8
500.6
17.6
Canada
98.3
89.5
-1.2
Latin America
6.5
3.4
-7.8
Mediterranean-EC
3.8
16.6
20.1
E.C.
243.4
393.0
6.2
Rest of Western Europe
69.1
48.6
-4.3
Middle East/North Africa
13.7
9.8
-4.1
Rest of Africa
1.9
1.6
-2.5
Far East
11.3
25.0
10.4
Oceania
1.0
3.0
15.4
Communist Bloc
5.2
4.2
-2.6
World Total
591.0
1095.2
8.0
Source: United Nations Trade Data Tapes.


METRIC TONS (Thousands)
YEAR
ACTUAL -H FITTED
Figure 5.14. Total Export Supply of Fresh Oranges from Latin America.
148


Table 5.14 United States C1F Price Linkage Equations
Partner
Region
Intercept
(+\->
FOB
Price
( + )
Year
Trend
<-\+)
Energy
Index
Price
( + )
@OBS
@RSQ
@DW
@FST
21 Years
Z of
Total
UTHEIL Imports
Average
X Total
Market
Demand
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
-17.809
15.161
-1.175
0.596
0.263
2.271
4.148
3.520
1.178
-0.116
0.186
-0.624
21
0.68
2.20
11.92
0.191764
0.035
0.001
LA
PARAM.VALUE
STD.ERROR
t STATISTIC
20.197
10.327
1.956
1.098
0.498
2.203
-4.576
2.169
-2.110
0.182
0.068
2.682
21
0.77
2.13
19.20
0.079479
83.619
2.194
MED-EC
PARAM. VALUE
STD. ERROR
t STATISTIC
6.593
7.346
0.898
0.884
0.315
2.804
-1.500
1.641
-0.914
0.163
0.109
1.493
21
0.93
2.36
72.38
0.048339
2.790
0.073
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
1.660
10.364
0.160
0.987
0.209
4.728
-0.272
2.442
-0.111
0.047
0.146
0.320
21
0.68
1.37
11.84
0.134116
0.036
0.001
RWE
PARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
-10.247
8.365
-1.225
0.736
0.331
2.223
2.403
1.953
1.230
0.032
0.160
0.202
21
0.83
2.34
27.35
0.100822
11.705
0.307
RAF
PARAM.VALUE
STD.ERROR
t STATISTIC
3
0.040
0.001
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
0.589
8.295
0.071
1.131
0.350
3.230
-0.067
1.882
-0.036
-0.043
0.064
-0.670
21
0.94
1.61
87.50
0.051425
1.671
0.044
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
-6.967
16.075
-0.433
1.001
0.490
2.042
1.749
3.827
0.457
-0.084
0.332
-0.254
21
0.59
2.47
8.22
0.334228
0.104
0.003
COMMB
PARAM. VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
Total
100.000
2.624


314
LEXQD1
LEXQD2
LEXQD3
LEXQD4
LEXQD5
LEXQD6
LEXQD7
LEXQD8
LEXQD9
LEXQD10
LEXQD11
LOG(EXPORT1)
LOG(EXPORT2)
LOG(EXPORT3)
L0G(EXP0RT4)
LOG(EXPORT5)
LOG(EXPORT)
LOG(EXPORT7)
LOG(EXPORT8)
LOG(EXPORT9)
LOG(EXPORTIO)
LOG(EXPORTll)
LIQ1D -
LIQ2D -
LIQ3D -
LIQ4D -
LIQ5D -
LIQ6D -
LIQ7D -
LIQ8D =
LIQ9D -
; LIQ10D
; LIQ11D
LOG(IQID)
LOG(IQ2D)
LOG(IQ3D)
L0G(IQ4D)
LOG(IQ5D)
LOG(IQ6D)
LOG(IQ7D)
LOG(IQ8D)
LOG(IQ9D)
LOG(IQIOD);
LOG(IQllD);
PARAM
RHOl
1
RH11
-1
RH21
1
RH31
1
DHOl
1
DH11
1
DH21
1
RH02
1
RH12
-1
RH22
1
RH32
1
DH02
1
DH12
1
DH22
1
RH03
1
RH13
-1
RH23
1
RH33
1
DH03
1
DH13
1
DH23
1
RH04
1
RH14
-1
RH24
1
RH34
1
DH04
1
DH14
1
DH24
1
RH05
1
RH15
-1
RH25
1
RH35
1
DH05
1
DH15
1
DH25
1
RH06
1
RH16
-1
RH26
1
RH36
1
DH06
1
DH16
1
DH26
1
RH07
1
RH17
-1
RH27
1
RH37
1
DH07
1
DH17
1
DH27
1
RH08
1
RH18
-1
RH28
1
RH38
1
DH08
1
DH18
1
DH28
1
RH09
1
RH19
-1
RH29
1
RH39
1
DH09
1
DH19
1
DH29
1
RH010 1 RH110 -1 RH210 1 RH310 1 DH010 1 DH110 1 DH210 1
RH011 1 RH111 -1 RH211 1 RH311 1 DH011 1 DH111 1 DH211 1;
PARAM
RH41 1 RH51 1 RH47 1 RH57 1
RH42 1 RH52 1 RH48 1 RH58 1
RH43 1 RH53 1 RH49 1 RH59 1
RH44 1 RH54 1 RH410 1 RH510 1
RH45 1 RH55 1 RH411 1 RH511 1
RH46 1 RH56 1;
REPD1 EPD1/CPI1;LREPD1
REPD2 EPD2/CPI2;LREPD2
REPD3 = EPD3/CPI3;LREPD3
REPD4 EPD4/CPI4;LREPD4
REPD5 = EPD5/CPI5;LREPD5
REPD6 = EPD6/CPI6;LREPD6
REPD7 = EPD7/CPI7;LREPD7
REPD8 EPD8/CPI8;LREPD8
REPD9 EPD9/CPI9;LREPD9
REPDIO EPDIO/CPIIO;LREPDIO
REPD11 = EPD11/CPI11;LREPD11
LOG(REPDl)
LOG(REPD2)
LOG(REPD3)
L0G(REPD4)
LOG(REPD5)
LOG(REPD)
LOG(REPD7)
LOG(REPD8)
LOG(REPD9)
LOG(REPDIO);
LOG(REPD11);
OLSQ LREPD1,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6
LREPD1H=@FIT
OLSQ
LREPD2,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6
LREPD2HH§FIT
OLSQ
LREPD3,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6
LREPD3HH§FIT
OLSQ
LREPD4,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6
LREPD4HH3FIT
OLSQ
LREPD5,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6
LREPD5H= OLSQ
LREPD6,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6
LREPD6H=@FIT
OLSQ
LREPD7,
C,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6
LREPD7H=@FIT


Table 2.3 Trade Flow Analysis for Selected Years (1966, 1976 and 1986) by Region in Relation to Partner Regions
Year
Region
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
RE
OCE
COtUB
Total
W/INTR*
Total
W/O INTRAb
Z
W/INTR
Z W/O
INTR
66
us
0
0
23285
222
13
0
4617
0
212
0
0
28349
28349
0.8
0.8
76
0
11
31094
196
7
0
300
0
952
89
0
32649
32649
0.7
0.8
86
0
19
22334
15808
25
0
9706
0
1477
0
0
49369
49369
0.9
1.0
66
CAN
140366
0
1513
81
20
0
8536
16101
13805
59
0
180501
180501
5.2
5.3
76
190943
0
231
24
0
0
3225
6445
23062
1616
0
225547
225547
4.5
5.3
86
123772
0
5297
18534
68
0
21490
0
10773
2168
0
182102
182102
3. A
3.7
66
LA
6397
0
3191
0
79
0
0
0
2
0
0
9669
6A78
0.3
0.2
76
1932
45
16110
0
56
2
0
0
0
0
0
18145
2035
0.4
0.0
86
582
3
2455
61
217
0
18
A
16
0
451
3807
1352
0.1
0.0
66
MED-EC
0
0
0
0
0
0
0
0
0
0
0
0
0
0.0
0.0
76
7
0
0
183
253
28
429
0
0
0
0
900
717
0.0
0.0
86
0
0
828
259
3167
158
497
3326
0
0
8
8243
7984
0.2
0.2
66
EC
56148
62
69439
1116393
29955
1079
908131
213821
23
3371
19
2398441
2368486
68.8
69.9
76
104359
2
39303
1516481
112937
817
760700
197487
36
2085
1983
2736190
2623253
54.5
91.3
86
9336
232
176180
2217655
173847
637
696538
176595
56
1825
11665
3464566
3290719
64.2
67.3
66
RWE
9889
0
2989
229717
1464
976
121588
17024
422
429
151
384649
383673
11.0
11.3
76
4704
0
2875
177279
3318
5067
223720
20800
7
1839
52
439661
434594
8.8
10.2
86
1619
0
6076
321349
15663
1073
186964
17922
36
276
265
551243
550170
10.2
11.3
66
ME/NA
17
0
0
1024
17
0
10612
0
9799
579
0
22048
11436
0.6
0.3
76
25
0
105
29447
5060
12
461007
54334
40379
0
0
590369
129362
11.8
3.0
86
15
0
3243
16333
116
0
224078
0
29574
5650
1
279010
54932
5.2
1.1
66
RAF
54
0
9
95
1045
0
7324
1764
2
300
0
10593
8829
0.3
0.3
76
7
0
178
747
742
0
3628
5493
57
36
0
10588
5095
0.2
0.1
86
0
0
5
490
761
0
4294
2119
0
542
0
8211
6092
0.2
0.1
66
FE
38555
9
595
358
9
19
13848
15479
41505
8964
0
119341
77836
3.4
2.3
76
145976
0
0
924
5
0
9988
1149
77821
3052
3
238918
161097
4.8
3.8
86
267916
14
324
1637
53
0
5730
8997
71939
26219
0
382829
310890
7.1
6.4
66
OCE
2210
0
1420
0
0
0
1618
225
0
9843
0
15316
5473
0.4
0.2
76
6440
0
0
0
0
0
23
0
0
2433
0
8896
6463
0.2
0.2
86
9802
0
0
0
8
0
0
0
5
10588
0
20403
9815
0. A
0.2
66
COMMB
4940
0
4130
166667
194
2
139601
11
10
0
0
315555
315555
9.1
9.3
76
9701
0
29292
212239
443
0
408836
0
0
0
58318
718829
660511
14.3
15.4
86
0
0
12603
241790
1035
314
167406
0
0
33
22656
445839
423181
8.3
8.7


221
the rest of Western Europe; EC and Far East products demanded in the rest
of Africa; Middle East/North Africa product demanded in the Far East; and
Middle East/North Africa product demanded in the Communist Bloc. Given
that, in most cases, the relationship is strong and the rest of the model
is well behaved with respect to this variable, two possible alternatives
could explain this situation: a data problem, or the existence of certain
structural change still not predicted by the model.
Conclusion: economic analysis
The economic analysis shows that the model results seem more
satisfactory for total market demands than it does for export supply
equations. In most cases the model is capturing major variations of total
market demands and export supplies for leading regions in world markets.
Most total market demand elasticities were between the expected signs and
magnitudes and made sense in most cases. In the events where wrong
elasticity signs were obtained, possible explanations were given.
The results for the export supply equations indicate that the FOB
export price is apparently not a major factor for export supplies. This
is an unexpected result. The other variables included in the model are
reflecting most of the export supply variations. Fresh production is the
strongest variable in the model. Nevertheless, export supply equations
for major world exporters behaved quite well as concluded in the
statistical analysis.
Once again, the results show that major trade flow equations are
captured by the model in most cases. Product demand equations for the


299
individual countries were averaged, using different methods to obtain the
regional tariff. Nontariff barriers were not considered in the study,
given that most of them are seasonal and the model uses annual data.
Fresh and processed orange utilization was not available for most
countries. It was necessary to do a detailed literature review, including
government reports, books, magazines, other publications and personal
contacts to obtain the necessary information for each country included in
the study. The information was then aggregated by region.
The regional CPIs based on Edwards and Ng (1985) theory were not
available for the regions considered. It was necessary to create the data
set for each country and region. Appendix H presents the detailed
procedure utilized to get the final numbers. The first step was to obtain
the domestic CPIs or inflation rates and the exchange rate index per
country. The domestic CPIs were divided by the exchange rate index to get
the CPIs per country. Finally, the CPIs per country were weighted using
the 1986 trade volume to obtain the regional CPIs.
Estimation and Sensitivity Analysis Difficulties
A nonlinear two stage least squares procedure was utilized for the
estimation of the model. While the model is simultaneous and large, it
was still possible to estimate the model by sections in a Personal
Computer, using TSP. Estimation capabilities have improved considerably
in the last two years, thus greatly facilitating the use of the personal
computer.


Table 1.12 World Fresh Orange Import Quantities by Region (Excluding Intraregional Trade)
Region
1966
1976
1978
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
Annual
Growth
Rate
1978-86
(000) Metric Tons
Percent of Change
United States
28.3
32.6
52.9
49.4
2.8
4.2
-0.8
Canada
180.5
225.5
180.2
182.1
.0
-2.1
0.1
Latin America
6.5
2.0
2.2
1.4
-7.5
-4.0
-5.9
Mediterranean-EC
0.0
0.7
1.5
8.0
N. A.
27.3
23.7
E.C.
2368.5
2623.3
2546.9
3290.7
1.7
2.3
3.3
Rest of Western Europe
383.7
434.6
433.2
550.2
1.8
2.4
3.0
Middle East/North Africa
11.4
129.4
72.2
54.9
8.2
-8.2
-3.4
Rest of Africa
8.8
5.1
8.7
6.1
-1.8
1.8
-4.3
Far East
77.8
161.1
162.0
310.9
7.2
6.8
8.5
Oceania
5.5
6.5
6.7
9.8
3.0
4.3
4.9
Communist Bloc
315.6
660.5
729.1
423.2
1.5
-4.4
-6.6
World Total
3386.6
4281.3
4195.5
4886.6
1.9
1.3
1.9
Source: United Nations Trade Data Tapes.
00


231
the right-hand side variables were obtained from the simulated data sets.
In order to compare the changes in the dependent variables, only one
right-hand side variable was modified at a time. The rest of the
variables remained at their original 1986 values.
The substitution of the new data sets into the equations generated
seven new dependent variable values, one for each percentage adjustment to
the base. The new simulated dependent variable values were indexed to the
1986 base value. The index helped to show the implications from changes
in the right-hand side variables in the different equations and markets.
It also helped to compare the results among the different regions. If the
index number was one, then the new dependent variable value was equal to
the 1986 value. If the index number was above or below one, the simulated
value was above or below the 1986 value, respectively.
The sensitivity analysis and resulting responses can be illustrated
with figures that show the behavior of the different dependent variables,
given changes in the right-hand side variable. The graphical approach
provides a useful interesting framework to illustrate the impact of
changes in the different equations and markets and to compare the results
among regions.
For example, Figure 6.1 shows the total market demand for major
world consumers. Total market demands are functions of average market
price, income (GDP), population, and substitute product price. The
average market price is one of the right-hand side variables. The figure
shows an index number ranging from .7 to 1.3 for the average market price
on the bottom axis. That is, the average market price has been modified
from 30% below to 30% above the 1986 price level. The left axis shows an


APPENDIX B
DERIVATION OF THE PRODUCT DEMAND EQUATIONS
If the market demand equations follow the CRES quantity index
function of the product demands, then:
(B.l) Xt. = [b^X,,*11 + bi2*Xi2ai2 + ... + bim*Ximaim](1/Ql-)
Defining the term in parenthesis as Q, the market demand equation can be
written as follows:
(1/a, )
(B. 2) X,. = Q' *'
Taking the partial derivative of the market demand (X, ) w.r.t. the product
demands (X,j) the following result is obtained:
(B. 3) (Xi.VaCXij) = [l/ai.]*[Q(1/ai-)1]*[a(Q)/3(XiJ)]
= [ 1/a,. ] [X,. *X,. 'ai ] [aij*bij*Xij (aiJ'1} ]
Equation (4) follows from the first order condition of utility
maximization:
(B.4) aXi.vaXy) P,. = Pij
which implies that
(b.5) p,. = Pij/tacxo/acXij)]
by substituting a(X, )/a(X,j) in (5), equation (6) holds:
(B. 6) P,. = Pij/f(l/ai.)*(Xi.'Qi-*Xi.)*(aij*bij*Xij(aiJ'1))]
Rearranging terms and solving for the product demands (X^), equations (7)
and (8) follow:
(B 7) X^'1 = [Oi./Co^by)] [Pij/P,.] tX1.(0iJ'1)]
307


344
Polak, J. J., "An International Economic System," St. Martin's Press,
London, 1954.
Porter, M. E. "The Competitive Advantage of Nations," The Free Press, New
York, 1990.
Prato, A. A., "An Econometric Study of Consumer Demands for Fresh Oranges
and Frozen Concentrated Orange Juice," Ph.D. Dissertation,
Department of Agricultural Economics, University of California
(Berkeley), 1969.
Prato, A. A., "Measurement of Citrus Demands when all Variables are
Subject to Error," Journal of the American Statistical Association
65(1970):1146-1158.
Priscott, R. H., "Demand for Citrus Products in the European Market,"
Master Thesis, University of Florida, Gainesville, 1969.
Resnick, S. A., and E. M. Truman, "An Empirical Examination of Bilateral
Trade in Western Europe," Journal of International Economics
3(1973):305-335.
Rhomberg, R. R. "A Short-term World Trade," Econometrica 34(1966) : 90-91.
Rhomberg, R. R. "Possible Approaches to a Model of World Trade and
Payments," IMF Staff Paper 17(1970):1-27
Rhomberg, R. R. and L. Boissonneault, "Effect of Income and Price Changes
on the U.S. Balance of Payments," IMF Staff Paper 11(1964):59-124.
Sarris, A. H., "European Community Enlargement and World Trade in Fruits
and Vegetables," American Journal of Agricultural Economics
65(1983):235-246.
Sarris, A. H., "World Trade in Fruits and Vegetables: Projections for an
Enlarged European Community," USDA ERS Foreign Agricultural Economic
Report 202, Washington D.C., 1984.
SAS Institute Inc., "SAS/ETS User's Guide," SAS Institute Inc., Cary,
N.C., 1982.
Sirham, G. A. and P. R. Johnson, "A Market-Share Approach to Foreign
Demand for U.S. Cotton," American Journal of Agricultural Economics
53(1970):593-599.
Solow, R. M. "The Production Function and the Theory of Capital," The
Review of Economic Studies, 23(1955-56):101-108.
Sparks, A. L. "A Simultaneous Econometric Model of World Vegetable Trade:
Implications for Market Development," Ph.D. Dissertation, University
of Florida, Gainesville, 1987.


289
PRODUCT DEMAND INDEX (1986=1)
IMPORT PRICE INDEX (1986 = 1)
Figure 6.27. Communist Bloc Imports Changing Import Prices.


209
regions are negative. However, only one is highly significant. This is
the case for the Far East product. The magnitudes in the three cases
indicates that relative price elasticities for this region are inelastic.
An increase in the relative import price implies a less than proportional
decrease in demand for the product.
To understand some of the implication of the empirical results, an
example will be developed. Suppose that there is no local production and
that only two suppliers exist for Canada: the Far East (-.92) and Oceania
(-2.55) (see Table 5.27). The elasticities obtained imply that an equal
change in the relative price variables will have different effects in each
product demand. A similar increase in relative prices will cause a shift
from consuming Oceania product to consuming relatively more product from
the Far East.
Total market demand elasticities for Canada's three major partners
are elastic and significant. The elasticity for the Middle East/North
Africa product is negative, while the ones for the United States and the
Far East are positive. The magnitudes of these elasticities indicate that
the demand for these products in Canada is highly sensitive to small
changes in the size of the market. The direction of change for the Middle
East/North Africa is different from the other regions. For example, if
Canadian market size grows consumers will shift from Middle East/North
Africa product to the United States or Far East products.
EC. EC imports represented 63.4% of total world imports in the 21-
year period studied. Major EC partners are Mediterranean-EC, Middle
East/North Africa, rest of Africa, Latin America, and United States.
Product demands for the United States and rest of Africa have significant


68
reflect differences due to commodity composition as well as differences
due to country of origin.
Trade-flow and market-share models represent a major improvement
over the other models developed to study international trade, since they
can more readily depict observed trade flows. The assumption that
products are differentiated by country of origin and prices may vary
between regions for reasons other than transportation costs and trade
barriers is intuitively appealing. Furthermore, Armington's
simplification by the introduction of an import quantity index function
is, in many cases, a necessary condition to operationalize the model and
obtain as much information as possible from the trade flows. As will be
shown later, the Armington model provides several practical solutions for
dealing with a large number of equations and parameters.
Trade Models: The Orange Industry
The fresh orange industry has been studied many times, usually in
the context of national markets. A few studies have been developed in the
international trade of fresh oranges. In addition, none of the research
developed so far considered a complete world trade model for this
particular good. Most of the studies have been either partial or
descriptive. One of the earliest international trade documents is a
descriptive study developed by the U.S. Department of Commerce (1940),
which showed citrus world production and trade statistics and trends.
Before the 1950s, little demand estimation for citrus fruits
existed. More attention has now been devoted to this economic area by the


191
following discussion will cover four of the five statistics. The "t"
statistic will be evaluated in the fourth part of this section of the
chapter together with the economic analysis of the parameter signs and
magnitudes.
The @DW statistic is generally used to determined the existence and
type of serial correlation in time series. If data are given on a yearly
basis, as is the case for the fresh orange trade model, evidence of serial
correlation would probably be related to model misspecification. If the
@DW is close to two, then there is no evidence of misspecification.
A criterion that is used to evaluate a simulation model is the fit
of the individual variables in a simulation context. It is expected that
the results of a historical simulation match the behavior of the real
world rather closely. It is therefore interesting to perform a historical
simulation and examine how closely each endogenous variable tracks the
historical data. This is especially important when the model is
nonlinear, given the weakness of the @RSQ and @FST in those cases.
Theil's Inequality Coefficient (UTHEIL) is a useful simulation statistic
related to the RMS (Root-Mean-Square) simulation error and applied to the
evaluation of historical simulations or ex post forecasts. The UTHEIL
will give an idea on how well the model captures the turning points of the
estimated equations. If the value of UTHEIL is zero, then the predicted
value is equal to the actual value and there is a perfect fit. If UTHEIL
is equal to one, then the predicted performance of the model is no better
than a random estimate.
Tables 5.1 to 5.24 present the empirical results of the estimated
fresh orange trade model and includes the major statistics discussed


236
equation analyzed. By construction, this index always goes from .7 to 1.3
independently of the variable considered. That is, the right-hand side
variable has been modified over that range. On the left axis, the index
represents the endogenous or response variable. The specific variable on
the left axis could be total market demand, export supply, or product
demand depending on the equation studied. The index varies, depending on
the type of response of the endogenous variable in each case. The
response depends on the percentage change in the right-hand side variable
and the magnitude and sign of the estimated parameter.
The right-hand side variable index and the endogenous or response
variable index were used to construct the figures. Each figure shows the
regions in order of importance. The first region presented corresponds to
the most important region in the figure, the second to the next most
important, etc. The most important region corresponds to the largest
consuming region or importer for total market demands and to the largest
exporting region for export supplies. The criterion for product demands
was based on trade-flow volumes between partner regions and the final
market.
The first section will center on total market demands, the second on
export supplies, and the third on product demands. In each case, the
discussion will address consumers, importers, exporters, and trading
partners. A summary regarding this section of the chapter will be
presented at the end.


214
inelastic one for Mediterranean-EC. Given an increase in the Communist
Bloc market size, consumers will consume relatively more from Latin
America and the Middle East/North Africa than from Mediterranean-EC.
Latin America. Mediterranean-EC, rest of Africa, and Oceania. Latin
America, Mediterranean-EC, rest of Africa, and Oceania are net exporters.
Their imports represented only .5% of total world imports during the 21-
year period considered. Latin America major partners are the United
States and EC. The results for relative price elasticities indicate that
in Latin America the product demand for the United States has a positive
elasticity and, for the EC is negative but not significant.
Mediterranean-EC major partners are Latin America, EC, and Middle
East/North Africa. Demands for EC and Middle East/North Africa products
have negative and significant elasticities. Relative price elasticity for
Latin America product demand is positive but not significant.
Rest of Africa major partners are EC, Middle East/North Africa, and
Oceania. Elasticities for EC and Middle East/North Africa products are
positive. However, the elasticity from the EC is not significant. The
demand for the Oceania product is negative and significant.
Oceania major partners are the United States, Latin America, and the
Middle East/North Africa. Elasticities for the United States and Middle
East/North Africa products are negative and significant. Latin America
product has a positive and significant elasticity.
Relative price elasticities turned out to be positive and
significant in three cases. They are the demands for United States
product in Latin America, Middle East/North Africa product in rest of
Africa, and Latin America product in Oceania. As mentioned before, for


ACKNOWLEDGEMENTS
I wish to give particular thanks to Dr. Ronald W. Ward,
chairman of my doctoral committee, for the guidance, assistance and
constructive criticism given to me during the preparation of this
dissertation. Similarly, I would like to express my gratitude to the
other members of my advisory committee--Drs. Kenneth R. Tefertiller, James
L. Seale, Jr., Max R. Langham, and Terry L. McCoy, for their advice and
support in the preparation of this document. Special thanks are given to
Sharon Bullivant and Melissa Bracewell in helping me put the dissertation
in its final form for the Graduate School. Special thanks are also given
to Bill Messina for his support and encouragement during the preparation
of the document.
I also appreciate the funding I received from the the United States
government through the Fulbright Scholarship Program. I want to thank my
friends from Liberty Church of Gainesville for their support,
understanding, and prayers. I give special thanks to my parents for they
taught me to appreciate education and always gave me the necessary support
to accomplish my academic achievements.
Finally, to my wife and son, I owe the most gratitude. They were
patient, understanding, and supportive all these years, and thanks to them
I finished this document. So, I would like to dedicate this dissertation
to my wife and son.
ii


Table 1.9 World Fresh Orange Export Values by Region (Excluding Intraregional Trade)
Region
1966
1976
1978
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
Annual
Growth
Rate
1978-86
In
Million of U.
, S. Dollars
Percent of Chang<
2
United States
47.0
118.8
144.0
233.0
8.3
7.0
6.2
Canada
.0
.0
.0
0.1
11.5
20.7
26.6
Latin America
6.2
14.7
27.9
61.5
12.1
15.4
10.4
Mediterranean-EC
178.9
406.1
530.0
1025.2
9.1
9.7
8.6
E.C.
0.7
2.8
5.7
12.6
15.2
16.1
10.3
Rest of Western Europe
0.2
0.3
1.3
0.6
5.4
9.2
-8.3
Middle East/North Africa
172.7
322.1
410.7
361.6
3.8
1.2
-1.6
Rest of Africa
35.7
59.1
103.6
83.5
4.3
3.5
-2.7
Far East
4.6
21.1
17.2
17.6
7.0
-1.8
0.2
Oceania
2.5
3.1
6.5
15.2
9.4
17.1
11.2
Communist Bloc
.0
0.4
0.8
3.9
28.7
25.0
21.9
World Total
448.7
948.6
1247.7
1814.7
7.2
6.7
4.8
Source: United Nations Trade Data Tapes.


273
Rest of Western Europe
Table 6.2 indicate that four regions accounted for 98.7% of total
imports in the rest of Western Europe. The sensitivity analysis is
presented in Figures 6.17 and 6.18. Figure 6.17 shows the relationships
for the import price and product demand indices. The figure indicates
that three of the four relationships have the correct negative direction.
Only the demand for the rest of Africa product has a positive and
significant association (see Table 5.8). The responses of the other three
product demands with negative relationships are very similar, indicating
that consumers in the rest of Western Europe will not shift their
consumption among sources if all prices change proportionally.
Figure 6.18 indicates that the relationships between the total
market demand indices and the product demand indices are positive in all
cases. The parameter for the Middle East/North Africa is not significant,
indicating that demand for this product is not affected by change in
market size. The figure shows that responses are different among regions.
It shows that, if market size increases, consumers will increase their
consumption first from the EC product, second from the Mediterranean-EC,
and third from the rest of Africa.
The results indicate that the rest of Africa import price/product
demand relationship is not significant. The positive and relatively
strong association between product demand and market size gives this
region an opportunity to take advantage of potential market growth.


34
Trade Flow Analysis
Table 2.3 shows the quantities traded between the 11 regions for
1966, 1976, and 1986. These years were selected to illustrate changes
through time. The first column of this table represents the different
years, the second column and the top row of the table represent the region
and the partner region names, respectively. Each of the 11 columns
depicts the quantities exported from the partner region to each region.
The following two columns show the total product imported by each region,
with the first one including the intraregional trade and the other
including interregional trade. Since the first and second regions consist
of a single country, both columns display the same values. The last two
columns exhibit the percentages associated with the previous two columns
in relation to total world imports. Similarly, the last four rows of the
table contain total exports from each partner region. The first row
includes intraregional trade, the second one only interregional trade, and
the last two rows show the percentages associated with total world
exports.
Tables 2.4 and 2.5 contain the percentages needed to illustrate the
allocation of exports, imports, and trade flows in total and among the 11
regions. Table 2.4 shows the percentages from the exporter or partner
region position and Table 2.5 from the importer or region perspective. In
both tables, intraregional trade was excluded, given that the major
interest of the present study has to do with trade among the regions.
Intraregional quantities are part of the region's production that is
consumed domestically.


285
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.24. Far East Imports Changing Total Market Demand.


84
Aj is the percentage of total orange production utilized in the
processed industry in region j and is assumed to be exogenous.
Export Supply Equations
Exporters will respond to export prices by adjusting their level of
exports accordingly. Given changes in total production and fresh
utilization, exports will also tend to adjust accordingly.
Export supply equations are consequently assumed to be a function of
the average export price from region j (average Free On Board price = F j)
and total fresh orange utilization (PRDXj) in the region of origin. The
export supply equation for fresh oranges is the following:
(4.13) X.j = Si Xij = v(F+j,PRDij) for i*j
where
The summations represent total exports of fresh oranges from region
j to all other regions,
v represents some functional relationships between variables,
F j represents the average export price of fresh oranges from
region j to all other regions.
The demand equations for local product will follow from the
difference between total fresh utilization (PRDij) plus the change in
inventories (when applicable) and the export supply from region j Demand
for domestically produced product is:
(4.14) Xjj = PRDu + A INVj X.j
Given that fresh oranges can be stored only for short periods of
time, it is assumed that inventory levels are zero. Accordingly, the
change in inventories will be zero and equation 4.15 will be given by
(4.15) Xjj = PRDj X.j


METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.1. Total Market Demand for Fresh Oranges in the United States.
135


213
This result confirms the discussion maintained in previous sections about
the characteristics of the markets and consumers in the Far East with
regards to fast growth and interest in quality and high-grade fruit. This
is especially true for trade between the United States and the Far East
markets.
Total market demand elasticities for Far East's four major partners
are positive except for the case of the Middle East/North Africa.
Elasticities are 2.66 for the United States, 2.69 for the rest of Africa,
and 1.3 for Oceania. Middle East/North Africa elasticity is negative but
not significantly different from zero. Elasticity magnitudes indicate
that product demands from the different sources are very sensitive to
changes in the size of the market. Since the Far East market has been
growing fast in the last 21 years, shifts from one region to another are
common and will probably continue in the future.
Communist Bloc. The Communist Bloc imports represented 13.6% of the
world's total imports in the period considered. Major partners are Middle
East/North Africa, Mediterranean-EC, and Latin America. Two of them have
significant negative relative price elasticities. Mediterranean-EC has a
positive but not significant elasticity. Middle East/North Africa and
Latin America results show product demands with highly elastic relative
price elasticities.
Total market demand elasticities for Communist Bloc's three major
partners are positive and significant. Elasticities are 3.38 for Latin
America, .28 for Mediterranean-EC, and 1.24 for the Middle East/North
Africa. The results show an elastic response with respect to total market
demand for Latin America and Middle East/North Africa products and an


235
Product demand equations were defined as functions of the relative
price and total market demand. Relative price variables refer to the
import price of a product coming from a certain region relative to the
final market average price. The import price could change relative to the
final market average price, due to changes in tariffs, taxes, FOB export
prices, other factors included in the CIF equations, and other causes. If
the import price for a certain region increases relative to the average
market price, less consumption relative to other suppliers is expected in
the final market. Total market demand variables measure total consumption
of fresh oranges in the final market. It is a measure of the size of the
market. Economic theory and the empirical results in Chapter 5 indicate
that both variables help to determine trade flows. Relative price and
total market demand variables were included in the sensitivity analysis.
Sensitivity Analysis
This section of the chapter will present and evaluate the results of
the sensitivity analysis. Each type of equation will be addressed
separately. The analysis and discussion will focus on major trading
regions. To ease the presentation and discussion, a graphical analysis
constructed using the sensitivity analysis indices was developed. The
figures will be used to evaluate individual market behavior and to compare
them among regions. The indices generated in the sensitivity analysis are
included in Appendix J.
The figures show, on the bottom axis, an index that represents the
right-hand side variable modified. The variable varies, depending on the


206
will be addressed separately and both relative price and total market
demand variables will be analyzed in each case. Latin America,
Mediterranean-EC, rest of Africa, and Oceania are net exporters; their
imports represented only .5% of total world imports during the 21-year
period considered. These regions will be analyzed briefly following the
discussion about leading importers.
Tables 5.27 and 5.28 show the estimated parameters and their
associated "t" statistics for the product demands. Product demands
measure the demand in a given region for fresh oranges coming from another
region. There will be one product demand in each region for each one of
the partner regions. Since the model has a total of 11 regions, the
number of estimated product demands should be 110 (11 regions with ten
partners each).
The variables considered in the estimation of this section of the
model are relative price and total market demand. The relative price
variable refers to the price of the imported product (for example, the
price of fresh oranges from Latin America in the EC) relative to the
average market price (in the EC). If the relative price variable
increases, imported product price is going up faster than the average
market price. In that case, demand for that product should decrease in
the final market. This example implies that the exporting region will be
losing part of its market share in the final market. Therefore, the
expected sign for relative price elasticities is negative.
The other variable included in the model is total market demand.
This variable measures apparent consumption or total size of the market
for fresh oranges in a given region. If market size increases


293
regarding production and exports. This region has better opportunities to
take advantage of new or growing markets in the future. However, product
quality has to be improved in order to take advantage of any new
opportunity.
Product demand analysis provides important information for exporters
and importers. The following conclusions will be made from the point of
view of the exporting regions. The United States has good opportunities
to increase and/or maintain its market share in four regions: Canada,
Latin America, Far East, and Oceania. In most regions, the United States
holds a premium position. The most promising region is the Far East,
which is one of the faster growing regions in the world. Latin America
apparently has better opportunities in the EC, Mediterranean-EC, Middle
East/North Africa, and Communist Bloc. Its best market position is found
in the Middle East/North Africa, which is one of the faster growing
regions. Mediterranean-EC has better opportunities in the United States,
EC, rest of Africa, and the Communist Bloc. Its holds a premium position
in all regions, indicating that its fruit quality levels are high and
consumers are willing to pay the price. The Middle East/North Africa has
opportunities in the United States and the Communist Bloc. Its
competitive position in most markets is not good. In most regions, its
market size parameter is negative, probably indicating that quality
standards are poor. The rest of Africa has opportunities in the rest of
Western Europe and Middle East/North Africa. Its position is especially
strong in the rest of Western Europe. The Far East holds a strong second
position in the Canadian market. Oceania has an important opportunity in
the Far East market. Its geographical position is better than its major


METRIC TONS (Millions)
10
8
0
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
YEAR
ACTUAL
FITTED
Figure 5.9. Total Market Demand for Fresh Oranges in the Far East.


203
have significant parameters. The two regions with insignificant
parameters are the Far East, which is a net importer, and Oceania, which
is a minor exporter.
The results show that the export supply behavior for major world
exporters is good. Mediterranean-EC, with exports accounting for 44% of
total world exports, has a well-behaved export supply equation with
insignificant FOB export price elasticity but a strong positive fresh
production elasticity. The Middle East/North Africa, accounting for 35%
of total world exports, has a well-behaved export supply equation with
strong positive FOB export price and fresh production elasticities. The
United States and Latin America, accounting for 12.8% of total world
exports, have well-behaved export supply equations. Both regions have
strong positive fresh production elasticities. The United States has
negative but weak FOB export price elasticity, and Latin America's price
elasticity is insignificant.
Product demand
In previous sections, most of the equations were analyzed
graphically and in terms of fit, performance, and simulation ability. In
order to avoid unnecessary repetition, the emphasis of the following
discussion will be on important trade flows. The fresh orange trade model
developed here is basically interested in understanding the demand factors
that make regions shift their imports from one source to another.
Decisions about relevance have to be made first by selecting major world
importers and then by identifying their major suppliers. Relevant regions


Table 5.18 EC CIF Price Linkage Equations
Energy 21 Years Average
Partner
Region
Intercept
( + \-)
FOB
Price
( + )
Year
Trend
(-\+)
Index
Price
( + )
SOBS
@RSQ
@DW
@FST
UTHEIL
Z of
Total
Imports
Z Total
Market
Demand
us
PARAM.VALUE
STD.ERROR
t STATISTIC
-2.268
8.547
-0.265
0.556
0.264
2.103
0.399
1.945
0.205
0.128
0.082
1.557
21
0.85
2.10
31.16
0.087405
1.821
1.812
CAN
PARAM.VALUE
STD.ERROR
t STATISTIC
5.290
13.454
0.393
1.021
0.468
2.183
-1.167
3.070
-0.380
0.028
0.193
0.144
21
0.62
2.62
9.32
0.193149
0.002
0.002
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-6.349
7.108
-0.893
0.210
0.211
0.995
1.167
1.581
0.738
0.176
0.074
2.370
21
0.90
1.92
52.34
0.073432
3.278
3.265
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
6.695
3.022
2.216
1.065
0.1A 7
7.241
-1.468
0.671
-2.190
-0.011
0.051
-0.220
21
0.97
1.97
184.49
0.042842
54.192
53.955
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
2.582
5.782
0.447
1.274
0.356
3.576
-0.405
1.349
-0.300
-0.121
0.123
-0.981
21
0.74
2.20
15.87
0.09092
0.030
0.029
ME/NA
PARAM.VALUE
STD.ERROR
t STATISTIC
-2.449
3.136
-0.781
1.063
0.233
4.557
0.671
0.705
0.952
-0.030
0.064
-0.467
21
0.94
2.10
89.99
0.045453
32.690
32.548
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
-0.561
4.433
-0.127
0.579
0.230
2.523
0.005
0.986
0.005
0.102
0.068
1.490
21
0.93
2.77
71.77
0.056549
7.691
7.656
FE
PARAM.VALUE
STD.ERROR
t STATISTIC
2.336
8.875
0.263
0.935
0.239
3.915
-0.479
2.110
-0.227
-0.033
0.157
-0.212
21
0.66
2.21
10.97
0.126091
0.007
0.007
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
0.820
5.792
0.142
0.886
0.386
2.295
-0.145
1.406
-0.103
-0.016
0.168
-0.092
21
0.78
1.83
20.58
0.075877
0.108
0.106
COMMB
PARAM.VALUE
STD. ERROR
t STATISTIC
2.951
4.699
0.628
0.895
0.244
3.671
-0.682
1.051
-0.649
0.042
0.072
0.590
21
0.88
2.34
40.39
0.05886
0.181
0.180
Total 100.000 99.560


328
XREPD10-REPD10; XPRDIO-PRDIO;
XREPD11-REPD11; XPRD11-PRD11;
? SIMULATION #1 TOTAL MARKET DEMAND VARYING AVERAGE MARKET;
SMPL 1,1;
I-.5;
SMPL 2,20; I-I(-l)+.l;
SMPL 1,20;
ID-1;
RMP1D-XRMPID*I; RMP2D-XRMP2D*I; RMP3D-XRMP3D*I; RMP4D=XRMP4D*I;
RMP5D-XRMP5D*I; RMP6D-XRMP6D*I; RMP7D-XRMP7D*I; RMP8D-XRMP8D*I;
RMP9D=XRMP9D*I; RMP10D-XRMP10D*I; RMP11D-XRMP11D*I;
GENR EQ1#20;GENR EQ2y/20; GENR EQ3#20;GENR EQ4#20;GENR EQ5#20;
GENR EQ6#20;GENR EQ7#20;GENR EQ8#20;GENR EQ9#20;GENR EQ10#20;
GENR EQliy/20;
WRITE (FORMAT-LOTUS,FILE-'C:\LOTUS\SIM#1.WK1')
ID I RMP1D RMP2D RMP3D RMP4D RMP5D RMP6D RMP7D RMP8D RMP9D
RMP10D RMP11D
IQ1D IQ2D IQ3D IQ4D IQ5D IQ6D IQ7D IQ8D IQ9D IQ10D IQ11D
EXPORT1 EXP0RT2 EXPORT3 EXP0RT4 EXPORT5 EXPORT6 EXPORT6
EXP0RT7 EXPORT8 EXPORT9 EXPORTIO EXP0RT11;
RMP 1D-XRMP1D; RMP 2 D-XRMP 2 D; RMP3D-XRMP3D; RMP4D-XRMP4D; RMP5D-XRMP5D;
RMP6D-XRMP6D; RMP 7 D-XRMP 7 D; RMP8D-XRMP8D; RMP 9 D-XRMP 9 D; RMP10D-XRMP10D;
RMP11D-XRMP11D;
? SIMULATION #2 TOTAL MARKET DEMAND VARYING INCOME (GDP);
SMPL 1,1;
I-.5;
SMPL 2,20; I-I(-l)+.l;
SMPL 1,20;
ID-1;
GDP1=XGDP1*I;GDP2=XGDP2*I;GDP3-XGDP3*I;GDP4-XGDP4*I;GDP5-XGDP5*I;
GDP6-XGDP6*I;GDP7-XGDP7*I;GDP8-XGDP8*I; GDP9-XGDP9*I;GDP10=XGDP10*I;
GDP11=XGDP11*I;
GENR EQ1#20;GENR EQ2#20;GENR EQ3#20;GENR EQ4#20;GENR EQ5#20;
GENR EQ6y/20; GENR EQ7#20;GENR EQ8#20;GENR EQ9#20;GENR EQ10#20;
GENR EQ11#20;
WRITE (FORMAT-LOTUS,FILE-'C:\LOTUS\SIM#2.WK1')
ID I GDP1 GDP2 GDP3 GDP4 GDP5 GDP6 GDP7 GDP8 GDP9
GDP10 GDP11
IQ1D IQ2D IQ3D IQ4D IQ5D IQ6D IQ7D IQ8D IQ9D IQ10D IQ11D;
GDP1-XGDP1;GDP2-XGDP2;GDP3-XGDP3;GDP4-XGDP4;GDP5-XGDP5;GDP6-XGDP6;
GDP7-XGDP7;GDP8-XGDP8;GDP9-XGDP9;GDP10-XGDP10;GDP11-XGDP11;


86
associated to a similar set of prices through the FOB (Free On Board)
export price.
In this section, the price linkage equations among regions are
presented. In each region there is an export price that corresponds to
each region. This price is the FOB export price and will be denoted F^.
Accordingly, the average export price for fresh oranges (F j) from region
j to all regions is defined as follows:
(4.19) F j = tSt ^(Fy XtJ)] / i,j Xij]
The numerator in equations (4.19) represents the total export value
from region j to all other regions and the denominator represents the
total quantity exported. The use of i*j is because no data are available
for within-region export price (Fjj) and the equations represent the
weighted average export price, which should not include the local price.
Since Fjj is not available for all regions and will be used in the
following calculations, it will be assumed to equal F j.
The CIF price is the price of a product in the port of final
destination. The represent the CIF price of fresh oranges coming from
region j to region i. These prices do not include any trade barriers and
are a function of the Fi price. Changes in the FOB export prices (F^)
are not expected to have the same impact on the CIF import prices (C^)
across regions. This follows from the assumption that certain market
structures could exist and prevent perfect transmission of prices from the
regions of origin to the regions of destination. Spoilage or product
deterioration during the transportation process from region j to region i
could be different than from region r to region i. Therefore, the general
relationship between CIF (C^) and FOB (F^) prices might not be linear.


6.20 Middle East/North Africa Imports Changing Total
Market Demand 278
6.21 Rest of Africa Imports Changing Import Prices 280
6.22 Rest of Africa Imports Changing Total Market
Demand 281
6.23 Far East Imports Changing Import Prices 283
6.24 Far East Imports Changing Total Market Demand 285
6.25 Oceania Imports Changing Import Prices 287
6.26 Oceania Imports Changing Total Market Demand 288
6.27 Communist Bloc Imports Changing Import Prices 289
6.28 Communist Bloc Imports Changing Total Market
Demand 291
xiv


69
Florida Agricultural Experiment Station and by the Florida Department of
Citrus (FDOC). As reported by Chapman (1963), the first major step in
this area was the work on experimental pricing techniques applied to the
orange demand analysis developed by Godwin and Powell during the 1950s.
Chapman (1963) and Godwin et al. (1965) developed a study on demand and
substitution relationships for California and Florida Indian River and
Interior Valencia fresh orange market. Their research was basically
concentrated in the U.S. market and focused on questions regarding own-
price elasticities and cross-price elasticities between the three regions'
in the Grand Rapids, Michigan market.
Dean and Collins (1967, 1968) studied the effects of the European
Community (EC) tariff policies in a model of world trade for fresh
oranges. Their paper included a summary of world production, consumption,
and trade of fresh oranges. Projections of orange production and
consumption, estimates of transportation costs, possible future tariffs,
and income and price elasticities of demand in the EC for 22 regions were
also included. The price elasticities were estimated at the import demand
level, i.e., at the location of consumption, but before retail margins
were added to the wholesale price. Transportation costs as well as
tariffs and any special import taxes were included in determining the
wholesale price level. Using a transportation model analysis, the impact
of possible future tariff policies in the EC was procured on producer and
consumer prices in each of the major countries and on trade flows.
Finally, using the results obtained in the different tariff scenarios, the
welfare effect on consumers and producers was also captured. The major
implication of this document and the ones by Chapman (1963) and Godwin et


245
Figure 6.2 presents the same indices as Figure 6.1, but for major
world importers. The bottom axis shows the average market price index.
The left axis shows the total market demand index. In this case, the
response index goes from below one up to approximately 1.5 times the 1986
level. This figure indicates that major world importers have different
responses to changes in the average market price, but they are closer to
each other than to the responses of major world consumers. The responses
are negative as expected, with the exception of the Far East. The EC is
Che major world importer. It has an elastic price parameter or elasticity
which is very similar to the ones from the Communist Bloc, Middle
East/North Africa, and the rest of Western Europe. The empirical results
indicate that the four elasticities are significant and lie between 1 and
1.22 (see Table 5.1). Canada's response is inelastic, implying that it is
less sensitive to changes in the average market price than former regions.
Figures 6.1 and 6.2 indicate that regions with high import levels,
such as the EC, Communist Bloc, rest of Western Europe, and Middle
East/North Africa, have similar total market demand indices; and their
parameters or elasticities are elastic. Regions with high consumption
levels and low imports relative to consumption will tend to have lower
average market price indices. Therefore, they are not very sensitive to
changes in average market prices. This is the case for Latin America and
Mediterranean-EC. These conclusions show that if world prices increase,
major importers will tend to consume proportionally less than regions with
low import levels. If world prices decrease, importers will tend to
consume relatively more than regions with low import levels.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ECXWNQ8R9_EY6MW0 INGEST_TIME 2015-08-13T21:31:53Z PACKAGE AA00029756_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


47
last decade. Some countries of the Middle East/North Africa region
utilized the fresh product to produce Frozen Concentrated Orange Juice
(FCOJ). The rest of Western Europe is another important market for the
Latin American product. An interesting issue about this region is that
its percentage of participation has not changed significantly over the
years.
The rest of Africa used to be the third largest exporter of the
world, but its share of the market has been decreasing, especially from
1976 to 1986. With intraregional trade considered, this region's share of
the world's export market was 7.6% in 1966, 5.7% in 1976, and 3.9% in 1986
(see Table 2.3). Given that most of its trade was external, these
percentages increased to 7.8, 6.5, and 4.2, respectively, when only
interregional trade is considered. The region's share of the market
indicates that it occupied the fifth position relative to the other
regions in 1986.
The major export market for the rest of Africa was the EC region.
This region represented 81.4%, 70.5% and 85.4% of the total rest of Africa
exports in 1966, 1976, and 1986, respectively (see Table 2.4). The second
most important partner was the rest of Western Europe, which absorbed
6.5%, 7.4%, and 8.7% of total exports in the same years. The rest of
Africa exports to Canada represented 6.1% in 1966 but decreased to 0% in
1986. In that year, the Far East region was the third largest market for
the rest of Africa. Exports to that region represented 5.9% in 1966, .4%
in 1976, and 4.3% in 1986. During the 1970s, exports from the rest of
Africa to the Middle East/North Africa region increased sharply and later
decreased.


301
signs and magnitudes of the estimated parameter meet economic expectations
in a majority of cases.
The model consists of 11 regions, including all countries of the
world. It was found that trade is concentrated in a few regions. The
performance of the model is better where significant trade took place.
However, regions with small participation have important growing export or
import markets; reducing the size of the model will hide important
information and opportunities for some regions.
The analysis of the demand parameters showed the likely future
direction of trade. Price elasticities were used to predict responses in
the different markets to changes in prices. The role of prices as an
allocative tool was shown. Income and population elasticities gave an
indication of possible adjustments in consumption and trade patterns.
Fresh production was found to be the most important factor contributing to
world exports. Relative import price and market size were found to be
important product demand drivers for most trade flows.
The model made forecasts of trade patterns among importers and
exporters possible. The model was used to construct a sensitivity
analysis to predict and compare total market demands, export supplies, and
product demands responses among regions. Simulations were completed
giving shocks in the different variables including average market price,
income, relative prices, market size, FOB average export price and fresh
production.


303
Further Research
There are many areas to which future research could be directed.
The first would be to use the conceptual framework and model developed
here in a different agricultural product. The changes to be made are
minor and, obviously, related to the individual characteristics and trade
patterns of the product selected.
Another interesting area of research would be to work with the fresh
orange industry, modifying the model by reconsidering the number of
regions and the country composition. It is also important to investigate
and evaluate alternative functional forms for some of the equations. This
would represent a tremendous amount of work, but it would probably provide
a better model that could be used for many different products. It is
important to recognize the significant changes in Eastern Europe which may
affect some of the conclusions of the present study.
It was not possible to obtain the reduced form of the fresh orange
trade model. It is important that future research pursue the possibility
of obtaining the reduced-form parameters of the model. The procedure
developed will be useful in many ways, since the same model can be used
for other products.
The results of the present study suggest the presence of some
specification problems. Specification tests other than the Durwin Watson
where not conducted. An interesting area for future research will be to
apply specification tests for this particular model and evaluate and
measure the specification errors properly. Models such as used here with
large number of equations do not lend themselves to certain types of test.


218
Table 5.30 CIF Price Linkage Year Trend Elasticity3
Region ib
REGION jb
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
CCKMB
US
3.462
0.999d
-6.093
-1.452
0.399
0.205
0.014
0.009
0.033
0.005
0.478
0.198
0.086
0.060
1.268
1.042
-1.620
-0.421
CAN
4.U6
1.178
-2.441
-0.556
-1.167
-0.380
-2.603
-0.999
LA
-4.576
-2.110
3.902
1.907
5.846
1.634
1.167
0.738
3.696
2.394
0.507
0.217
0.715
0.424
1.269
0.563
0.478
0.291
-2.826
-1.199
MED-EC
-1.500
-0.914
1.098
0.473
0.128
0.047
-1.468
-2.190
1.132
1.991
-3.046
-0.973
5.798
1.909
-0.660
-0.430
1.526
0.583
-0.078
-0.108
EC
-0.272
-0.111
-4.019
-1.342
0.775
0.498
2.425
1.097
-1.738
-1.563
2.450
0.756
0.066
0.107
-2.652
-0.844
0.001
0.000
RWE
4.199
1.406
-0.405
-0.300
-3.583
-0.716
0.997
0.385
-1.873
-0.472
ME/NA
2.403
1.230
-0.090
-0.048
2.005
0.305
-6.370
-1.326
0.671
0.952
1.217
2.819
2.698
2.073
-2.509
-0.877
-0.720
-0.233
-0.406
-0.263
RAF
0.260
0.151
0.005
0.005
0.429
0.768
0.746
0.646
0.916
0.432
FE
-0.067
-0.036
0.388
0.159
-0.479
-0.227
1.539
0.607
-2.223
-0.606
2.607
0.860
-0.005
-0.001
-3.520
-1.450
OCE
1.749
0.457
-1.048
-0.628
-0.145
-0.103
-1.860
-1.808
-3.679
-1.311
2.251
1.008
0.753
0.963
COMB
1.640
0.443
-0.682
-0.649
0.627
0.515
-2.049
-0.411
-0.116
-0.112
aCIF price linkage equals Cjj and Year trend equals TRD.
^Region i across the top is the region importing from region j down the column.
cThe first line in each region represents parameter values.
dThe second line in each region represents the t Statistic.


249
Export Supplies
As shown in Table 6.1, three regions exported 87.8% of total world
fresh orange exports in the period considered. The regions were the
Mediterranean-EC, Middle East/North Africa, and the United States. The
sensitivity analysis will focus on these three regions.
The variables analyzed were the FOB average export price and fresh
orange production. Figures 6.5 and 6.6 present the sensitivity analyses
for export supply equations. Figure 6.5 shows export supplies while
changing the FOB average export price, and Figure 6.6 shows export
supplies while changing fresh production.
FOB average export price
Figure 6.5 has the FOB average export price index on the bottom axis
and the export supply index in the left axis. The expected economic
relationship between export supply and the FOB average export price is
positive. The figure shows that responses differ dramatically among
regions. Mediterranean-EC curve indicates that its response to changes in
the FOB average export price is close to zero. The response for the
Middle East/North Africa is positive as expected, but the one for the
United States is negative. The empirical results show that the "t"
statistic for the United States is one, indicating that it is significant
(see Table 5.2). The insignificance of the parameter obtained for
Mediterranean-EC and the sign of the parameter for the United States show
that the FOB average export price is not a major factor for export


23
exported. Per capita consumption of fresh oranges within Brazil is quite
low. The Mediterranean countries are major producers and exporters of
fresh oranges. The rest of Europe as well as Canada have always been net
importers of fresh and processed oranges. Other regions, especially the
Middle East/North Africa and the Far East, have increasingly become
important producers and traders in the orange industry. Chapter 2 will
outline production and trade flows of the fresh orange industry in more
detail.
Problem and Objectives
World consumption trends indicate that consumers are interested in
healthy and natural products. Fresh product consumption is increasing and
its potential growth is promising. Given the changes in consumption
patterns and the improvement in the transportation systems, studying the
fresh markets is of increasing importance. Fresh oranges, in particular,
provide consumers with natural flavor and important vitamins and minerals.
Fresh oranges and FCOJ are direct substitutes in the supply decision
process, but are not considered substitutes in the consumption side.
Consumer satisfaction is considered to be different for each good. Most
recent literature has studied mainly processed trade, which has been
growing faster than fresh trade in the last two decades. However, the
value and quantity of fresh trade are two and four times that for the
processed trade, respectively (see Tables 1.8 to 1.15). World fresh
utilization represents 65% of total orange production.


FRML EQ7#19 EXPORT7 -
FRML EQ8//19 EXPORT8 -
FRML EQ9//19 EXPORT9 -
FRML EQ10//19 EXPORTIO
DH210*LOG(PRD10));
FRML EQ11//19 EXPORT11
DH211*LOG(PRDll));
EXP(DH07 + DH17*LOG(REPD7) + DH27*LOG(PRD7))
EXP(DH08 + DH18*LOG(REPD8) + DH28*LOG(PRD8))
EXP(DH09 + DH19*LOG(REPD9) + DH29*LOG(PRD9))
= EXP(DH010 + DH110*LOG(REPD10) +
= EXP(DH011 + DH11l*LOG(REPD11) +
FRML EQ1//20 IQ1D
+ RH31*LOG(POPl)
FRML EQ2#20 IQ2D
+ RH32*LOG(POP2)
FRML EQ3#20 IQ3D
+ RH33*LOG(POP3)
FRML EQ4//20 IQ4D
+ RH34*LOG(P0P4)
FRML EQ5#20 IQ5D
+ RH35*LOG(POP5)
FRML EQ6//20 IQ6D
+ RH36*LOG(POP6)
FRML EQ7#20 IQ7D
+ RH37*LOG(POP7)
FRML EQ8//20 IQ8D
+ RH38*LOG(POP8)
FRML EQ9#20 IQ9D
+ RH39*LOG(POP9)
FRML EQ10#20
RH210*LOG(GDP10/CPI10) +
FRML EQ11#20 IQ11D
EXP(RH01 + RH11*L0G(RMP1D)
RH41*L0G(BAVAL1/CPI1));
EXP(RH02 + RH12*LOG(RMP2D)
RH42*LOG(BAVAL2/CPI2));
EXP(RH03 + RH13*LOG(RMP3D)
RH43*LOG(BAVAL3/CPI3));
EXP(RH04 + RH14*LOG(RMP4D)
RH44*L0G(BAVAL4/CPI4));
EXP(RH05 + RH15*LOG(RMP5D)
RH45*LOG(BAVAL5/CPI5));
EXP(RH06 + RH16*LOG(RMP6D)
RH46*LOG(BAVAL6/CPI6));
EXP(RH07 + RH17*LOG(RMP7D)
RH47*LOG(BAVAL7/CPI7));
EXP(RH08 + RH18*LOG(RMP8D)
RH48*LOG(BAVAL8/CPI8));
EXP(RH09 + RH19*LOG(RMP9D)
RH49*LOG(BAVAL9/CPI9));
IQIOD EXP(RH010
RH310*LOG(POP10) +
EXP(RH011
+ RH21*L0G(GDP1/CPI1)
+ RH22*LOG(GDP2/CPI2)
+ RH23*LOG(GDP3/CPI3)
+ RH24*LOG(GDP4/CPI4)
+ RH25*LOG(GDP5/CPI5)
+ RH26*LOG(GDP6/CPI6)
+ RH27*LOG(GDP7/CPI7)
+ RH28*LOG(GDP8/CPI8)
+ RH29*LOG(GDP9/CPI9)
+ RH110*LOG(RMP10D)
RH410*LOG(BAVAL10/CPI10)) ;
+ RH111*L0G(RMP11D)
RH211*LOG(GDPl1/CPI11) + RH311*LOG(POPll) + RH411*L0G(BAVAL11/CPI11)) ;
XRMP1D-RMP1D;
XRMP2D-RMP2D;
XRMP3D=RMP3D;
XRMP4D-RMP4D;
XRMP5D-RMP5D;
XRMP6D-RMP6D;
XRMP7D-RMP7D;
XRMP8D-RMP8D;
XRMP9D-RMP9D;
XRMP10D=RMP10D;
XRMP11DRMP11D;
XPOP1-POP1;
XPOP2-POP2;
XPOP3=POP3;
XP0P4-P0P4;
XPOP5-POP5;
XPOP6=POP6;
XPOP7=POP7;
XPOP8=POP8;
XPOP9=POP9;
XPOPIO-POPIO;
XPOPll=POPll;
XGDP1-GDP1;
XGDP2-GDP2;
XGDP3-GDP3;
XGDP4-GDP4;
XGDP5=GDP5;
XGDP6-GDP6;
XGDP7-GDP7;
XGDP8-GDP8;
XGDP9=GDP9;
XGDP10GDP10;
XGDP11GDP11;
XREPD1*=REPD1
XREPD2-REPD2
XREPD3-REPD3
XREPD4-REPD4
XREPD5-REPD5
XREPD6-REPD6
XREPD7-REPD7
XREPD8-REPD8
XREPD9-REPD9
XPRD1=PRD1
XPRD2-PRD2
XPRD3=PRD3
XPRD4-PRD4
XPRD5-PRD5
XPRD6-PRD6
XPRD7-PRD7
XPRD8=PRD8
XPRD9-PRD9


6
perspective of two different goods, fresh and processed oranges. To
improve our understanding and ease the following analysis of the fresh
orange industry, Tables 1.1 to 1.15 show 11 regions of significant trade.
These regions have been selected by considering similarities in supply or
demand among the countries and their importance in the production of and
international trade in fresh oranges. The regions identified are the
United States, Canada, Latin America, Mediterranean-European Community
countries, the rest of the European Community (EC), the rest of Western
Europe, Middle East/North Africa countries, the rest of Africa, the Far
East, Oceania and the Communist Bloc. The Communist Bloc is defined as it
existed prior to the recent political changes of 1991. Cuba is included
in the Bloc given the existance of trade agreements with Eastern Europe.
Appendix A shows the composition of these regions.
As shown in Table 1.1, world orange production increased at a rate
of 3.3% a year from 1966 to 1986. Table 1.2 shows that world fresh
utilization increased at a rate of 2.6% a year for the same period. The
processed industry increased faster than fresh utilization in the last
decade. From 1978 to 1986, world processed production increased at a rate
of 2.7% a year while fresh utilization increased 2.4% a year (see Tables
1.3 and 1.2).
Fresh orange world trade increased by 2.2% a year from 1966 to 1986,
and 1.1% from 1978 to 1986 (see Tables 1.4 and 1.6). If intraregional
trade or trade between countries of the same region is not considered,
international trade in fresh oranges showed an increase of 1.9% a year for
the same period (see Tables 1.8 and 1.12). This percentage is higher than
1.1%, meaning that trade among regions increased in the last decade.


331
XPM1_5-PM1_5;
XPM1_6-PM1_6;
XPM1_7-PM1_7;
XPM1_8-PM1_8;
XPM1_9=PM1_9;
XPM1_10=PM1_10;
XPM1_11PM1_11;
XIQ1D-IQ1D;
? SIMULATION #1 -
SMPL 1,1;
i-.5;
SMPL 2,7;
i-K-U+.l;
SMPL 1,7;
PM1_2-XPM1_2*I; PM1_3-XPM1_3*I; PM1_4=XPM1_4*I; PM1_5=XPM1_5*I;
PM1_6-XPM1_6*I;PM1_7-XPM1_7*I;PM1_8=XPM1_8*I;PM1_9=XPM1_9*I;
PM1_10=XPM1_10*I;PM1_11-XPM1_11*I;
EP2_1=XEP2_1*I;EP3_1-XEP3_1*I;EP4_1=XEP4_1*I;EP5_1=XEP5_1*I;
EP6_1=XEP6_1*I;EP7_1=XEP7_1*I;EP8_1-XEP8_1*I;EP9_1=XEP9_1*I;
EP10_1-XEP10_1*I;EP11_1-XEP11_1*I;
GENR EQ1#50;GENR EQ1#22;GENR EQ1#24;GENR EQ1#26;GENR EQ1#28;
GENR EQiy/30; GENR EQ1#32;GENR EQ1#34;GENR EQ1#36;GENR EQ1#38;
GENR EQ1#21;GENR EQ1#23;GENR EQ1//25;GENR EQ1#27 ;GENR EQ1#29;
GENR EQ1#31;GENR EQ1//33 ;GENR EQ1#35;GENR EQ1#37;GENR EQ1#39;
WRITE (FORMAT-LOTUS,FILE-'C:\L0TUS\SIM1#1.WK1')
I PM1_2 PM1_3 PM1_4 PM1_5 PM1_6 PM1_7 PM1_8 PM1_9
PM1_10 PM1_11
IQ1_2 IQ1_3 IQ1_4 IQ1_5 IQ1_6 IQ1_7 IQ1_8 IQ1_9
IQ1_10 IQ1_11
PM1_2-XPM1_2 ; PM1_3=XPM1_3 ; PM1_4-XPM1_4; PM1_5-XPM1_5 ;
PM1_6-XPM1_6 ; PM1_7-XPM1_7 ; PM1_8-XPM1_8 ; PM1_9-XPM1_9 ;
PM1_10=XPM1_10;PM1_11XPM1_11;
XEP5_1EP5_1;
XEP6_1EP6_1;
XEP7_1EP7_1;
XEP8_1EP8_1;
XEP9_1EP9_1;
XEP10_1=EP10_1;
XEP11_1=EP11_1;
PRODUCT DEMAND VARYING RELATIVE PRICES;
EP2_1=XEP2_1;EP3_1=XEP3_1;EP4_1=XEP4_1;EP5_1-XEP5_1;
EP6_1-XEP6_1;EP7_1=XEP7_1;EP8_1=XEP8_1;EP9_1-XEP9_1;
EP10_1=XEP10_1;EP11_1XEP11_1;
? SIMULATION y/2 PRODUCT DEMAND VARYING TOTAL MARKET DEMAND OR MARKET
SIZE;
SMPL 1,1;
i-.5;
SMPL 2,7;
I-K-D + .l;


Tab]o 2.2 World Fresh Utilization by Region
Region
1966
1976
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
--- (000)
Metric
Tons
Percent
of Change
United States
2575
2294
2322
-0.5
0.1
Canada
0
0
0
N.A.
N.A.
Latin America
5290
8336
9394
2.9
1.2
Mediterranean-EC
3689
4807
5787
2.3
1.9
E.C.
4
31
34
11.3
0.9
Rest of Western Europe
0
0
0
N.A.
N.A.
Middle East/North Africa
2862
4273
5122
3.0
1.8
Rest of Africa
692
894
919
1.4
0.3
Far East
3801
5634
7619
3.5
3.1
Oceania
202
196
251
1.1
2.5
Communist Bloc
192
287
638
6.2
8.3
World Total
19307
26752
32086
2.6
1.8


270
PRODUCT DEMAND INDEX (1986=1)
0.7 0.8 0.9 1 1.1 1.2 1.3
IMPORT PRICE INDEX (1986 = 1)
Figure 6.15. EC Imports Changing Import Prices.


70
al. (1965) is that it is possible to argue that consumers actually see
products of the same kind coming from different regions as non-perfect
substitutes.
Weisenborn et al. (1970) estimated the price-quantity relationships
at the processor or packer FOB level in foodstore, institutional, and
export market channels for Florida oranges and orange products. The
products included fresh and processed oranges. As reported by Weisenborn
et al., virtually no previous demand analysis had been completed for the
institutional and export sectors at that time.
Prato (1970) used the concept of separability to separate food from
non-food items. Once the demand equation was defined for only food items,
he showed that the correlation between first differences in the prices of
orange products and first differences in the prices of each of the other
food items were not significantly different from zero. Therefore,
individual demand equations for fresh and processed oranges without the
introduction of other food item prices could be defined. As reported by
Prato, research findings appear reasonable when compared with estimates
derived using other and more conventional approaches.
Tang (1977) studied the world demand for United States fresh
grapefruit in four markets: the United States, Japan, Europe, and Canada.
In his research, Tang identified and measured the effects of the different
factors that affect domestic and export demand in order to determine the
optimal allocation of United States fresh grapefruit to the domestic and
export markets. The results were used to simulate the grapefruit industry


13
Table 1.7 World Processed Orange Imports by Region
Region
1978
1986
Annual
Growth
Rate
1978-86
(000) 65 Degree
Brix Metric Tons
Percent
of Change
United States
136.8
500.6
17.6
Canada
98.3
89.5
-1.2
Latin America
6.6
7.1
0.9
Mediterranean-EC
4.0
16.7
19.6
E.C.
324.6
568.4
7.3
Rest of Western Europe
69.8
49.6
-4.2
Middle East/North Africa
14.1
10.0
-4.2
Rest of Africa
2.0
1.6
-2.4
Far East
11.7
29.4
12.3
Oceania
1.0
3.7
18.1
Communist Bloc
5.1
4.1
-2.7
World Total
673.8
1280.7
8.4
Source: United Nations Trade Data Tapes.


MULTIPLE-REGION EQUILIBRIUM WORLD TRADE
MODEL: THE ORANGE INDUSTRY
By
ESTEBAN R. BRENES
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFTLT.MF.NT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1992
UNIVERSITY OF FLORIDA LIBRARIES

ACKNOWLEDGEMENTS
I wish to give particular thanks to Dr. Ronald W. Ward,
chairman of my doctoral committee, for the guidance, assistance and
constructive criticism given to me during the preparation of this
dissertation. Similarly, I would like to express my gratitude to the
other members of my advisory committee--Drs. Kenneth R. Tefertiller, James
L. Seale, Jr., Max R. Langham, and Terry L. McCoy, for their advice and
support in the preparation of this document. Special thanks are given to
Sharon Bullivant and Melissa Bracewell in helping me put the dissertation
in its final form for the Graduate School. Special thanks are also given
to Bill Messina for his support and encouragement during the preparation
of the document.
I also appreciate the funding I received from the the United States
government through the Fulbright Scholarship Program. I want to thank my
friends from Liberty Church of Gainesville for their support,
understanding, and prayers. I give special thanks to my parents for they
taught me to appreciate education and always gave me the necessary support
to accomplish my academic achievements.
Finally, to my wife and son, I owe the most gratitude. They were
patient, understanding, and supportive all these years, and thanks to them
I finished this document. So, I would like to dedicate this dissertation
to my wife and son.
ii

TABLE OF CONTENTS
page
ACKNOWLEDGEMENTS ii
LIST OF TABLES vi
LIST OF FIGURES x
ABSTRACT xv
CHAPTER
1 INTERNATIONAL TRADE AND AGRICULTURAL PRODUCTS:
THE ORANGE INDUSTRY 1
Introduction 1
International Trade of Agricultural Products ... 3
Problem and Objectives 23
Scope 26
Methodology 27
Overview 28
2 FRESH ORANGE WORLD PRODUCTION AND TRADE 29
Introduction 29
Production Analysis 29
Trade Flow Analysis 34
Partner Region Perspective 39
Region Perspective 50
Conclusions 56
3 LITERATURE REVIEW 58
International Agricultural Trade Models 58
Trade Models: The Orange Industry 68
4 WORLD FRESH ORANGE TRADE MODEL 75
Introduction 75
Fresh Orange Trade Model 77
Demand Side 77
Supply Side 83
Export Supply Equations 84
Equilibrium Conditions 85
iii

Price Linkage Equations 85
CRES Model Restrictions 88
Model Implications 92
Trade Data Base 93
5 ECONOMETRIC PROCEDURE AND EMPIRICAL RESULTS 98
Introduction 98
Econometric and Estimation Procedure 98
Empirical Results and Implications 107
Empirical Results 107
Empirical Results: Graphical Analysis 134
Empirical Results: Statistical Analysis 190
Empirical Results: Economic Analysis 195
Application for Policy Purposes 222
Conclusion 225
6 ECONOMIC IMPLICATIONS FROM SENSITIVITY ANALYSIS .... 226
Introduction 226
Sensitivity Analysis Procedure 228
Rationale for Region, Equation, and
Variable Selections 232
Sensitivity Analysis 235
Total Market Demands 237
Export Supplies 249
Product Demands 253
Summary 290
Conclusions 294
7 SUMMARY AND CONCLUSIONS 296
Introduction 296
Data Limitations 298
Estimation and Sensitivity Analysis Difficulties 299
Performance of the Model and Results 300
Contributions to Agricultural Economics
Research 302
Further Research 303
APPENDIX A COUNTRY COMPOSITION OF THE REGIONS 305
APPENDIX B DERIVATION OF THE PRODUCT DEMAND EQUATIONS 307
APPENDIX C PROCEDURE TO OBTAIN REGIONAL CPIs 309
APPENDIX D PROCESSED ORANGE UTILIZATION 310
APPENDIX E TARIFF DATA 311
APPENDIX F PRINCIPAL COMPONENT PROCEDURE AND PROGRAM 312
iv

APPENDIX G ESTIMATION PROGRAM 313
APPENDIX H EMPIRICAL RESULTS: PRODUCT DEMAND AND CIF PRICE
LINKAGE EQUATIONS STATISTICS 321
APPENDIX I SENSITIVITY ANALYSIS PROGRAM 326
APPENDIX J INDICES OBTAINED FROM THE SENSITIVITY ANALYSIS .... 333
REFERENCES 339
BIOGRAPHICAL SKETCH 348
v

LIST OF TABLES
Table Page
1.1 World Orange Production by Region 7
1.2 World Fresh Utilization by Region 8
1.3 World Processed Production by Region 9
1.4 World Fresh Orange Exports by Region 10
1.5 World Processed Orange Exports by Region 11
1.6 World Fresh Orange Imports by Region 12
1.7 World Processed Orange Imports by Region 13
1.8 World Fresh Orange Export Quantities
by Region (Excluding Intraregional Trade) 14
1.9 World Fresh Orange Export Values by Region
(Excluding Intraregional Trade) 15
1.10 World Processed Orange Export Quantities
by Region (Excluding Intraregional Trade) 16
1.11 World Processed Orange Export Values by Region
(Excluding Intraregional Trade) 17
1.12 World Fresh Orange Import Quantities by Region
(Excluding Intraregional Trade) 18
1.13 World Fresh Orange Import Values by Region
(Excluding Intraregional Trade) 19
1.14 World Processed Orange Import Quantities by
Region (Excluding Intraregional Trade) 20
1.15 World Processed Orange Import Values by Region
(Excluding Intraregional Trade) 21
2.1 World Orange Production by Region 30
2.2 World Fresh Utilization by Region 32
vi

2.3 Trade Flow Analysis for Selected Years (1966,
1976 and 1986) by Region in Relation to Partner
Regions 35
2.4 Trade Flow Analysis for Selected Years (1966,
1976 and 1986) Without Intraregional Trade
"Relative Partner Region Exports by Region" 37
2.5 Trade Flow Analysis for Selected Years (1966,
1976 and 1986) Without Intraregional Trade
"Relative Region Imports from Partner Regions 38
2.6 Trade Flow Analysis for Selected Periods of Five
Years (1966-70,1974-78 and 1982-86) 40
2.7 Trade Flow Analysis for Selected Periods of Five
Years (1966-70,1974-78 and 1982-86) Without
Intraregional Trade "Relative Partner Region
Exports by Region" 42
2.8 Trade Flow Analysis for Selected Periods of Five
Years (1966-70,1974-78 and 1982-86) Without
Intraregional Trade "Relative Region Imports
from Partner Regions" 43
5.1 Total Market Demand Equations 108
5.2 Export Supply Equations 109
5.3 United States Product Demands 110
5.4 Canada Product Demands Ill
5.5 Latin America Product Demands 112
5.6 Mediterranean-EC Product Demands 113
5.7 EC Product Demands 114
5.8 Rest of Western Europe Product Demands 115
5.9 Middle East/North Africa Product Demands 116
5.10 Rest of Africa Product Demands 117
5.11 Far East Product Demands 118
5.12 Oceania Product Demands 119
5.13 Communist Bloc Product Demands 120
5.14 United States CIF Price Linkage Equations 121
vii

5.15 Canada CIF Price Linkage Equations 122
5.16 Latin America CIF Price Linkage Equations 123
5.17 Mediterranean-EC CIF Price Linkage Equations 124
5.18 EC CIF Price Linkage Equations 125
5.19 Rest of Western Europe CIF Price Linkage
Equations 126
5.20 Middle East/North Africa CIF Price Linkage
Equations 127
5.21 Rest of Africa CIF Price Linkage Equations 128
5.22 Far East CIF Price Linkage Equations 129
5.23 Oceania CIF Price Linkage Equations 130
5.24 Communist Bloc CIF Price Linkage Equations 131
5.25 Total Market Demand and Export Supply Equations
Statistics 193
5.26 Market Demand and Export Supply Equations
Elasticities 197
5.27 Product Demands Relative Price Elasticities 204
5.28 Product Demands Total Market Demand
Elasticities 205
5.29 CIF Price Linkage FOB Export Price Elasticities .... 217
5.30 CIF Price Linkage Year Trend Elasticity 218
5.31 CIF Price Linkage Index Price for Energy
Elasticity 219
6.1 World Demand, Imports and Exports Share Per
Region (Cumulative 21 Year Period 1966-1986) 238
6.2 Region's Relative Imports Per Partner Region (%
Cumulative 21 Year Period 1966-1986) 255
H.l United States, Canada and Latin America Product
Demand and CIF Price Linkage Equations
Statistics 322
viii

H.2 Mediterranean-EC, EC, Rest of Western Europe
Product Demand and CIF Price Linkage Equations
Statistics 323
H.3 Middle East/North Africa, Rest of Africa and Far
East Product Demand and CIF Price Linkage
Equations Statistics 324
H.4 Oceania and Communist Bloc Product Demand and
CIF Price Linkage Equations Statistics 325

LIST OF FIGURES
Figure Page
5.1 Total Market Demand for Fresh Oranges in the
United States 135
5.2 Total Market Demand for Fresh Oranges in Canada .... 136
5.3 Total Market Demand for Fresh Oranges in Latin
America 137
5.4 Total Market Demand for Fresh Oranges in the
Mediterranean-EC 138
5.5 Total Market Demand for Fresh Oranges in the EC ... 139
5.6 Total Market Demand for Fresh Oranges in the
Rest of Western Europe 140
5.7 Total Market Demand for Fresh Oranges in the
Middle East/North Africa 141
5.8 Total Market Demand for Fresh Oranges in the
Rest of Africa 142
5.9 Total Market Demand for Fresh Oranges in the Far
East 143
5.10 Total Market Demand for Fresh Oranges in
Oceania 144
5.11 Total Market Demand for Fresh Oranges in the
Communist Bloc 145
5.12 Total Export Supply of Fresh Oranges from the
United States 146
5.13 Total Export Supply of Fresh Oranges from
Canada 147
5.14 Total Export Supply of Fresh Oranges from Latin
America 148
x

5.15 Total Export Supply of Fresh Oranges from the
Mediterranean-EC 149
5.16 Total Export Supply of Fresh Oranges from EC 150
5.17 Total Export Supply of Fresh Oranges from the
Rest of Western Europe 151
5.18 Total Export Supply of Fresh Oranges from Middle
East/North Africa 152
5.19 Total Export Supply of Fresh Oranges from the
Rest of Africa 153
5.20 Total Export Supply of Fresh Oranges from the
Far East 154
5.21 Total Export Supply of Fresh Oranges from
Oceania 155
5.22 Total Export Supply of Fresh Oranges from the
Communist Bloc 156
5.23 United States Imports of Fresh Oranges from
Latin America (Product Demand 1_3) 163
5.24 United States Imports of Fresh Oranges from the
Middle East/North Africa (Product Demand 1_7) 164
5.25 Canada Imports of Fresh Oranges from the United
States (Product Demand 2_1) 165
5.26 Canada Imports of Fresh Oranges from the Middle
East/North Africa (Product Demand 2_7) 166
5.27 Canada Imports of Fresh Oranges from the Far
East (Product Demand 2_9) 167
5.28 Latin America Imports of Fresh Oranges from the
United States (Product Demand 3_1) 168
5.29 Latin America Imports of Fresh Oranges from the
EC (Product Demand 3_5) 169
5.30 Mediterranean-EC Imports of Fresh Oranges from
Latin America (Product Demand 4_3) 170
5.31 Mediterranean-EC Imports of Fresh Oranges from
the EC (Product Demand 4_5) 171
5.32 EC Imports of Fresh Oranges from the
Mediterranean-EC (Product Demand 5_4) 172
xi

5.33 EC Imports of Fresh Oranges from the Middle
East/North Africa (Product Demand 5_7) 173
5.34 EC Imports of Fresh Oranges from the Rest of
Africa (Product Demand 5_8) 174
5.35 Rest of Western Europe Imports of Fresh Oranges
from Mediterranean-EC (Product Demand 6_4) 175
5.36 Rest of Western Europe Imports of Fresh Oranges
from the Middle East/North Africa (Product
Demand 6_7) 176
5.37 Rest of Western Europe Imports of Fresh Oranges
from the Rest of Africa (Product Demand 6_8) 177
5.38 Middle East/North Africa Imports of Fresh
Oranges from Latin America (Product Demand 7_3) .... 178
5.39 Middle East/North Africa Imports of Fresh
Oranges from the Rest of Africa (Product Demand
7_8) 179
5.40 Middle East/North Africa Imports of Fresh
Oranges from the Far East (Product Demand 7_9) .... 180
5.41 Rest of Africa Imports of Fresh Oranges from the
EC (Product Demand 8_5) 181
5.42 Rest of Africa Imports of Fresh Oranges from the
Middle East/North Africa (Product Demand 8_7) 182
5.43 Far East Imports of Fresh Oranges from the
United States (Product Demand 9_1) 183
5.44 Far East Imports of Fresh Oranges from the
Middle East/North Africa (Product Demand 9_7) 184
5.45 Far East Imports of Fresh Oranges from Oceania
(Product Demand 9_10) 185
5.46 Oceania Imports of Fresh Oranges from the United
States (Product Demand 10_1) 186
5.47 Oceania Imports of Fresh Oranges from the Middle
East/North Africa (Product Demand 10_7) 187
5.48 Communist Bloc Imports of Fresh Oranges from the
Mediterranean-EC (Product Demand 11_4) 188
5.49 Communist Bloc Imports of Fresh Oranges from the
Middle East/North Africa (Product Demand 11_7) .... 189
xii

6.1 Total Market Demand Changing Average Market
Price (Major World Consumers) 239
6.2 Total Market Demand Changing Average Market
Price (Major World Importers) 240
6.3 Total Market Demand Changing Income (GDP) (Major
World Consumers) 241
6.4 Total Market Demand Changing Income (GDP) (Major
World Importers) 242
6.5 Export Supply Changing FOB Average Export Price
(Major World Exporters) 250
6.6 Export Supply Changing Fresh Production (Major
World Exporters) 251
6.7 United States Imports Changing Import Prices 256
6.8 United States Imports Changing Total Market
Demand 257
6.9 Canada Imports Changing Import Prices 260
6.10 Canada Imports Changing Total Market Demand 261
6.11 Latin America Imports Changing Import Prices 264
6.12 Latin America Imports Changing Total Market
Demand 265
6.13 Mediterranean-EC Imports Changing Import Prices .... 267
6.14 Mediterranean-EC Imports Changing Total Market
Demand 268
6.15 EC Imports Changing Import Prices 270
6.16 EC Imports Changing Total Market Demand 272
6.17 Rest of Western Europe Imports Changing Import
Prices 274
6.18 Rest of Western Europe Imports Changing Total
Market Demand 275
6.19 Middle East/North Africa Imports Changing Import
Prices 277
xiii

6.20 Middle East/North Africa Imports Changing Total
Market Demand 278
6.21 Rest of Africa Imports Changing Import Prices 280
6.22 Rest of Africa Imports Changing Total Market
Demand 281
6.23 Far East Imports Changing Import Prices 283
6.24 Far East Imports Changing Total Market Demand 285
6.25 Oceania Imports Changing Import Prices 287
6.26 Oceania Imports Changing Total Market Demand 288
6.27 Communist Bloc Imports Changing Import Prices 289
6.28 Communist Bloc Imports Changing Total Market
Demand 291
xiv

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
MULTIPLE-REGION EQUILIBRIUM WORLD TRADE MODEL:
THE ORANGE INDUSTRY
By
Esteban R. Brenes
May, 1992
Chairman: Dr. Ronald W. Ward
Major Department: Food and Resource Economics
A multiple-region equilibrium trade model for the fresh orange
industry including 11 regions of the world was developed and estimated.
The model is used to understand the major driving factors affecting fresh
orange consumption and trade. The model is a modified spatial equilibrium
model that takes into account that products are differentiated by country
of origin. Armington developed the demand theory underlying this
assumption. The model assumes a constant ratio of elasticity of
substitution (CRES) index which makes the model somewhat less restrictive.
The model is estimated using a nonlinear two stage least square
procedure. Graphical, statistical and economic analyses of the results
are used to evaluate the performance of the model and the implications for
the fresh orange industry. The results indicate that the model performs
well. A sensitivity analysis of the system is developed to evaluate the
consequences of changes in the main variables of the model.
xv

Total market demand analysis shows that market prices and income are
the major drivers for world consumption of fresh oranges in most regions.
Major world importers are more sensitive to changes in the average market
price than major world consumers with domestic production. Export supply
equations show weak FOB export price parameters versus strong fresh
production parameters. This indicates that major export decisions are
driven mainly by fresh production.
Product demand analysis shows the role of prices as an allocative
tool and the importance of the market size to determine consumer
preferences when facing several product sources. Market positions and
opportunities for all regions were determined. The regions included the
United States, Canada, South America, the Mediterranean-EEC countries, the
rest of the EEC, the rest of western Europe, the Middle East and North
Africa countries, the rest of Africa, the Far East, Oceania, and the
Communist Bloc. The Communist Bloc was defined as it existed prior to the
recent political changes of 1991.
xvi

CHAPTER 1
INTERNATIONAL TRADE AND AGRICULTURAL PRODUCTS:
THE ORANGE INDUSTRY
Introduction
Developing countries have long recognized the importance of trade to
their national welfare. Exchanging the goods that they produce with their
endowments and experience for the goods in which other countries have a
comparative advantage provides the potential for both growth and
development. The world is becoming smaller in terms of communication and
the ability to trade. International trade has been expanding at an
increasing rate, especially in the last three decades. Most countries are
dependent to some degree on the foreign currency generated through trade.
Some countries consider international trade as one of the most important
means for development, especially the ones that have a high foreign debt.
The United States continues to implement important macro-economic and
commercial policies to improve its competitive position in the
international trade arena.
Less developed countries use international trade as a means for
development and subsistence because their domestic economies are poor and
small in most cases. Therefore, it is very important to develop new
markets for their products as well as to develop new marketable products.
Most of these nations possess a high foreign debt that must be paid with
1

2
foreign currency. Imports to these countries are frequently higher than
exports, which implies a need for more foreign currency, i.e., they are
often net importers.
In the 1980s, many less developed countries changed their
development strategy from an import substitution scheme to export
promotion. While the reasons for change differ among countries, the main
reasons include interruption of multilateral and bilateral agreements, the
increase of the fiscal deficit due to subsidies, recognition of
inefficiencies due to high import barriers, and the need of foreign
currency.
Developed countries such as Japan, Germany, Italy, and France are
highly interdependent on international trade. The United States also
recognizes the importance of being competitive and the need of
interdependence with potential trade partners. These changes increased
the importance of international trade and market development worldwide
with the consequence being an increase in the relative importance of
international trade.
In theory, international trade is dependent on comparative
advantage. This means that countries will tend to trade goods and
services that they produce efficiently for goods and services in which
other countries have a comparative advantage. Porter (1990) mentions four
broad attributes to shape the environment in which local firms compete in
order to achieve international success in a particular industry. First,
factor conditions, which mean the nation's position in factors of
production necessary to compete in a given industry; second, demand
conditions which consider the natures of domestic demand for the

3
industry's product or service; third, related and supporting industries
which refer to the presence or absence in the nation of supplier and
related industries that are internationally competitive; fourth, firm
strategy, structure, and rivalry which refer to the conditions of the
nation governing how companies are created, organized, and managed, and
the nature of rivalry. However, in the real world, international trade is
not solely driven on comparative advantage but also on variables including
tariffs, quotas, subsidies, international agreements, and domestic
policies.
World real-value trade increased at an average rate of 6.7% a year
while real Gross National Product (GNP) increased at an average rate of
4.1% from 1966 to 1986 (International Monetary Fund [IMF] Direction of
Trade). During that period the United States economy became increasingly
interdependent with world economies. The value of United States total
trade as a percentage of GNP increased from 7.5 to 14 for the same period.
International Trade of Agricultural Products
International agricultural trade depends heavily on national and
regional policies. For agricultural products, import protection and
export subsidies are usual in most countries; especially in cases where a
particular product is socially desirable and support groups are
politically strong. For example, grain production is usually protected
from imports in most countries. Reasons given to support such policies
have to do with income distribution given the amount of people involved in
production, process and distribution of grains; self-sufficiency; and less

4
need of foreign currency. In many cases, the government has to support
production through higher prices for these people to continue in business.
This implies higher fiscal deficits, given subsidies and underpriced
exports to get rid of excess inventories. This promotes inefficiencies
and a waste of resources.
A controversial agricultural policy is the Common Agricultural
Policy (CAP) of the European Community (EC). The CAP is primarily a
market-regulation and price-support policy. It currently covers grains,
rice, poultry and eggs, dairy products, pork, beef and veal, sugar,
certain fruits and vegetables, and certain processed agricultural
products. The protectionism system includes common customs tariffs for
imports and internal regulations designed to protect EC producers. The
system gradually eliminates trade barriers among nations within the bloc
but imposes a common external trade barrier. The variable levy system is
the major instrument used by the EC to protect domestic markets from
foreign competition. It imposes a levy equal to the difference between
the world price and the domestic support price. This tends to make EC
imports from other countries have a perfectly inelastic demand within a
considerable price range, with outside countries the residual suppliers.
Some of the proceeds from the levy are further used to subsidize EC
exports (Tweeten, 1979). The CAP regulations in fresh fruits set quality
standards for a variety of products and outline a price and intervention
system in most cases.
Tariffs and nontariff barriers, along with preferential treatment,
have become increasingly important factors influencing agricultural trade.
In the case of fresh fruits, CAP regulations are of particular importance

5
with the recent enlargement of the EC to 12 nations with the inclusion of
Spain and Portugal. Both are major producers and exporters of fresh
fruits to the rest of Europe.
Other factors affecting agricultural trade are income, population,
demographic variables within the trading regions, and exchange rates.
Income and population are important to determine the level of consumption.
As these two variables increase, higher levels of consumption and shifts
from one bundle of goods to another are expected. Exchange rates affect
the real terms of trade among countries, especially in cases when they are
managed by governments and are not allowed to fluctuate freely in the
market. Transportation is an important linkage variable in world trade.
The linkage is between the Free On Board (FOB) export price from any
country and the Cost Insurance Freight (CIF) import price at the final
market. Substitute product prices should have a positive effect on the
consumption of a specific commodity.
Total agricultural trade increased at an annual rate of 2.8% while
agricultural production increased by 2.3% a year from 1966 to 1986 (Food
and Agricultural Organization [FAO] Trade and Production Yearbooks). The
United States agricultural sector represents about 15% of total exports.
Hence, United States agricultural market prices are strongly influenced by
supply and demand conditions among major world markets (Statistical
Abstract of the United States and FAO Trade Yearbook). These statistics
reflect the increasing importance of trade in the world economy and, in
particular, the importance of agriculture in world trade.
The focus of the present study will be the fresh orange industry.
World trade in the fresh orange industry must be studied from the

6
perspective of two different goods, fresh and processed oranges. To
improve our understanding and ease the following analysis of the fresh
orange industry, Tables 1.1 to 1.15 show 11 regions of significant trade.
These regions have been selected by considering similarities in supply or
demand among the countries and their importance in the production of and
international trade in fresh oranges. The regions identified are the
United States, Canada, Latin America, Mediterranean-European Community
countries, the rest of the European Community (EC), the rest of Western
Europe, Middle East/North Africa countries, the rest of Africa, the Far
East, Oceania and the Communist Bloc. The Communist Bloc is defined as it
existed prior to the recent political changes of 1991. Cuba is included
in the Bloc given the existance of trade agreements with Eastern Europe.
Appendix A shows the composition of these regions.
As shown in Table 1.1, world orange production increased at a rate
of 3.3% a year from 1966 to 1986. Table 1.2 shows that world fresh
utilization increased at a rate of 2.6% a year for the same period. The
processed industry increased faster than fresh utilization in the last
decade. From 1978 to 1986, world processed production increased at a rate
of 2.7% a year while fresh utilization increased 2.4% a year (see Tables
1.3 and 1.2).
Fresh orange world trade increased by 2.2% a year from 1966 to 1986,
and 1.1% from 1978 to 1986 (see Tables 1.4 and 1.6). If intraregional
trade or trade between countries of the same region is not considered,
international trade in fresh oranges showed an increase of 1.9% a year for
the same period (see Tables 1.8 and 1.12). This percentage is higher than
1.1%, meaning that trade among regions increased in the last decade.

Table 1.1 World Orange Production by Region
Region
1966
1976
1978
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
Annual
Growth
Rate
1978-86
(000) Metric Tons
Percent of Change
United States
7598
10183
9268
7192
-0.3
-3.4
-3.1
Canada
0
0
0
0
N.A.
N.A.
N.A.
Latin America
5540
12117
11832
18535
6.2
4.3
5.8
Mediterranean-EC
4208
5472
5267
6840
2.5
2.3
3.3
E.C.
4
31
29
34
11.3
0.9
2.0
Rest of Western Europe
0
00
0
0
N.A.
N.A.
N.A.
Middle East/North Africa
3067
4664
5364
5794
3.2
2.2
1.0
Rest of Africa
773
1034
1132
1022
1.4
-0.1
-1.3
Far East
4023
6532
6771
8354
3.7
2.5
2.7
Oceania
249
360
410
574
4.3
4.8
4.3
Communist Bloc
192
292
408
728
6.9
9.6
7.5
World Total
25654
40685
40481
49073
3.3
1.9
2.4
Source: FAO Production Yearbook. Various issues.

Table 1.2 World Fresh Utilization by Region
Region
1966
1976
1978
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
Annual
Growth
Rate
1978-86
- (000) Metric Tons -
Percent of Changi
e
United States
2575
2294
2031
2322
-0.5
0.1
1.7
Canada
0
0
0
0
N.A.
N.A.
N.A.
Latin America
5290
8336
7342
9394
2.9
1.2
3.1
Mediterranean-EC
3689
4807
4600
5787
2.3
1.9
2.9
E.C.
4
31
29
34
11.3
0.9
2.0
Rest of Western Europe
0
0
0
0
N.A.
N.A.
N.A.
Middle East/North Africa
2862
4273
4943
5122
3.0
1.8
0.4
Rest of Africa
692
894
980
919
1.4
0.3
-0.8
Far East
3801
5634
5934
7619
3.5
3.1
3.2
Oceania
202
196
202
251
1.1
2.5
2.7
Communist Bloc
192
287
398
638
6.2
8.3
6.1
World Total
19307
26752
26459
32086
2.6
1.8
2.4
CO

9
Table 1.3 World Processed Production by Region
Region
1978
1986
Annual
Growth
Rate
1978-86
(000) 65 Degree
Brix Metric Tons
Percent
of Change
United States
732.5
481.3
-5.1
Canada
0.0
0.0
N. A.
Latin America
406.7
895.3
10.4
Mediterranean-EC
60.4
103.2
6.9
E.C.
0.0
0.0
N. A.
Rest of Western Europe
0.0
0.0
N. A.
Middle East/North Africa
38.1
65.8
7.1
Rest of Africa
13.8
10.1
-3.8
Far East
75.8
72.0
-0.6
Oceania
18.8
31.6
6.7
Communist Bloc
0.9
8.8
32.9
World Total
1347.1
1668.0
2.7

Table 1.4 World Fresh Orange Exports by Region
Region
1966
1976
1978
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
Annual
Growth
Rate
1978-86
(000) Metric Tons
Percent of Change
United States
258.6
464.1
355.9
413.0
2.4
-1.2
1.9
Canada
0.1
0.1
0
0.3
6.9
16.1
31.5
Latin America
106.6
119.2
164.7
229.3
3.9
6.8
4.2
Mediterranean-EC
1514.6
1937.5
1802.2
2833.9
3.2
3.9
5.8
E.C.
32.8
122.6
121.5
195.0
9.3
4.8
6.1
Rest of Western Europe
2.1
5.9
7.8
2.2
0.2
-9.5
-14.7
Middle East/North Africa
1215.9
1871.9
1893.2
1316.7
0.4
-3.5
-4.4
Rest of Africa
264.4
285.7
274.4
209.0
-1.2
-3.1
-3.3
Far East
65.8
142.3
137.5
113.9
2.8
-2.2
-2.3
Oceania
23.5
11.2
30.1
47.3
3.5
15.5
5.8
Communist Bloc
0.2
60.4
140.9
35.0
30.5
-5.3
-16.0
World Total
3484.5
5020.7
4928.3
5395.6
2.2
0.7
1.1
Source: United Nations Trade Data Tapes.
o

11
Table 1.5 World Processed Orange Exports by Region
Region
1978
1986
Annual
Growth
Rate
1978-86
(000) 65 Degree
Brix Metric Tons
Percent
of Change
United States
148.1
76.4
-7.9
Canada
0.5
2.8
23.1
Latin America
296.1
831.4
13.8
Mediterranean-EC
29.1
33.5
1.8
E.C.
97.6
207.2
9.9
Rest of Western Europe
3.3
7.1
10.1
Middle East/North Africa
95.9
112.3
2.0
Rest of Africa
2.3
1.4
-5.6
Far East
0.9
6.7
29.0
Oceania
0.1
1.5
45.3
Communist Bloc
.0
0.3
46.2
World Total
673.8
1280.7
8.4
Source: United Nations Trade Data Tapes.

Table 1.6 World Fresh Orange Imports by Region
Region
1966
1976
1978
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
Annual
Growth
Rate
1978-86
(000) Metric
Tons
Percent of Changi
e
United States
28.3
32.6
52.9
49.4
2.8
4.2
-0.8
Canada
180.5
225.5
180.2
182.1
.0
-2.1
0.1
Latin America
9.7
18.1
14.2
3.8
-4.6
-14.5
-15.2
Mediterranean-EC
0.0
0.9
1.5
8.2
N. A.
24.8
24.2
E.C.
2398.4
2736.2
2655.5
3464.6
1.9
2.4
3.4
Rest of Western Europe
384.6
439.7
436.0
551.2
1.8
2.3
3.0
Middle East/North Africa
22.0
590.4
440.5
279.0
13.5
-7.2
-5.5
Rest of Africa
10.6
10.6
14.5
8.2
-1.3
-2.5
-6.8
Far East
119.3
238.9
251.7
382.8
6.0
4.8
5.4
Oceania
15.3
8.9
15.5
20.4
1.4
8.7
3.5
Communist Bloc
315.6
718.8
865.8
445.8
1.7
-4.7
-8.0
World Total
3484.5
5020.7
4928.3
5395.6
2.2
0.7
1.1
Source: United Nations Trade Data Tapes.

13
Table 1.7 World Processed Orange Imports by Region
Region
1978
1986
Annual
Growth
Rate
1978-86
(000) 65 Degree
Brix Metric Tons
Percent
of Change
United States
136.8
500.6
17.6
Canada
98.3
89.5
-1.2
Latin America
6.6
7.1
0.9
Mediterranean-EC
4.0
16.7
19.6
E.C.
324.6
568.4
7.3
Rest of Western Europe
69.8
49.6
-4.2
Middle East/North Africa
14.1
10.0
-4.2
Rest of Africa
2.0
1.6
-2.4
Far East
11.7
29.4
12.3
Oceania
1.0
3.7
18.1
Communist Bloc
5.1
4.1
-2.7
World Total
673.8
1280.7
8.4
Source: United Nations Trade Data Tapes.

Table 1.8 World Fresh Orange Export Quantities by Region (Excluding Intraregional Trade)
Region
1966
1976
1978
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
Annual
Growth
Rate
1978-86
(000) Metric Tons
Percent of Change
United States
258.6
464.1
355.9
413.0
2.4
-1.2
1.9
Canada
0.1
0.1
.0
0.3
6.9
16.1
31.5
Latin America
103.4
103.1
152.7
226.9
4.0
8.2
5.1
Mediterranean-EC
1514.6
1937.3
1802.2
2833.7
3.2
3.9
5.8
E.C.
2.8
9.6
13.0
21.1
10.5
8.2
6.3
Rest of Western Europe
1.1
0.9
4.9
1.1
.0
2.6
-17.0
Middle East/North Africa
1205.3
1410.9
1524.9
1092.6
-0.5
-2.5
-4.1
Rest of Africa
262.7
280.2
268.6
206.9
-1.2
-3.0
-3.2
Far East
24.3
64.5
47.9
41.9
2.8
-4.2
-1.6
Oceania
13.7
8.7
21.2
36.7
5.1
15.5
7.1
Communist Bloc
0.2
2.0
4.2
12.4
23.9
19.8
14.6
World Total
3386.6
4281.3
4195.5
4886.6
1.9
1.3
1.9
Source: United Nations Trade Data Tapes.

Table 1.9 World Fresh Orange Export Values by Region (Excluding Intraregional Trade)
Region
1966
1976
1978
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
Annual
Growth
Rate
1978-86
In
Million of U.
, S. Dollars
Percent of Chang<
2
United States
47.0
118.8
144.0
233.0
8.3
7.0
6.2
Canada
.0
.0
.0
0.1
11.5
20.7
26.6
Latin America
6.2
14.7
27.9
61.5
12.1
15.4
10.4
Mediterranean-EC
178.9
406.1
530.0
1025.2
9.1
9.7
8.6
E.C.
0.7
2.8
5.7
12.6
15.2
16.1
10.3
Rest of Western Europe
0.2
0.3
1.3
0.6
5.4
9.2
-8.3
Middle East/North Africa
172.7
322.1
410.7
361.6
3.8
1.2
-1.6
Rest of Africa
35.7
59.1
103.6
83.5
4.3
3.5
-2.7
Far East
4.6
21.1
17.2
17.6
7.0
-1.8
0.2
Oceania
2.5
3.1
6.5
15.2
9.4
17.1
11.2
Communist Bloc
.0
0.4
0.8
3.9
28.7
25.0
21.9
World Total
448.7
948.6
1247.7
1814.7
7.2
6.7
4.8
Source: United Nations Trade Data Tapes.

16
Table 1.10 World Processed Orange Export Quantities by Region (Excluding
Intraregional Trade)
Region
1978
1986
Annual
Growth
Rate
1978-86
(000) 65 Degree
Brix Metric Tons
Percent
of Change
United States
148.1
76.4
-7.9
Canada
0.5
2.8
22.9
Latin America
296.1
827.8
13.7
Mediterranean-EC
28.9
33.3
1.8
E.C.
16.4
31.9
8.6
Rest of Western Europe
2.6
6.1
11.5
Middle East/North Africa
95.5
112.1
2.0
Rest of Africa
2.3
1.4
-5.7
Far East
0.5
2.2
19.4
Oceania
0.1
0.9
34.0
Communist Bloc
.0
0.3
39.4
World Total
591.0
1095.2
8.0
Source: United Nations Trade Data Tapes.

17
Table 1.11 World Processed Orange Export Values by Region (Excluding
Intraregional Trade)
Region
1978
1986
Annual
Growth
Rate
1978-86
In Millions of
U.S. Dollars
Percent
of Change
United States
98.0
66.6
-4.7
Canada
0.6
5.1
31.7
Latin America
288.6
671.8
11.1
Mediterranean-EC
23.4
32.9
4.4
E.C.
16.2
29.3
7.6
Rest of Western Europe
2.9
4.2
4.4
Middle East/North Africa
59.8
102.2
6.9
Rest of Africa
4.2
1.2
-14.2
Far East
0.3
1.5
23.0
Oceania
0.1
0.7
32.5
Communist Bloc
.0
0.2
52.4
World Total
494.1
915.8
8.0
Source: United Nations Trade Data Tapes.

Table 1.12 World Fresh Orange Import Quantities by Region (Excluding Intraregional Trade)
Region
1966
1976
1978
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
Annual
Growth
Rate
1978-86
(000) Metric Tons
Percent of Change
United States
28.3
32.6
52.9
49.4
2.8
4.2
-0.8
Canada
180.5
225.5
180.2
182.1
.0
-2.1
0.1
Latin America
6.5
2.0
2.2
1.4
-7.5
-4.0
-5.9
Mediterranean-EC
0.0
0.7
1.5
8.0
N. A.
27.3
23.7
E.C.
2368.5
2623.3
2546.9
3290.7
1.7
2.3
3.3
Rest of Western Europe
383.7
434.6
433.2
550.2
1.8
2.4
3.0
Middle East/North Africa
11.4
129.4
72.2
54.9
8.2
-8.2
-3.4
Rest of Africa
8.8
5.1
8.7
6.1
-1.8
1.8
-4.3
Far East
77.8
161.1
162.0
310.9
7.2
6.8
8.5
Oceania
5.5
6.5
6.7
9.8
3.0
4.3
4.9
Communist Bloc
315.6
660.5
729.1
423.2
1.5
-4.4
-6.6
World Total
3386.6
4281.3
4195.5
4886.6
1.9
1.3
1.9
Source: United Nations Trade Data Tapes.
00

Table 1.13 World Fresh Orange Import Values by Region (Excluding Intraregional Trade)
Region
1966
1976
1978
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
Annual
Growth
Rate
1978-86
In Millions of U.
S. Dollars
Percent of Change
United States
3.7
8.8
14.2
28.9
10.8
12.6
9.3
Canada
38.6
99.1
76.7
133.1
6.4
3.0
7.1
Latin America
1.0
0.8
2.6
1.0
0.2
2.6
-11.1
Mediterranean-EC
.0
0.2
0.8
4.5
N. A.
38.4
24.7
E.C.
405.3
751.0
966.2
1367.7
6.3
6.2
4.4
Rest of Western Europe
68.7
153.1
198.4
333.9
8.2
8.1
6.7
Middle East/North Africa
1.1
44.3
28.0
14.5
13.6
-10.5
-7.8
Rest of Africa
3.0
2.8
4.5
6.2
3.7
8.4
3.9
Far East
19.1
66.0
97.4
241.1
13.5
13.8
12.0
Oceania
1.5
2.2
3.0
6.7
7.6
12.1
10.9
Communist Bloc
43.9
168.8
246.9
203.9
8.0
1.9
-2.4
World Total
586.0
1297.0
1638.6
2341.5
7.2
6.1
4.6
Source: United Nations Trade Data Tapes.

20
Table 1.14 World Processed Orange Import Quantities by Region (Excluding
Intraregional Trade)
Region
1978
1986
Annual
Growth
Rate
1978-86
(000) 65 Degree
Brix Metric Tons
Percent
of Change
United States
136.8
500.6
17.6
Canada
98.3
89.5
-1.2
Latin America
6.5
3.4
-7.8
Mediterranean-EC
3.8
16.6
20.1
E.C.
243.4
393.0
6.2
Rest of Western Europe
69.1
48.6
-4.3
Middle East/North Africa
13.7
9.8
-4.1
Rest of Africa
1.9
1.6
-2.5
Far East
11.3
25.0
10.4
Oceania
1.0
3.0
15.4
Communist Bloc
5.2
4.2
-2.6
World Total
591.0
1095.2
8.0
Source: United Nations Trade Data Tapes.

21
Table 1.15 World Processed Orange
Intraregional Trade)
Import
Values by Region
(Excluding
Region
1978
1986
Annual
Growth
Rate
1978-86
In Millions of
U.S. Dollars
Percent
of Change
United States
150.6
518.9
16.7
Canada
106.0
104.4
-0.2
Latin America
5.9
8.1
3.9
Mediterranean-EC
3.7
16.7
20.6
E.C.
230.2
414.1
7.6
Rest of Western Europe
83.0
55.0
-5.0
Middle East/North Africa
12.5
16.1
3.2
Rest of Africa
2.3
2.1
-1.3
Far East
19.5
49.5
12.4
Oceania
1.4
3.5
12.1
Communist Bloc
3.3
3.4
0.5
World Total
618.4
1191.9
8.5
Source: United Nations Trade Data Tapes.

22
World trade in the processed industry showed a higher average increase
from 1978 to 1986, reaching 8.4% a year (see Tables 1.5 and 1.7). If
intraregional trade is not considered, the processed industry grew by 8%
a year during the same period.
Tables 1.9, 1.11, 1.13, and 1.15 show world fresh and processed
orange exports and imports excluding intraregional trade. Trade is
expressed in value terms measured in United States dollars.
The United States fresh orange production decreased at an average
rate of .3% a year from 1966 to 1986. During the 1970s, production
increased rapidly and later decreased mainly due to unfavorable weather
conditions. Oranges used for fresh consumption decreased at a rate of .5%
a year from 1966 to 1986. Total United States trade increased at a rate
of 2.4% a year for the same period. This shows that the United States has
actually decreased its real participation in world fresh utilization. It
has increased the use of oranges in the processed industry, along with a
slight increase in its role in the international trade arena for fresh
oranges. In the processed industry, the United States has decreased its
production participation relative to the rest of the world and passed from
a net exporter to a net importer of FCOJ (Frozen Concentrated Orange
Juice) (Tables 1.3, 1.5, and 1.7).
As shown in the different tables introduced in this chapter, in the
last two decades trade patterns in the orange industry have changed
dramatically. The United States, once the world's major producer of fresh
oranges and orange juice, today is no longer the leading producer or
exporter. Latin America, mainly Brazil, is the major producer of oranges
in the world. Most of Brazil's production is used for FCOJ and is

23
exported. Per capita consumption of fresh oranges within Brazil is quite
low. The Mediterranean countries are major producers and exporters of
fresh oranges. The rest of Europe as well as Canada have always been net
importers of fresh and processed oranges. Other regions, especially the
Middle East/North Africa and the Far East, have increasingly become
important producers and traders in the orange industry. Chapter 2 will
outline production and trade flows of the fresh orange industry in more
detail.
Problem and Objectives
World consumption trends indicate that consumers are interested in
healthy and natural products. Fresh product consumption is increasing and
its potential growth is promising. Given the changes in consumption
patterns and the improvement in the transportation systems, studying the
fresh markets is of increasing importance. Fresh oranges, in particular,
provide consumers with natural flavor and important vitamins and minerals.
Fresh oranges and FCOJ are direct substitutes in the supply decision
process, but are not considered substitutes in the consumption side.
Consumer satisfaction is considered to be different for each good. Most
recent literature has studied mainly processed trade, which has been
growing faster than fresh trade in the last two decades. However, the
value and quantity of fresh trade are two and four times that for the
processed trade, respectively (see Tables 1.8 to 1.15). World fresh
utilization represents 65% of total orange production.

24
The discussion in the previous section described some of the factors
affecting trade patterns and market shares of the fresh orange industry
during the last 21 years. Even though the market has experienced
important increases, several countries, including the United States, have
experienced pronounced changes in their trade patterns for value and
quantity. The dynamics of the marketplace are illustrated with
participation of the Middle East/North Africa countries in the European
market; Latin America's increasing share of the European market; the
increasing portion of the United States in the Far East markets; the
Middle East/North Africa's increasing participation as consumers of fresh
and processed oranges; and the potential growth of China as a supplying
and consuming country.
The fresh orange industry is quite important for some regions,
especially for the United States, Latin America, Mediterranean-EC, Middle
East/North Africa, and Far East, as producers, consumers, and exporters.
Producers and exporters need to understand the major driving factors for
fresh consumption and their competitive position in foreign markets. It
will allow them to compete with more information, possibly achieve
international success, and help to develop new markets. The fresh orange
industry is also important for net importers such as Canada, EC, rest of
Western Europe, and the Communist Bloc. These regions will be interested
in knowing which are the major driving factors for fresh consumption, and
demand and price linkages between the region and its major trading
partners.
Given the changes in the fresh orange market, studying the world
trade flows becomes important for the future of the United States orange

25
industry as well as for other partner regions. Modeling these changes is
the major objective of this study. The analysis will provide information
to help understand the reasons for changes in market shares among major
suppliers and facilitate longer term forecasts and policy analyses.
To accomplish the objectives of the present study, international
trade linkages among the major trading regions must be identified. It is
also necessary to recognize the current and emerging problems in the
industry. This information will be helpful to study changes in trade
patterns arising from changes in supply and demand conditions and from
changes in policy variables such as tariff levels and institutional
constraints.
Analysis of the demand parameters will show the likely future
direction of trade. Using price elasticities, it will be possible to
predict responses in the different markets to changes in prices. The role
of prices as an allocative tool can be shown. Income and population
elasticities will give an indication of possible adjustments in
consumption and trade patterns. In general, it will be possible to
forecast trade patterns among importers and exporters. The system could
be used to construct a sensitivity analysis to study the behavior of the
fresh orange trade model given shocks in the different variables including
price, market size, income, population, fresh production, tariff and
nontariff barriers, and other variables.
The specific objectives of this research are
1. Specify a multiple-region equilibrium world trade model for the
fresh orange industry. Relative and substitute prices,

26
transportation costs, incomes, populations, exchange rates, and
policy variables were considered.
2. Estimate the demand, export supply and price equations that
explain the individual elements of the trade flows. A
simultaneous system was specified and estimated.
3. Analyze the implications contained in the estimated model.
The estimated parameters were used to study the reasons for
changes in market shares and to provide information for
specific policy issues.
4. Develop a sensitivity analysis of the model for the major
trading regions under different economic scenarios. To ease
forecasts, exogenous changes in the different variables such
as import and export price, market size, income, population,
fresh production, tariffs and nontariff barriers, and other
variables were considered.
Scope
The proposed study will develop a world trade model with demand and
export supply equations for selected major trading regions. The model
will include fresh oranges as defined by the United Nations Standard
International Trade Classification (SITC) (1975) code 05711. Fresh
oranges and orange juice are direct substitutes in the supply decision
process but are not considered substitutes on the consumption side. To be
complete, the model should take into consideration the supply relationship
between the two goods. The end product model includes market size, market

27
and relative prices, transportation costs, tariff barriers (national or
regional agricultural policies), price of substitutes, income levels,
exchange rates, and population.
International trade data including value and quantity were obtained
from United Nations' Commodity Trade Statistic Tapes (1987). These data
are gathered by each member country and sent to the statistics office in
New York. The data contain import and export value and quantity
information for each member, showing the partner country. The price data
used in this dissertation are unit prices obtained by dividing value by
quantity for each trade flow. As expected, many errors were found. Most
of them were probably related to gathering problems and inconsistencies.
Where errors were detected, the data were corrected in what seemed to be
an appropriate way. Tariff barriers were obtained from different sources.
The kinds of tariffs differ from country to country, from an ad valorem
basis, using CIF import prices or FOB export prices as a base, to fixed
dollar amounts per ton. Tariffs were averaged using different methods to
obtain the best possible regional tariff. Nontariff barriers are not
considered in the study, given that most of them are seasonal and the
model uses annual data. The period of study is 1966 to 1986.
Methodology
The present study develops a fresh orange multiple-region
equilibrium world trade model. To ease the estimation and the analysis,
world countries are aggregated into 11 regions. The regions have been
selected by considering similarities in supply and/or demand among the

28
countries and their importance in production and international trade in
fresh oranges.
The model is a nonlinear simultaneous system of equations that
contains 440 equations of which 242 were estimated. The equations to
estimate were total market demands (one per region), export supplies (one
per region), product demands (one per partner in each region), and price
linkage equations (one per partner in each region) The rest of the
equations in the model were identities.
A nonlinear simultaneous system estimator was used for the
estimation of the model. Model results were analyzed to evaluate the fit
of the model and its accuracy for simulation. The final model and its
parameters were used to develop a sensitivity analysis to investigate the
effects of changes in selected policy variables.
Overview
Chapter 2 discusses world production and trade flows for fresh
oranges. Chapter 3 covers the literature review for agricultural trade
and fresh orange trade models. Chapter 4 presents the fresh orange trade
model to be estimated. Chapter 5 discusses the methods used for the
estimation of the model. It also develops a graphical, statistical, and
economic analysis for the results of the estimation. Chapter 6 develops
the sensitivity analysis.

CHAPTER 2
FRESH ORANGE WORLD PRODUCTION AND TRADE
Introduction
This chapter discusses world production and trade flows of fresh
oranges. The discussion will be based on several tables for 11 specified
regions of the world. These regions were selected based on similarities
of supply and demand conditions among the different countries included in
each region with regard to the orange industry. The regions are the
United States (US), Canada (CAN), Latin America (LA), Mediterranean-
European Community countries (MED-EC), the rest of the European Community
countries (EC), rest of Western Europe (RWE), Middle East/North Africa
(ME/NA), rest of Africa (RAF), Far East (FE), Oceania (OCE), and Communist
Bloc (COMMB). The Communist Bloc is defined as it existed before the
political changes of 1991. Appendix A shows the countries included in
each region.
Production Analysis
Table 2.1 shows the production levels of oranges in the 11 regions
identified for 1966, 1976, and 1986. These years were selected to
illustrate changes through time. World orange production increased at an
annual rate of 3.3% in the last 20 years and increased at a rate of 1.9%
29

Table 2.1 World Orange Production By Region
Region
1966
1976
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
--- (000)
Metric
Tons
Percent
of Change
United States
7598
10183
7192
-0.3
-3.4
Canada
0
0
0
N.A.
N.A.
Latin America
5540
12117
18535
6.2
4.3
Mediterranean-EC
4208
5472
6840
2.5
2.3
E.C.
4
31
34
11.3
0.9
Rest of Western Europe
0
0
0
N.A.
N.A.
Middle East/North Africa
3067
4664
5794
3.2
2.2
Rest of Africa
773
1034
1022
1.4
-0.1
Far East
4023
6532
8354
3.7
2.5
Oceania
249
360
574
4.3
4.8
Communist Bloc
192
292
728
6.9
9.6
World Total
25654
40685
49073
3.3
1.9
Source: FAO Production Yearbook.
Various
issues.
LO
O

31
in the last decade. Table 2.2 shows the portion of that production used
as fresh product. Production utilization in fresh form decreased from
75.3% in 1966 to 65.4% in 1986 (compare data in Table 2.2 as a percent of
the corresponding figures in 2.1).
Table 2.1 also shows that the major world producer was Latin
America, with 21.6% in 1966, 29.8% in 1976, and 37.8% in 1986. This
region exhibited one of the faster annual growth rates, 6.2% during the 20
year period. However, as shown in Table 2.2, over 50% of the oranges of
this region went to the processed industry, leaving 9.4 million tons for
fresh utilization in 1986. This represented 29.3% of total world fresh
utilization.
The second largest producer of oranges was the Far East, with 15.7%
in 1966, 16.1% in 1976, and 17% in 1986. These percentages show that the
Far East region has not only maintained its participation in the total
world production of oranges in the last 20 years, but has also increased
it. Table 2.1 shows that the Far East region has doubled in absolute
terms its total production in the same period. In addition, 91.2% of
total production was used fresh in 1986.
The third largest producer was the United States, with 29.6% in
1966, 25% in 1976, and 14.7% in 1986. Even though the United States is
still a major world producer, its share of total production of oranges has
been decreasing, especially in the last decade. The United States used
most of its production in the processed industry. In 1986, 32.3% of total
production was used fresh, indicating that the United States was not the
third major supplier of oranges to the fresh markets.

Tab]o 2.2 World Fresh Utilization by Region
Region
1966
1976
1986
Annual
Growth
Rate
1966-86
Annual
Growth
Rate
1976-86
--- (000)
Metric
Tons
Percent
of Change
United States
2575
2294
2322
-0.5
0.1
Canada
0
0
0
N.A.
N.A.
Latin America
5290
8336
9394
2.9
1.2
Mediterranean-EC
3689
4807
5787
2.3
1.9
E.C.
4
31
34
11.3
0.9
Rest of Western Europe
0
0
0
N.A.
N.A.
Middle East/North Africa
2862
4273
5122
3.0
1.8
Rest of Africa
692
894
919
1.4
0.3
Far East
3801
5634
7619
3.5
3.1
Oceania
202
196
251
1.1
2.5
Communist Bloc
192
287
638
6.2
8.3
World Total
19307
26752
32086
2.6
1.8

33
In that year, the United States occupied the fifth position in fresh sales
worldwide.
The fourth largest producer of oranges was the Mediterranean-EC.
This region's share in world production of fresh oranges was 16.4% in
1966, 13.4% in 1976, and 13.9% in 1986. This region's production grew
rapidly in the last decade. This growth coincided with the incorporation
into the European Community (EC) of all the countries included in this
particular region. The Mediterranean-EC dedicated 15.4% of its total
orange production to the processed industry in 1986. Fresh utilization
represented 18% of total world fresh orange production. The region
occupied the third position in the fresh market in 1986.
The fifth major producer of oranges was the Middle East/North
Africa, with 12.0% in 1966, 11.5% in 1976, and 11.8% in 1986. In 1986,
88.4% of total production was used fresh, giving the Middle East/North
Africa region the fourth position in the world fresh orange market.
The rest of Africa was the sixth largest producer of oranges with
3.0% in 1966, 2.5% in 1976, and 2.1% in 1986. As shown in Table 2.2, out
of total orange production, this region dedicated 10.1% to the processed
industry in 1986. In that year, the region occupied the sixth position in
fresh sales worldwide.
The rest of world production of oranges was provided by the
Communist Bloc, Oceania, and the EC with .7%, 1.0%, and .02% in 1966 and
1.5%, 1.2%, and .1% in 1986, respectively. The Communist Bloc dedicated
99.1% of total production to the fresh orange industry in 1986, while in
Oceania 43.7% of total production was used for fresh consumption.

34
Trade Flow Analysis
Table 2.3 shows the quantities traded between the 11 regions for
1966, 1976, and 1986. These years were selected to illustrate changes
through time. The first column of this table represents the different
years, the second column and the top row of the table represent the region
and the partner region names, respectively. Each of the 11 columns
depicts the quantities exported from the partner region to each region.
The following two columns show the total product imported by each region,
with the first one including the intraregional trade and the other
including interregional trade. Since the first and second regions consist
of a single country, both columns display the same values. The last two
columns exhibit the percentages associated with the previous two columns
in relation to total world imports. Similarly, the last four rows of the
table contain total exports from each partner region. The first row
includes intraregional trade, the second one only interregional trade, and
the last two rows show the percentages associated with total world
exports.
Tables 2.4 and 2.5 contain the percentages needed to illustrate the
allocation of exports, imports, and trade flows in total and among the 11
regions. Table 2.4 shows the percentages from the exporter or partner
region position and Table 2.5 from the importer or region perspective. In
both tables, intraregional trade was excluded, given that the major
interest of the present study has to do with trade among the regions.
Intraregional quantities are part of the region's production that is
consumed domestically.

Table 2.3 Trade Flow Analysis for Selected Years (1966, 1976 and 1986) by Region in Relation to Partner Regions
Year
Region
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
RE
OCE
COtUB
Total
W/INTR*
Total
W/O INTRAb
Z
W/INTR
Z W/O
INTR
66
us
0
0
23285
222
13
0
4617
0
212
0
0
28349
28349
0.8
0.8
76
0
11
31094
196
7
0
300
0
952
89
0
32649
32649
0.7
0.8
86
0
19
22334
15808
25
0
9706
0
1477
0
0
49369
49369
0.9
1.0
66
CAN
140366
0
1513
81
20
0
8536
16101
13805
59
0
180501
180501
5.2
5.3
76
190943
0
231
24
0
0
3225
6445
23062
1616
0
225547
225547
4.5
5.3
86
123772
0
5297
18534
68
0
21490
0
10773
2168
0
182102
182102
3. A
3.7
66
LA
6397
0
3191
0
79
0
0
0
2
0
0
9669
6A78
0.3
0.2
76
1932
45
16110
0
56
2
0
0
0
0
0
18145
2035
0.4
0.0
86
582
3
2455
61
217
0
18
A
16
0
451
3807
1352
0.1
0.0
66
MED-EC
0
0
0
0
0
0
0
0
0
0
0
0
0
0.0
0.0
76
7
0
0
183
253
28
429
0
0
0
0
900
717
0.0
0.0
86
0
0
828
259
3167
158
497
3326
0
0
8
8243
7984
0.2
0.2
66
EC
56148
62
69439
1116393
29955
1079
908131
213821
23
3371
19
2398441
2368486
68.8
69.9
76
104359
2
39303
1516481
112937
817
760700
197487
36
2085
1983
2736190
2623253
54.5
91.3
86
9336
232
176180
2217655
173847
637
696538
176595
56
1825
11665
3464566
3290719
64.2
67.3
66
RWE
9889
0
2989
229717
1464
976
121588
17024
422
429
151
384649
383673
11.0
11.3
76
4704
0
2875
177279
3318
5067
223720
20800
7
1839
52
439661
434594
8.8
10.2
86
1619
0
6076
321349
15663
1073
186964
17922
36
276
265
551243
550170
10.2
11.3
66
ME/NA
17
0
0
1024
17
0
10612
0
9799
579
0
22048
11436
0.6
0.3
76
25
0
105
29447
5060
12
461007
54334
40379
0
0
590369
129362
11.8
3.0
86
15
0
3243
16333
116
0
224078
0
29574
5650
1
279010
54932
5.2
1.1
66
RAF
54
0
9
95
1045
0
7324
1764
2
300
0
10593
8829
0.3
0.3
76
7
0
178
747
742
0
3628
5493
57
36
0
10588
5095
0.2
0.1
86
0
0
5
490
761
0
4294
2119
0
542
0
8211
6092
0.2
0.1
66
FE
38555
9
595
358
9
19
13848
15479
41505
8964
0
119341
77836
3.4
2.3
76
145976
0
0
924
5
0
9988
1149
77821
3052
3
238918
161097
4.8
3.8
86
267916
14
324
1637
53
0
5730
8997
71939
26219
0
382829
310890
7.1
6.4
66
OCE
2210
0
1420
0
0
0
1618
225
0
9843
0
15316
5473
0.4
0.2
76
6440
0
0
0
0
0
23
0
0
2433
0
8896
6463
0.2
0.2
86
9802
0
0
0
8
0
0
0
5
10588
0
20403
9815
0. A
0.2
66
COMMB
4940
0
4130
166667
194
2
139601
11
10
0
0
315555
315555
9.1
9.3
76
9701
0
29292
212239
443
0
408836
0
0
0
58318
718829
660511
14.3
15.4
86
0
0
12603
241790
1035
314
167406
0
0
33
22656
445839
423181
8.3
8.7

Table 2.3--continued.
Year
Region
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
RE
OCE
COMMB
Total
W/INTR1
Total
W/0 INTRAb
z
W/INTR
Z W/0
INTR
66
Total
258596
71
106571
1514557
32796
2076
1215875
264425
65780
23545
170
3484462
3386616
100.0
100.0
76
W/INTR
464094
58
119188
1937490
122551
5926
1871856
285709
142314
11150
60356
5020692
4281323
100.0
100.0
86
413042
268
229345
2833916
194960
2182
1316721
208963
113876
47301
35046
5395620
4886606
100.0
100.0
66
Total
258596
71
103380
1514557
2841
1100
1205263
262661
24275
13702
170
3386616
76
W/0 INTR
464094
58
103078
1937307
9614
859
1410849
280216
64493
8717
2038
4281323
86
413042
268
226890
2833657
21113
1109
1092643
206844
41937
36713
12390
4886606
66
Z W/INTR
7.4
.0
3.1
43.5
0.9
0.1
34.9
7.6
1.9
0.7
.0
100.0
76
9.2
.0
2.4
38.6
2.4
0.1
37.3
5.7
2.8
0.2
1.2
100.0
86
7.7
.0
4.3
52.5
3.6
.0
24.4
3.9
2.1
0.9
0.6
100.0
66 Z
W/0 INTR
7.6
.0
3.1
44.7
0.1
.0
35.6
7.8
0.7
0.4
.0
100.0
76
10.8
.0
2.4
45.3
0.2
.0
33.0
6.5
1.5
0.2
.0
100.0
86
8.5
.0
A .6
58.0
0.4
.0
22.4
4.2
0.9
0.8
0.3
100.0
'Total includes intraregional trade.
Total does not Include intraregional trade.

37
Table 2.
4 Trade Flow Analysis for Selected Years (1966, 1976 and 1986)
Without Intraregional Trade "Relative Partner Region Exports
by Region"
Year
Region US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
COMMB
- Percentages -
66
US
0.0
0.0
22.5
0.0
0.5
0.0
0.4
0.0
0.9
0.0
0.0
76
0.0
19.0
30.2
0.0
0.1
0.0
0.0
0.0
1.5
1.0
0.0
86
0.0
7.1
9.8
0.6
0.1
0.0
0.9
0.0
3.5
0.0
0.0
66
CAN
54.3
0.0
1.5
0.0
0.7
0.0
0.7
6.1
56.9
0.4
0.0
76
41.1
0.0
0.2
0.0
0.0
0.0
0.2
2.3
35.8
18.5
0.0
86
30.0
0.0
2.3
0.7
0.3
0.0
2.0
0.0
25.7
5.9
0.0
66
LA
2.5
0.0
0.0
0.0
2.8
0.0
0.0
0.0
0.0
0.0
0.0
76
0.4
77.6
0.0
0.0
0.6
0.2
0.0
0.0
0.0
0.0
0.0
86
0.1
1.1
0.0
0.0
1.0
0.0
0.0
0.0
0.0
0.0
3.6
66
MED-EC
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
76
0.0
0.0
0.0
0.0
2.6
3.3
0.0
0.0
0.0
0.0
0.0
86
0.0
0.0
0.4
0.0
15.0
14.2
0.0
1.6
0.0
0.0
0.1
66
EC
21.7
87.3
67.2
73.7
0.0
98.1
75.3
81.4
0.1
24.6
11.2
76
22.5
3.4
38.1
78.3
0.0
95.1
53.9
70.5
0.1
23.9
97.3
86
2.3
86.6
77.6
78.3
0.0
57.4
63.7
85.4
0.1
5.0
94.1
66
RWE
3.8
0.0
2.9
15.2
51.5
0.0
10.1
6.5
1.7
3.1
88.8
76
1.0
0.0
2.8
9.2
34.5
0.0
15.9
7.4
0.0
21.1
2.6
86
0.4
0.0
2.7
11.3
74.2
0.0
17.1
8.7
0.1
0.8
2.1
66
ME/NA
0.0
0.0
0.0
0.1
0.6
0.0
0.0
0.0
40.4
4.2
0.0
76
0.0
0.0
0.1
1.5
52.6
1.4
0.0
19.4
62.6
0.0
0.0
86
0.0
0.0
1.4
0.6
0.5
0.0
0.0
0.0
70.5
15.4
0.0
66
RAF
0.0
0.0
0.0
0.0
36.8
0.0
0.6
0.0
0.0
2.2
0.0
76
0.0
0.0
0.2
0.0
4.9
0.0
0.3
0.0
0.1
0.4
0.0
86
0.0
0.0
0.0
0.0
3.6
0.0
0.4
0.0
0.0
1.5
0.0
66
FE
14.9
12.7
0.6
0.0
0.3
1.7
1.1
5.9
0.0
65.4
0.0
76
31.5
0.0
0.0
0.0
0.1
0.0
0.7
0.4
0.0
35.0
0.1
86
64.9
5.2
0.1
0.1
0.3
0.0
0.5
4.3
0.0
71.4
0.0
66
OCE
0.9
0.0
1.4
0.0
0.0
0.0
0.1
0.1
0.0
0.0
0.0
76
1.4
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
86
2.4
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
66
COMMB
1.9
0.0
4.0
11.0
6.8
0.2
11.6
0.0
0.0
0.0
0.0
76
2.1
0.0
28.4
11.0
4.6
0.0
29.0
0.0
0.0
0.0
0.0
86
0.0
0.0
5.6
8.5
4.9
28.3
15.3
0.0
0.0
0.1
0.0
66
TOTAL
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
76
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
86
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0

38
Table 2.5 Trade Flow Analysis for Selected Years (1966, 1976 and 1986)
Without Intraregional Trade "Relative Region Imports from
Partner Regions"
Year
Region
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE COMMB
Total
Percentages -
66
US
0.0
0.0
82.1
0.8
0.0
0.0
16.3
0.0
0.7
0.0
0.0
100.0
76
0.0
0.0
95.2
0.6
0.0
0.0
0.9
0.0
2.9
0.3
0.0
100.0
86
0.0
0.0
45.2
32.0
0.1
0.0
19.7
0.0
3.0
0.0
0.0
100.0
66
CAN
77.8
0.0
0.8
0.0
0.0
0.0
4.7
8.9
7.6
0.0
0.0
100.0
76
84.7
0.0
0.1
0.0
0.0
0.0
1.4
2.9
10.2
0.7
0.0
100.0
86
68.0
0.0
2.9
10.2
0.0
0.0
11.8
0.0
5.9
1.2
0.0
100.0
66
LA
98.7
0.0
0.0
0.0
1.2
0.0
0.0
0.0
0.0
0.0
0.0
100.0
76
94.9
2.2
0.0
0.0
2.8
0.1
0.0
0.0
0.0
0.0
0.0
100.0
86
43.0
0.2
0.0
4.5
16.1
0.0
1.3
0.3
1.2
0.0
33.4
100.0
66
MED-EC
N. A
N.A
N.A
N.A
N.A
N.A
N.A
N.A
N.A
N.A
N.A
N.A
76
1.0
0.0
0.0
0.0
35.3
3.9
59.8
0.0
0.0
0.0
0.0
100.0
86
0.0
0.0
10.4
0.0
39.7
2.0
6.2
41.7
0.0
0.0
0.1
100.0
66
EC
2.4
0.0
2.9
47.1
0.0
0.0
38.3
9.0
0.0
0.1
0.0
100.0
76
4.0
0.0
1.5
57.8
0.0
0.0
29.0
7.5
0.0
0.1
0.1
100.0
86
0.3
0.0
5.4
67.4
0.0
0.0
21.2
5.4
0.0
0.1
0.4
100.0
66
RWE
2.6
0.0
0.8
59.9
0.4
0.0
31.7
A. A
0.1
0.1
0.0
100.0
76
1.1
0.0
0.7
40.8
0.8
0.0
51.5
4.8
0.0
0.4
0.0
100.0
86
0.3
0.0
1.1
58.4
2.8
0.0
34.0
3.3
0.0
0.1
0.0
100.0
66
ME/NA
0.1
0.0
0.0
9.0
0.1
0.0
0.0
0.0
85.7
5.1
0.0
100.0
76
0.0
0.0
0.1
22.8
3.9
0.0
0.0
42.0
31.2
0.0
0.0
100.0
86
0.0
0.0
5.9
29.7
0.2
0.0
0.0
0.0
53.8
10.3
0.0
100.0
66
RAF
0.6
0.0
0.1
1.1
11.8
0.0
83.0
0.0
0.0
3.4
0.0
100.0
76
0.1
0.0
3.5
14.1
9.3
0.0
71.2
0.0
1.1
0.7
0.0
100.0
86
0.0
0.0
0.1
8.0
12.5
0.0
70.5
0.0
0.0
8.9
0.0
100.0
66
FE
49.5
0.0
0.8
0.5
0.0
0.0
17.8
19.9
0.0
11.5
0.0
100.0
76
90.6
0.0
0.0
0.6
0.0
0.0
6.2
0.7
0.0
1.9
0.0
100.0
86
86.2
0.0
0.1
0.5
0.0
0.0
1.8
2.9
0.0
8.4
0.0
100.0
66
0CE
40.4
0.0
25.9
0.0
0.0
0.0
29.6
4.1
0.0
0.0
0.0
100.0
76
99.6
0.0
0.0
0.0
0.0
0.0
0.4
0.0
0.0
0.0
0.0
100.0
86
99.9
0.0
0.0
0.0
0.1
0.0
0.0
0.0
0.1
0.0
0.0
100.0
66
C0f*ffl
1.6
0.0
1.3
52.8
0.1
0.0
44.2
0.0
0.0
0.0
0.0
100.0
76
1.5
0.0
4.4
32.1
0.1
0.0
61.9
0.0
0.0
0.0
0.0
100.0
86
0.0
0.0
3.0
57.1
0.2
0.1
39.6
0.0
0.0
0.0
0.0
100.0

39
Tables 2.6, 2.7, and 2.8 correspond to the same characteristics
described for Tables 2.3, 2.4, and 2.5, respectively. The difference
between these two sets of tables is that the former tables present
cumulative information for five years instead of yearly information. The
periods considered are 1966 to 1970, 1974 to 1978, and 1982 to 1986. The
discussion that follows will be based mainly on the first set of tables,
because both sets draw similar conclusions. However, given that yearly
information could be biased for uncommon reasons, the results in Table 2.6
to 2.8 are useful to support general conclusions.
Partner Region Perspective
In this section, the discussion will be oriented from the exporters'
viewpoint. In all cases, the relative importance of each region is set
forth and then a trade flow analysis is developed.
Table 2.3 shows that the world's major fresh orange exporter was the
Mediterranean-EC region. With intraregional trade considered, this
region's share of total exports was 43.5% in 1966, 38.6% in 1976, and
52.5% in 1986. With intraregional trade not considered, the relative
importance of the region in world trade increased to 44.7%, 45.3%, and 58%
respectively. These values show the importance of this region in world
trade of fresh oranges.
Table 2.4 shows that the major partner of the Mediterranean-EC was
the EC region. In 1966, 73.7% of the Mediterranean-EC region's total
exports went to the EC region. This percentage increased to 78.3 in 1976
and was the same in 1986. The EC region includes all EC countries except

Table
2.6
Trade
Flow
Analysis for
Selected Periods of
Five
Years
(1966-70,
1974-
78 and
1982-86)
Total
Total
Z
Z W/O
Period
Region
us
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
coroiB
W/INTR*
W/O INTR* W/INTR
INTR
Metric to U.S.
pi*
us
0
44
221119
524
25
0
35283
370
2142
47
0
259554
259554
1.4
1.5
P24
0
50
184606
570
46
0
16201
0
4263
101
0
205837
205837
0.8
1.0
P3*
0
162
177272
23466
200
0
34632
19
6130
685
0
242566
242566
1.0
1.2
PI
CAN
699870
0
15199
1489
133
0
36954
68392
73338
1307
0
896682
896682
4.8
5.1
P2
872603
0
2700
247
61
0
26027
36787
93671
8106
0
1040202
1040202
4.2
4.9
P3
698262
0
22155
39563
70
0
98322
0
76031
10741
0
945144
945144
4.0
4.6
PI
LA
16653
95
39255
2
599
0
812
0
2
0
0
57418
18163
0.3
0.1
P2
9149
45
75425
0
234
2
16
0
0
2
2
84875
9450
0.3
0.0
P3
7866
7
58938
212
1467
0
320
29
18
25
1173
70055
11117
0.3
0.1
PI
MED-EC
3
0
83
716
51
4
1801
0
0
0
0
2658
1942
0.0
0.0
P2
7
0
16
251
1315
77
2202
22
0
0
0
3890
3639
0.0
0.0
P3
9
0
8472
3456
15792
448
1879
5726
4
0
197
35983
32527
0.2
0.2
PI
EC
261380
127
287289
5486327
279949
3335
4843666
1074996
452
11516
1334
12250371
11970422
65.2
67.9
P2
388239
61
304034
7307729
570730
3660
3771137
1057439
354
10791
11940
13426114
12855384
54.1
60.3
P3
57895
307
693974
8117851
738464
4352
3443275
775683
607
4 764
57678
13894850
13156386
59.1
63.8
PI
RWE
30694
0
24337
1033534
10465
6541
785484
103987
554
1154
1607
1998357
1991816
10.6
11.3
P2
26710
0
20179
867897
24061
23548
1126693
117572
1163
7224
888
2215935
2192387
8.9
10.3
P3
8319
9
20512
1011829
65269
8745
956474
105954
1136
1511
690
2180448
2171703
9.3
10.5
PI
ME/NA
55
0
0
2337
276
1166
490985
4643
22659
1136
0
523257
32272
2.8
0.2
P2
14826
0
23110
54432
12269
11175
2041502
144372
213855
10744
66
2526351
484849
10.2
2.3
P3
2108
0
125929
46085
738
1403
1581238
88428
164543
20389
4
2030865
449627
8.6
2.2
PI
RAF
106
0
2160
405
4147
0
24178
13071
14
1345
47
45473
32402
0.2
0.2
P2
43
0
2239
3778
3664
45
21629
27635
146
130
1
59310
31675
0.2
0.1
P3
11
0
506
2959
5110
0
23544
11321
10
2809
9
46279
34958
0.2
0.2
PI
FE
227715
9
6930
2223
341
55
81373
41497
293406
48004
0
701553
408147
3.7
2.3
P2
653456
1261
561
4411
64
5
61734
11860
376488
24833
13
1134686
758198
4.6
3.6
P3
1213817
105
1654
6258
349
24
39975
40106
357283
93041
27
1752639
1395356
7.5
6.8
PI
OCE
10170
0
6627
24
0
0
5768
1996
1
51174
0
75760
24586
0.4
0.1
P2
45305
0
977
0
0
0
193
1224
5
33201
0
80905
47704
0.3
0.2
P3
57437
0
0
3003
27
0
418
1207
243
43736
0
106071
62335
0.5
0.3
-O
o

Table 2.6--continued.
Period Region US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
cot*
Total
W/INTR*
Total
W/O INTRb
X
W/INTR
X W/O
INTR
Metric to U.S.
pi
COMMB 14202
0
30451
846525
1262
18
1089907
247
943
76
230
1983861
1983631
10.6
11.3
P2
36116
0
130615
1154096
1730
340
2368510
0
752
0
331498
4023657
3692159
16.2
17.3
P3
0
0
24227
912980
3138
1071
1171677
41
1
49
100226
2213410
2113184
9.4
10.3
PI
TOTAL 1260848
275
633450
7374106
297248
11119
7396211
1309199
393511
115759
3218
18794944
17619617
100.0
100.0
P2
W/INTR 2046454
1417
744462
9393411
614174
38852
9435844
1396911
690697
95132
344408
24801762
21321484
100.0
100.0
P3
2045724
590
1133639
10167662
830624
16043
7351754
1028514
606006
177750
160004
23518310
20614903
100.0
100.0
PI
TOTAL 1260848
275
594195
7373390
17299
4578
6905226
1296128
100105
64585
2988
17619617
P2
W/O INTR 2046454
1417
669037
9393160
43444
15304
7394342
1369276
314209
61931
12910
21321484
P3
2045724
590
1074701
10164206
92160
7298
5770516
1017193
248723
134014
59778
20614903
PI
X W/INTR 6.7
0.0
3.4
39.2
1.6
0.1
39.4
7.0
2.1
0.6
0.0
100.0
P2
8.3
0.0
3.0
37.9
2.5
0.2
38.0
5.6
2.8
0.4
1.4
100.0
P3
8.7
0.0
4.8
43.2
3.5
0.1
31.3
A. A
2.6
0.8
0.7
100.0
PI
X W/O INTR 7.2
0.0
3.4
41.8
0.1
0.0
39.2
7. A
0.6
0.4
0.0
100.0
P2
9.6
0.0
3.1
A A 1
0.2
0.1
34.7
6. A
1.5
0.3
0.1
100.0
P3
9.9
0.0
5.2
A9.3
0. A
0.0
28.0
A .9
1.2
0.7
0.3
100.0
Total includes intraregional trade.
bTotal does not include intraregional trade.
Represents period from 1966-70.
Represents period from 197A-78.
Represents period from 1982-86.

42
Table 2.7 Trade Flow Analysis for Selected Periods of Five Years, (1966-
70, 1974-78 and 1982-86) Without Intraregional Trade "Relative
Partner Region Exports by Region"
Period Region US CAN LA MED-EC EC RWE ME/NA RAF FE OCE COMMB
Percentages
PI
US
0.
.0
16
.0
37
.2
0.
.0
0
.1
0
.0
0
.5
0
.0
2
. 1
0
.1
0
.0
PE"
0.
.0
3
.5
27
.6
0,
.0
0
.1
0.
.0
0
.2
0
.0
1
. 4
0
.2
0
.0
P3*
0.
.0
27
.5
16
.5
0,
.2
0.
.2
0.
.0
0
.6
0
.0
2
.5
0
.5
0
.0
PI
CAN
55.
.5
0.
.0
2
.6
0.
.0
0
.8
0.
.0
0
.5
5
.3
73
.3
2
.0
0
.0
P2
42.
.6
0.
.0
0.
.4
0,
.0
0
.1
0.
.0
0
.4
2
.7
29
.8
13.
.1
0
.0
P3
34.
. 1
0.
.0
2.
.1
0.
.4
0
.1
0.
.0
1
.7
0
.0
30
.6
8
.0
0
.0
PI
LA
1.
.3
34
.5
0
.0
0,
.0
3
.5
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
P2
0.
.4
3
.2
0
.0
0,
.0
0
.5
0
.0
0
.0
0
.0
0
.0
0
.0
0
.0
P3
0.
.4
1
.2
0
.0
0,
,0
1
.6
0
.0
0
.0
0
.0
0
.0
0
.0
2
.0
PI
MED-EC
0
.0
0
.0
0
.0
0,
.0
0
.3
0
. 1
0
.0
0
.0
0
.0
0
.0
0
.0
P2
0.
.0
0
.0
0
.0
0,
,0
3
.0
0.
.5
0
.0
0
.0
0
.0
0
.0
0
.0
P3
0.
,0
0
.0
0.
.8
0,
,0
17
. 1
6.
. 1
0
.0
0
.6
0
.0
0.
.0
0
.3
PI
EC
20.
.7
46.
.2
48.
.3
74.
.4
0
.0
72.
.8
70
. 1
82
.9
0
.5
17.
.8
44
.6
P2
19.
.0
4.
.3
45.
.4
77,
,8
0
.0
23.
.9
51
.0
77
.2
0
.1
17
,4
92
.5
P3
2.
.8
52.
.0
64.
.6
79.
.9
0.
.0
59.
.6
59
. 7
76
.3
0.
.2
3.
.6
96
.5
PI
RWE
2.
.4
0.
.0
4,
.1
14.
,0
60
.5
0,
.0
11.
.4
8
.0
0.
.6
1.
.8
53
.8
P2
1.
.3
0.
.0
3.
.0
9.
.2
55.
.4
0.
.0
15.
.2
8
.6
0.
.4
11.
,7
6
.9
P3
0,
.4
1
.5
1.
.9
10.
.0
70
.8
0.
.0
16
.6
10.
.4
0.
.5
1.
.1
1
.2
PI
ME/NA
0.
.0
0.
.0
0.
.0
0.
,0
1.
.6
25.
.5
0
.0
0,
.4
22.
.6
1.
.8
0
.0
P2
0.
,7
0.
.0
3.
.5
0.
.6
28
.2
73,
.0
0
.0
10.
.5
68.
. 1
17.
.3
0
.5
P3
0.
.1
0.
.0
11.
.7
0.
.5
0.
.8
19.
.2
0
.0
8.
.7
66.
.2
15.
.2
0
.0
PI
RAF
0.
0
0.
.0
0.
.4
0.
.0
24,
,0
0,
,0
0
. 4
0,
.0
0.
.0
2.
, 1
1
.6
P2
0.
.0
0.
.0
0.
,3
0.
.0
8.
.4
0.
,3
0,
.3
0,
.0
0.
.0
0.
2
0
.0
P3
0.
0
0.
.0
0.
.0
0.
0
5.
.5
0.
.0
0.
.4
0,
,0
0.
.0
2.
.1
0.
,0
PI
FE
18.
.1
3,
.3
1.
,2
0.
0
2,
,0
1.
2
1.
.2
3.
.2
0.
.0
74.
3
0.
.0
P2
31.
9
89.
.0
0.
.1
0.
0
0,
,1
0,
,0
0,
,8
0,
.9
0.
.0
40.
1
0.
,1
P3
59.
3
17.
.8
0.
2
0.
1
0.
,4
0.
,3
0,
,7
3,
.9
0.
.0
69.
.4
0,
,0
PI
OCE
0.
8
0.
.0
1.
.1
0.
0
0,
.0
0.
.0
0.
.1
0.
.2
0.
.0
0.
0
0.
.0
P2
2.
2
0.
.0
0.
.1
0.
0
0.
,0
0.
.0
0.
.0
0.
,1
0.
.0
0.
0
0,
,0
P3
2.
8
0.
.0
0.
.0
0.
0
0.
.0
0.
.0
0,
.0
0.
.1
0.
1
0.
0
0.
.0
PI
COMiB
1.
1
0.
,0
5.
,1
11.
5
7.
.3
0.
.4
15.
.8
0.
.0
0.
9
0.
1
0.
.0
P2
1.
8
0.
0
19.
.5
12.
3
4.
.0
2.
,2
32.
,0
0.
.0
0.
2
0.
0
0.
.0
P3
0.
0
0.
.0
2.
.3
9.
0
3.
.4
14.
.7
20.
,3
0.
.0
0.
0
0.
0
0.
.0
PI
TOTAL
100.
0
100.
,0
100.
.0
100.
0
100.
,0
100.
,0
100.
.0
100.
.0
100.
0
100.
0
100.
.0
P2
100.
0
100.
0
100.
.0
100.
0
100.
.0
100.
,0
100.
.0
100.
.0
100.
0
100.
0
100.
0
P3
100.
0
100.
0
100.
.0
100.
0
100.
,0
100.
,0
100.
,0
100.
.0
100.
0
100.
0
100.
0
Represents period from 1966-70.
Represents period from 1974-78.
Represents period from 1982-86.

43
Table 2.8 Trade Flow Analysis for Selected Periods of Five Years (1966-
70, 1974-78 and 1982-86) Without Intraregional Trade "Relative
Region Imports from Partner Regions"
PeriodRegion
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
COMMB
Total
Percentages
PI*
US
0.0
0.0
85.2
0.2
0.0
0.0
13.6
0.1
0.8
0.0
0.0
100.0
P2b
0.0
0.0
89.7
0.3
0.0
0.0
7.9
0.0
2.1
0.0
0.0
100.0
P3*
0.0
0.1
73.1
9.7
0.1
0.0
14.3
0.0
2.5
0.3
0.0
100.0
PI
CAN
78.1
0.0
1.7
0.2
0.0
0.0
4.1
7.6
8.2
0.1
0.0
100.0
P2
83.9
0.0
0.3
0.0
0.0
0.0
2.5
3.5
9.0
0.8
0.0
100.0
P3
73.9
0.0
2.3
4.2
0.0
0.0
10.4
0.0
8.0
1.1
0.0
100.0
PI
LA
91.7
0.5
0.0
0.0
3.3
0.0
4.5
0.0
0.0
0.0
0.0
100.0
P2
96.8
0.5
0.0
0.0
2.5
0.0
0.2
0.0
0.0
0.0
0.0
100.0
P3
70.8
0.1
0.0
1.9
13.2
0.0
2.9
0.3
0.2
0.2
10.6
100.0
PI
MED-EC
0.2
0.0
4.3
0.0
2.6
0.2
92.7
0.0
0.0
0.0
0.0
100.0
P2
0.2
0.0
0.4
0.0
36.1
2.1
60.5
0.6
0.0
0.0
0.0
100.0
P3
0.0
0.0
26.0
0.0
48.6
1.4
5.8
17.6
0.0
0.0
0.6
100.0
PI
EC
2.2
0.0
2.4
45.8
0.0
0.0
40.5
9.0
0.0
0.1
0.0
100.0
P2
3.0
0.0
2.4
56.8
0.0
0.0
29.3
8.2
0.0
0.1
0.1
100.0
P3
0.4
0.0
5.3
61.7
0.0
0.0
26.2
5.9
0.0
0.0
0.4
100.0
FI
RWE
1.5
0.0
1.2
51.9
0.5
0.0
39.4
5.2
0.0
0.1
0.1
100.0
P2
1.2
0.0
0.9
39.6
1.1
0.0
51.4
5.4
0.1
0.3
0.0
100.0
P3
0.4
0.0
0.9
46.6
3.0
0.0
44.0
4.9
0.1
0.1
0.0
100.0
PI
ME/NA
0.2
0.0
0.0
7.2
0.9
3.6
0.0
14.4
70.2
3.5
0.0
100.0
F2
3.1
0.0
4.8
11.2
2.5
2.3
0.0
29.8
44.1
2.2
0.0
100.0
P3
0.5
0.0
28.0
10.2
0.2
0.3
0.0
19.7
36.6
4.5
0.0
100.0
PI
RAF
0.3
0.0
6.7
1.2
12.8
0.0
74.6
0.0
0.0
4.2
0.1
100.0
P2
0.1
0.0
7.1
11.9
11.6
0.1
68.3
0.0
0.5
0.4
0.0
100.0
P3
0.0
0.0
1.4
8.5
14.6
0.0
67.3
0.0
0.0
8.0
0.0
100.0
PI
FE
55.8
0.0
1.7
0.5
0.1
0.0
19.9
10.2
0.0
11.8
0.0
100.0
P2
86.2
0.2
0.1
0.6
0.0
0.0
8.1
1.6
0.0
3.3
0.0
100.0
P3
87.0
0.0
0.1
0.4
0.0
0.0
2.9
2.9
0.0
6.7
0.0
100.0
PI
OCE
41.4
0.0
27.0
0.1
0.0
0.0
23.5
8.1
0.0
0.0
0.0
100.0
P2
95.0
0.0
2.0
0.0
0.0
0.0
0.4
2.6
0.0
0.0
0.0
100.0
P3
92.1
0.0
0.0
4.8
0.0
0.0
0.7
1.9
0.4
0.0
0.0
100.0
PI
COPMB
0.7
0.0
1.5
42.7
0.1
0.0
54.9
0.0
0.0
0.0
0.0
100.0
P2
1.0
0.0
3.5
31.3
0.0
0.0
64.1
0.0
0.0
0.0
0.0
100.0
P3
0.0
0.0
1.1
43.2
0.1
0.1
55.4
0.0
0.0
0.0
0.0
100.0
'Represents period from 1966-70.
^Represents period from 1974-78.
'Represents period from 1982-86.

44
Spain, Italy, Portugal, and Greece. The second largest partner of the
Mediterranean-EC was the rest of Western Europe. Table 2.4 shows that the
relative importance of the rest of Western Europe in the Mediterranean-
EC' s total exports decreased from 15.2% in 1966 to 11.3% in 1986. The
third major partner of the Mediterranean-EC region was the Communist Bloc.
This region accounted for 11.0% of the Mediterranean-EC's total exports in
1966 and 8.5% in 1986. Exports to the rest of the partners were small,
but exports to the United States and Canada have increased in the last few
years.
The second major exporter region of the world was the Middle
East/North Africa. As opposed to the Mediterranean-EC region, this one
has been losing its share of the market in the last 20 years.
Participation in total world exports increased from 34.9% in 1966 to 37.3%
in 1976 (see Table 2.3). Nevertheless, the region's share of the export
market decreased to 24.4% in 1986. Examining exports without considering
intraregional trade shows that this region was losing its share of the
external market faster than its own regional market share. Table 2.3
shows that interregional percentages of the Middle East/North Africa
decreased from 35.6 in 1966 to 22.4 in 1986.
The Middle East/North Africa region's major partner was the EC
region. In 1966, 75.3% of total interregional exports from the Middle
East/North Africa countries went to the EC countries (see Table 2.4).
This percentage has since been decreasing, and in 1986 it represented only
63.7. In 1976, the percentage was lower, mainly due to an important shift
of exports to the Communist Bloc. The second and third largest partner
positions of the Middle East/North Africa region were closely shared by

45
two regions, the rest of Western Europe and the Communist Bloc. Exports
from the Middle East/North Africa region to the rest of Western Europe
represented 10.1% in 1966, 15.9% in 1976, and 17.1% in 1986. Exports to
the Communist Bloc region represented 11.6%, 29.0% and 15.3% in the same
years. Exports from the Middle East/North Africa region to the United
States and Canada decreased from 1966 to 1976; however, exports to these
countries have been increasing in recent years.
United States exports increased at a rate of 2.25% a year from 1966
to 1986. In relative terms, United States participation in world trade of
fresh oranges showed about the same level as 1966. Total United States
trade represented 7.4% of total world trade in 1966, increased to 9.2% in
1976 and decreased to 7.7% in 1986 (see Table 2.3). With intraregional
trade not considered, the relative importance of the United States trade
in the world trade increased. Table 2.3 shows that United States trade
represented 7.6% in 1966, 10.8% in 1976, and 8.5% in 1986. In relative
terms, these percentages show the United States to have been the third
largest exporter, exceeded by the Mediterranean-EC and the Middle
East/North Africa regions. In absolute terms, the Mediterranean-EC and
the Middle East/North Africa exports were 6.9 and 2.6 times the United
States exports, respectively, in 1986.
The relative importance of the United States partners has been
changing through the years. The major United States partner in 1966 was
Canada. Exports to Canada accounted for 54.3% of United States fresh
exports that year (see Table 2.4). The second largest partner was the EC
with 21.7% and the third largest was the Far East with 14.9%. Latin
America, rest of Western Europe, Oceania, and the Communist Bloc absorbed

46
2.5%, 3.8%, .9% and 1.9%, respectively. By 1976, Canada represented
41.1%, the EC region stayed almost the same, and the Far East region
absorbed 31.5% of the total United States fresh exports. In 1986, United
States exports to the Far East reached 64.9% of its total volume,
representing an important shift of the United States export partners. The
second largest partner was Canada, with 30% of the total volume. The EC
was no longer significant for the United States exports, given that it
represented only 2.3% in the same year. The rest of the regions also
decreased their participation relative to previous years.
Table 2.3 shows that exports from Latin America have doubled in
absolute terms in the last two decades. However, exports did not increase
from 1966 to 1976, which implies that the increase took place during the
last ten years. Total world trade participation of Latin America passed
from 3.1% in 1966 to 2.4% in 1976 and 4.3% in 1986. With intraregional
trade excluded, its participation in world fresh trade increased to 4.6%
in 1986. Note that Brazil generally does not export fresh oranges.
The major export market for Latin American product has been the EC
region, which in 1966 absorbed 67.2% of the total product exported (see
Table 2.4). This percentage decreased to 38.1 in 1976 and increased to
77.6 in 1986. The United States was the second largest market for Latin
America exports in 1966, with 22.5% of the total export level. This
percentage increased slightly in 1976 and decreased to 9.8 in 1986. The
third largest market for Latin America was the Communist Bloc, which took
most of the reduction shown in the EC region during the 1970s and part of
the United States share in the 1980s. Latin American exports to Canada
and the Middle East/North Africa have been increasing, especially in the

47
last decade. Some countries of the Middle East/North Africa region
utilized the fresh product to produce Frozen Concentrated Orange Juice
(FCOJ). The rest of Western Europe is another important market for the
Latin American product. An interesting issue about this region is that
its percentage of participation has not changed significantly over the
years.
The rest of Africa used to be the third largest exporter of the
world, but its share of the market has been decreasing, especially from
1976 to 1986. With intraregional trade considered, this region's share of
the world's export market was 7.6% in 1966, 5.7% in 1976, and 3.9% in 1986
(see Table 2.3). Given that most of its trade was external, these
percentages increased to 7.8, 6.5, and 4.2, respectively, when only
interregional trade is considered. The region's share of the market
indicates that it occupied the fifth position relative to the other
regions in 1986.
The major export market for the rest of Africa was the EC region.
This region represented 81.4%, 70.5% and 85.4% of the total rest of Africa
exports in 1966, 1976, and 1986, respectively (see Table 2.4). The second
most important partner was the rest of Western Europe, which absorbed
6.5%, 7.4%, and 8.7% of total exports in the same years. The rest of
Africa exports to Canada represented 6.1% in 1966 but decreased to 0% in
1986. In that year, the Far East region was the third largest market for
the rest of Africa. Exports to that region represented 5.9% in 1966, .4%
in 1976, and 4.3% in 1986. During the 1970s, exports from the rest of
Africa to the Middle East/North Africa region increased sharply and later
decreased.

48
The Far East includes the Asian continent except for the Middle
Eastern countries. This region's intraregional trade was more intense
than its interregional trade. Its participation in total exports was
higher in relative terms with intraregional trade considered (see Table
2.3). The Far East region's share of total world exports was 1.9% in
1966, 2.8% in 1976, and 2.1% in 1986. With intraregional trade not
considered, these percentages decreased to .7, 1.5, and .9, respectively.
China is one of the countries with the potential to become an important
exporter in this particular region and worldwide. Given the level of
exports of this region, it occupied position six among the 11 regions
considered in 1986.
The Far East major partners were Canada and the Middle East/North
Africa. In 1966, Canada was the major partner with 56.9% of total Far
East exports. In 1986 this percentage decreased to only 25.7. The share
of the Middle East/North Africa increased from 40.4% in 1966 to 70.5% in
1986 (see Table 2.4). The United States participation has been increasing
slowly through the two decades, passing from .9% in 1966 to 3.5% in 1986.
With these exceptions, the rest of the regions were not major partners of
the Far East.
Oceania held the seventh place relative to the rest of the regions
considered. Table 2.3 shows that, with intraregional trade included, the
percentages representing this region's participation in world total
exports were .7 in 1966, .2 in 1976, and .9 in 1986. Excluding
intraregional trade, these percentages decreased slightly to .4, .2 and
.8, respectively. This suggests that Oceania's intraregional trade was
relatively more important than its external trade.

49
In 1966, the major partners of Oceania included the Far East and the
EC regions with 65.4% and 24.6% of exports, respectively (see Table 2.4).
In 1986, the main partners were the Far East and the Middle East/North
Africa regions with 71.4% and 15.4%, respectively. The EC share decreased
from 23.9% in 1976 to 5.0% in 1986. Canada's share of Oceania's exports
was .4% in 1966, 18.5% in 1976, and 5.9% in 1986. Similarly, the rest of
Western Europe sharply increased its participation in the 1970s from 3.1%
in 1966 to 21.1% in 1976. In the 1980s, this percentage decreased again
to the 1966 level. The rest of Africa was another important partner of
Oceania's exports with 2.2% in 1966, .4% in 1976, and 1.5% in 1986.
EC production of oranges is relatively small and mainly concentrated
in southern France. Nevertheless, trade data reveal some intraregional
trade and a small amount of external trade. Including intraregional
trade, world export participation of the region was .9% in 1966, 2.4% in
1976, and 3.6% in 1986 (see Table 2.3). With intraregional trade
excluded, these percentages decreased to .1, .2, and .4, respectively.
This indicates that the EC region occupied position number eight relative
to the rest of the regions with regard to world fresh orange export share
in 1986.
The main partner of EC exports is the rest of Western Europe, with
51.5% in 1966, 34.5% in 1976, and 74.2% in 1986 (see Table 2.4).
Interestingly, in 1966, 36.8% of the EC's total exports were directed to
the rest of Africa and, in 1976, 52.6% were sent to the Middle East/North
Africa. In both cases, the participation of these regions rapidly
decreased to 3.6% and .5%, respectively, in 1986. The rest of the regions
were not significant partners to the EC except for the Communist Bloc.

50
This region's participation was 6.8% in 1966, 4.6% in 1976, and 4.9% in
1986.
The Communist Bloc has been increasing its participation in world
total exports from almost zero in 1966 to .6% including intraregional
trade and .3% excluding intraregional trade in 1986 (see Table 2.3). With
these percentages, the Communist Bloc held position number nine concerning
world exports of fresh oranges in 1986.
The Communist Bloc region has two principal partners, the EC and the
rest of Western Europe (see Table 2.4). The EC and the rest of Western
Europe absorbed 11.2% and 88.8% in 1966, 97.3% and 2.6% in 1976, and 94.1%
and 2.1% in 1986, respectively. The rest of the Communist Bloc exports in
1986 went to the Latin America region.
Finally, Canada and rest of Western Europe are not exporters of
fresh oranges. Weather conditions in these regions do not allow them to
produce oranges (see Table 2.3). However, trade data revealed some
exports out of those regions. Most of that trade was related to re
exports reported as exports.
Region Perspective
In this section, the discussion will be based on importers'
viewpoint. Once again, the relative importance of each one of the regions
will be set forth and then trade flows will be discussed.
As shown in Table 2.3, the leading importer of fresh oranges
including intraregional trade was the EC region, with shares of 68.8% in
1966, 54.5% in 1976, and 64.3% in 1986. Considering only interregional

51
trade, the shares for the same years were 69.9%, 61.3%, and 67.3%. These
percentages show that EC trade with other regions was more important than
its own within-region trade.
Table 2.5 shows that, in 1966, 47.1% of the total EC imports came
from the Mediterranean-EC region, while in 1986 this percentage reached
67.4. The second largest exporter to the EC region was the Middle
East/North Africa region with 38.3% in 1966 and 21.2% in 1986. The third
largest exporter was the rest of Africa region with 9% in 1966 and 5.4% in
1986. Other important exporters to the EC region included Latin America
and the United States. These two regions' EC market shares were 2.9% and
2.4% in 1966, and 5.4% and .3% in 1986, respectively. The major portion
of the EC market growth has been captured through the years by the
Mediterranean-EC region. Latin America was the only other region whose
share of the EC market grew in the last two decades. The rest of the
regions' exports to the EC were minimal.
The second largest importer of fresh oranges was the rest of Western
Europe with 11.0% in 1966, 8.8% in 1976, and 10.2% in 1986 (see Table
2.3). With only interregional trade used, the percentages increased to
11.3, 10.2, and 11.3 for the same years.
As shown in Table 2.5, the leading exporter to the rest of Western
Europe region was the Mediterranean-EC region, with 59.9% in 1966, 40.8%
in 1976, and 58.4% in 1986. The second largest exporter to this region
was the Middle East/North Africa with 31.7%, 51.5%, and 34.0% for the same
years. Another important exporter to the rest of Western Europe was the
rest of Africa with 4.4% in 1966, 4.8% in 1976, and 3.3% in 1986. Even
though the EC was not a major producer of fresh oranges, it was the fourth

52
major exporter to the rest of Western Europe region in 1986. Latin
America increased its participation in this market passing, from .8% in
1966 to 1.1% in 1986. The United States, once the fourth major exporter
to the region, was a very small participant in the rest of Western Europe
fresh orange trade. The rest of the regions' exports to the rest of
Western Europe were relatively small.
The third largest importer was the Communist Bloc with 9.1% in 1966,
14.3% in 1976, and 8.3% in 1986 (see Table 2.3). With only interregional
trade considered, the percentages increased to 9.3, 15.4, and 8.7,
respectively.
The major supplier of fresh oranges to the Communist Bloc was the
Mediterranean-EC with 52.8% in 1966, 32.1% in 1976, and 57.1% in 1986 (see
Table 2.5). The second largest exporter to the region was the Middle
East/North Africa with 44.2%, 61.9%, and 39.6%, respectively. The third
largest exporter was the Latin America region with 1.3%, 4.4%, and 3.0% in
the same years. During the 1970s, the Communist Bloc countries
drastically increased their consumption and the deficit was mainly
supplied by the Middle East/North Africa region. During the 1980s,
consumption went back to the original trend. The United States exported
1.6% of the Communist Bloc's total imports in 1966, and 1.5% in 1976. In
1986, the United States did not export fresh oranges to the Communist Bloc
region. The rest of the regions were not very important relative to total
Communist Bloc's imports of fresh oranges.
The three principal regions mentioned above have been relatively
stable in their participation in the world's fresh orange imports in the

53
last two decades. As a whole, they represented 90.5% in 1966, 86.9% in
1976, and 87.3% in 1986 of total world imports (see Table 2.3).
The Far East has been consistently growing as an importing region in
the last two decades. It passed from 3.4% in 1966 to 4.8% in 1976, and to
7.1% in 1986 (see Table 2.3). With intraregional trade omitted, these
percentages decreased to 2.3, 3.8 and 6.4, respectively. This shows that
trade among countries belonging to the Far East region was important
relative to the rest of the world's trade with the same region.
The United States was the leading exporter to the Far East region in
the period considered. Its exports represented 49.5% in 1966, 90.6% in
1976, and 86.2% in 1986 (see Table 2.5). The second exporter to the Far
East was Oceania, with 11.5%, 1.9%, and 8.4%, respectively. The third
exporter was the rest of Africa. However, its participation has been
decreasing, from 19.9% in 1966 to 2.9% in 1986. The fourth exporter to
the Far East was the Middle East/North Africa region, whose participation
has also decreased from 17.8% in 1966 to 1.8% in 1986. The Mediterranean-
EC region represented the fifth exporter to the Far East region, with a
consistent participation in the market of only .5%. Latin America's share
of the market decreased from .8% in 1966 to .1% in 1986. It is clear from
these numbers that the United States was the only exporting region whose
market share grew in the Far East.
The Middle East/North Africa region was another significant importer
of fresh oranges. Its participation grew from .6% in 1966 to 11.8% in
1976, but decreased later to 5.2% in 1986 (see Table 2.3). The table
shows that the percentages excluding intraregional trade were .3 in 1966,

54
3.0 in 1976, and 1.1 in 1986. Therefore, the principal trade of this
region was among the countries constituting the region.
The major exporter to the Middle East/North Africa countries was the
Far East region, with 85.7% in 1966, 31.2% in 1976, and 53.8% in 1986 (see
Table 2.5). The second principal exporter to this region was the
Mediterranean-EC with 9.0% in 1966, 22.8% in 1976, and 29.7% in 1986. The
third exporter was Oceania with 5.1% in 1966, 0% in 1976, and 10.3% in
1986. Even though exports from Latin America appear insignificant
compared to other exporters to the Middle East/North Africa, they have
been growing very rapidly in the last few years, passing from 0% in 1966
to 5.9% in 1986.
Canada was an important importer of fresh oranges. During 1966, its
imports represented 5.2% of the world's trade. This percentage decreased
to 4.5 in 1976 and 3.4 in 1986 (see Table 2.3). With only interregional
trade considered, these percentages increased slightly to 5.3 in 1966 and
1976, and 3.7 in 1986.
The major supplier of fresh oranges to Canada was the United States,
with 77.8% in 1966, 84.7% in 1976, and 68.0% in 1986 (see Table 2.5). The
second largest exporter to Canada was the Middle East/North Africa region,
holding 4.7%, 1.4%, and 11.8% for these years. The third major exporter
was the Far East region with 7.6% in 1966, 10.2% in 1976, and 5.9% in
1986. In the last few years, the Mediterranean-EC region, whose share was
insignificant during the 1960s and the 1970s, have increased their
participation in this market. In 1986, Mediterranean-EC supplied 10.2% of
the Canadian market. Latin America increased its share of the market from
.8% in 1966 to 2.9% in 1986.
Similarly, Oceania increased its

55
participation in recent years, passing from 0% in the 1960s to 1.2% in
1986. The rest of the regions were not very important with regard to
exports to the Canadian region.
The rest of the regions represented small percentages of total
imports in the world's fresh orange industry (see Table 2.3). The United
States import share was .8% in 1966, .7% in 1976, and .9% in 1986. These
percentages changed very little if only interregional trade were
considered. Major exporters to the United States were Latin America with
45.2%, Mediterranean-EC with 32.0%, and Middle East/North Africa with
19.7% in 1986. The Mediterranean-EC only had .8% and .6% share of the
United States market in 1966 and 1976, respectively, indicating that
Mediterranean-EC region's participation in the United States has been
growing rapidly in the last decade.
The Oceania portion of total world imports was .4% in 1966, .2% in
1976, and .4% in 1986 (see Table 2.3). These percentages switched to .2
each reported year if only interregional trade were included.
The exporter with the major portion of the Oceania region's market
was the United States, with 40.4% in 1966, 99.6% in 1976, and 99.9% in
1986 (see Table 2.5). Middle East/North Africa and Latin America regions
used to have an important share of the Oceania market, reaching 29.6% and
25.9%, respectively in 1966. These regions lost their portion of the
market to the United States in the 1970s. The rest of the regions were
not major exporters to Oceania.
The rest of Africa's share of total world imports was .3% in 1966,
and .2% in 1976 and 1986 including intraregional trade (see Table 2.3).

56
If intraregional trade were excluded, these percentages changed to .1 for
the last two years reported.
The four major suppliers of fresh oranges to the rest of Africa were
the Middle East/North Africa with 83.0% in 1966, 71.2% in 1976, and 70.5%
in 1986; the EC with 11.8%, 9.3%, and 12.5%, respectively; Oceania with
3.4%, .7%, and 8.9%, respectively; and the Mediterranean-EC with .1%,
14.1% and 8.0%, respectively (see Table 2.5). The United States share of
the rest of Africa market was .6% in 1966. However, the United States
lost its share totally by 1986.
Latin America's portion of total world imports was .3% in 1966, .4%
in 1976, and .1% in 1986 (see Table 2.3). Given that most of its trade
was among countries of the region, these percentages decreased to .2 in
1966 and to 0 in 1976 and 1986. Imports in Latin America came from the
United States in the 1960s and 1970s (see Table 2.5). In 1986 the United
States share was only 68.0% of total imports. The rest of the product
came mainly from the Communist Bloc with 33.4%, the EC region with 16.1%,
and the Mediterranean-EC with 4.5%.
The Mediterranean-EC region has only a small share of total world's
fresh orange imports. Imports reached .2% in 1986 with and without
considering intraregional trade (see Table 2.3).
Conclusions
In summary, it is possible to describe most of the world production
and trade flows of the fresh orange industry with few regions. On the
production side, the major producers of oranges were Latin America, Far

57
East, United States, Mediterranean-EC, and Middle East/North Africa.
Latin America and United States had high percentages of processed
utilization. The Far East had an intense within-region trade. Therefore,
as shown above, large orange productions did not necessarily imply high
participation in interregional fresh orange trade.
On the supply side, the major exporters were the Mediterranean-EC,
Middle East/North Africa and United States. However, United States share
of total fresh exports was small compared to the other two regions. The
Mediterranean-EC region includes Spain, Greece, Italy, and Portugal. The
Middle East/North Africa includes the Middle East and the North African
countries.
The Middle East/North Africa region has been losing its share of the
world market to the Mediterranean-EC in the last few years. It is clear
that the leading world exporter was the Mediterranean-EC region. The
United States, once a major exporter to the European markets, shifted to
the Far East and Oceania markets. United States share declined in most
markets, with the exceptions mentioned above. Finally, the Latin America
region increased its share of the total market in last two decades.
On the demand side, the major importer was the EC region which
includes the EC countries except for Spain, Greece, Italy, and Portugal.
The second largest importer was the rest of Western Europe, which
represents the rest of the Western European countries. The third largest
importer was the Communist Bloc, among which the major importers were the
Eastern European countries.

CHAPTER 3
LITERATURE REVIEW
International Agricultural Trade Models
Several models or approaches to study international trade have been
developed in the last two decades. These models were developed mainly due
to the need for knowledge and understanding of increasing world trade.
Thompson (1981) presented an interesting survey of new developments in
international agricultural trade models. In his document, each model was
reviewed in three sections: a historical survey, an evaluation, and a
summary and implications section. The different modeling approaches were
divided into two basic groups determined by the number of regions
considered in the model. The two groups were two-region models and
multiple-region models of agricultural trade. The latter was further
divided into three groups: non-spatial price equilibrium, spatial price
equilibrium, and trade-flow and market-share models.
A different classification system for international trade models was
developed by Thompson and Abbott (1982) Each modeling approach was
grouped based on the assumptions made about the homogeneity of the
commodity traded. The two major categories identified in their research
were single homogeneous commodity models and multiple-product models. The
single homogeneous commodity models were divided into three groups: non-
spatial price equilibrium, spatial price equilibrium, and two-region
58

59
models. The multiple-product trade models were also divided into three
groups: general equilibrium (including agricultural and non-agricultural
products), multiple related commodity products (including only
agricultural products), and differentiated product models (differentiated
by place of origin). The two-region and the general equilibrium models
were special cases of non-spatial price equilibrium models. Thompson and
Abbott's (1982) classification procedure added important insights into the
discussion about new developments in international agricultural trade
models. The major contribution was their extensive treatment of and
emphasis on the characteristics of the products traded and how consumers
perceived them.
In the following discussion, Thompson's (1981) approach will be
followed. His classification was basically the same as the one presented
in Thompson and Abbott's (1982) investigation. The most important
difference between the two studies was the emphasis that the latter
researchers gave to product differentiation.
The first type of model covered by Thompson (1981) was the
two-region model. The model divided all countries of the world into two
groups, the country of interest and the rest of the world. This version
was basically a domestic agricultural sector model enlarged with
exogenously driven exported or imported quantities. Export equations or
excess demand equations were developed for the rest of the world. The
model included linkages between the domestic and world prices to reflect
the simultaneous determination of domestic consumption, supply, and prices
with the rest of the world. The models did not take into consideration
trade flows (destination) but instead accounted for the net trade between

60
the country of interest and the rest of the world. They did not provide
information on demand and supply for individual foreign regions or on the
share of the market that any particular country has in a specific region.
Without knowledge of the structure of supply and demand in each
major trading region, it is impossible to say how the excess demand
function will change given an exogenous shock or a change in policy. It
is then very difficult under the two-region models to evaluate the impact
of shocks or policies in a given country. Such models do, however,
provide a good framework to analyze domestic farm and trade policies.
According to Thompson (1981) multiple-region world trade models
were developed to answer broader questions regarding the impact of
exogenous shocks and policy changes for trading regions in the world.
They also provide information about the market share of each region by
destination. The non-spatial price equilibrium models treat the
interrelations among trading regions by assuming that the world market
price is determined simultaneously by the demand-supply balance in all
trading regions such that the world market clears. Solution of the model
gives the world market prices and the net trade for each region, but it
does not provide any information on source or destination of trade flows.
Multiple-region world trade models allow for the introduction of
transportation costs, tariffs and nontariff barriers, and other policy
variables through the price linkage equations. These models are for many
reasons an improvement over the two-region models, since they endogenously
determine the demand and supply in each of the trading regions. However,
they usually have an important drawback. The price linkage frequently
used is not consistent with the spatial price equilibrium theory. This is

61
so because in some cases a unique world price is assumed and in other
cases a base country or region price is used and linked with the rest of
the regions by the transportation cost. The model ignores the fact that
some regions may not trade at all with the base region. Solutions to
these models are obtained by solving an econometric simultaneous system of
equations.
The second type is the spatial price equilibrium models. These
models differ from the non-spatial and the two-region models in the fact
that they consider endogenous trade flows and market shares. Prices are
linked only between those pairs of countries that actually trade with each
other. The rest of the characteristics are similar to the ones mentioned
for the non-spatial equilibrium models, except for the solution method.
They usually follow a quadratic programming procedure for estimation.
None of the models described above can replicate all of the observed
trade flows since they are designed to predict trade flows of homogeneous
products (Grennes et al., 1977 and 1978; Thompson, 1981; and Thompson and
Abbott, 1982). If products are homogeneous, then price differences
between regions are given only by transportation costs and trade barriers.
Products may not be perfectly homogeneous and may be differentiated by
country of origin. Therefore, prices may vary between regions for reasons
other than transportation costs and trade barriers.
A serious formulation of a spatial price equilibrium model will be
to determine trade flows exclusively by minimizing the transportation
cost. According to Grennes et al. (1978) "nearly everyone who has
employed spatial models concedes that the world does not behave this way".
This situation is intuitively appealing, and indeed there is enough

62
empirical evidence that this may be the case for wheat (Grennes et al.,
1977 and 1978; Thompson, 1981) and other agricultural products. Spatial
price equilibrium models have few capabilities except for the weak and
incomplete explanation of trade flows given the problems mentioned above.
Trade-flow and market-share models are the third type of
multiple-region models considered by Thompson (1981). These models were
developed to account for the observed variation in trade flows more
adequately than do the spatial equilibrium models.
Taplin (1967) and Johnston (1976) in a partial sense surveyed world
trade models concerned primarily with trade flows. They studied the ones
that analyzed the structure of world trade and the short-run trade
fluctuations among countries. In his paper, Taplin classified them in two
categories: the ones that have separate functions for total exports and
total imports but do not attempt to estimate the individual flows between
countries; and the ones that look at individual flows directly.
In the first part, Taplin's discussion went from an import-export
matrix developed by the League's Network of World Trade (1942), passing by
Woolley's (1965) transactions matrices on payments for trade, services,
and capital flows, to Beckerman's (1956) input-output approach. These
studies provided important insight into the structure of the international
economy. However, they did not represent a formal model where hypotheses
could be tested, measured or forecasted.
The second part of Taplin's investigation continued with a survey
covering other studies (Tinbergen, 1962; Linnemann, 1966; Waelbroeck, 1962
and 1965) where individual trade flows (from the import-export matrix)
between countries were considered to be a function of income, population,

63
trade preference, and distance variables. In these models, prices were
normally omitted given that cross-section models were used, with data at
the same point of time. Prices were assumed not to change. These models
did not capture shifts or changes of trade which might develop in the long
run because of more complicated interrelationships among prices, income,
and imports.
Taplin continued his study by reviewing four different transmission
models that tried to establish the main relationships between the level of
domestic economic activities in the various countries and their
international transactions. The four models surveyed and reported by
Taplin were: Metzler (1950) who focused on changes in investment; Neisser
and Modigliani (1953) on income and capital flows; Polak (1954) on
autonomous investment and price changes; and Rhomberg (1966) and Rhomberg
and Boissonneault (1964) who focused on income, prices, and capacity.
Rhomberg and Boissonneault (1964) developed a trade-flow and market-share
model that considered three regions, the United States, Western Europe,
and the rest of the World. An aggregated commodity called merchandise,
including all commodities traded among the regions, was defined and used
to estimate income and price elasticities.
Taplin concluded that a model was needed that incorporates the type
of disaggregation possible with a constant share approach and the
flexibility and economic content provided by a transmission model. Taplin
also proposed a three stage procedure to accomplish his recommendations:
consider the import demand for 10 to 12 regions for six good's classes;
determine what share of the import market the other countries have in
supplying the given country's imports; the export-supply schedules should

64
tie into the model. These conclusions provided guidelines for continued
research in trade-flow and market-share models during the latter part of
the 1960s.
Trade-flow and market-share models are based on the idea that
products are differentiated by country of origin. Three alternative
solution approaches exist: mechanical procedures that transform trade flow
matrices from one year to the next without regard for price; econometric
models designed to explain one or more elements of the trade flow matrix
(an example is Ward, 1976); and modified spatial equilibrium models that
take into account that products are differentiated by country of origin.
The latter implies that the elasticity of substitution is less than
infinite (examples are Hickman and Lau, 1973; Grennes et al., 1978;
Johnson et al. 1979; Sarris, 1983 and 1984; Sparks, 1987; Penson and
Babula, 1988; Deardorff and Stern, 1986). None of the examples, except
for Sparks and Deardorff and Stern, used a simultaneous equation approach
to estimate the world trade model. Hence, the results obtained suffer
from simultaneity bias (Maddala, 1977, p. 231-251).
The modified spatial equilibrium model approach has been used to
estimate a total import demand equation for each importing region and
separate market share equations for each region. This approach rests on
the assumption that products have unique characteristics distinguishing
them from similar products of other exporters. Most studies mentioned
above have proved that consumers view goods of the same kind from
different suppliers as imperfect substitutes. This is especially true in
agricultural trade, where quality and variety characteristics, national
factors, variations in harvest time, and monopolistic competition are

65
normally present. Therefore, different countries faced different
elasticities that may vary when market shares differ (Grennes et al. ,
1977) .
Armington (1969a, 1969b, 1970a, 1970b, 1973) developed the theory
for market share demand studies which considered goods differentiated by
place or origin. Most market-share demand studies have used this theory
because of important variations obtained in price and income elasticities
among suppliers in the foreign markets (examples are Sirham and Johnson,
1970; Ito et al. 1988; Lin et al. 1988). Later, Rhomberg (1970)
concluded that a complete demand and supply model for a world trade and
payments model could also be developed following Armington's approach.
Armington assumed a weakly separable utility function, so that
consumers' decision process may be viewed as occurring in two stages
(Varian, 1984). Equations can be derived that relate a particular trade
flow between two countries to the importing country's index of total
imports and a price ratio or relative price. Each region's market share
of a commodity may be affected by changes in the size of the market of
destination even if relative prices remain unchanged. The price ratio is
between the price of the exporting country and an average of the import
prices of the same type of product from other origins in the importing
country. The total quantity of a commodity to be imported is first
determined, and then the quantity is allocated among the competing
suppliers.
Armington assumed that the total quantity of the product imported is
a constant elasticity of substitution (CES) index of the quantities
imported from the regions of origin. The assumption was made to simplify

66
the model and reduce the number of parameters to be estimated, especially
when the number of trading regions is large.
Under these assumptions, the cross-price elasticities between all
pairs of regions need not be estimated, since they can be obtained from
the estimated price elasticities and the estimated "constant" elasticity
of substitution (Learner and Stern, 1970). The CES assumption is highly
restrictive. In fact, the model assumes that products are differentiated
by country of origin and at the same time assumes that the elasticities of
substitution are constant and equal between all pairs of exporting regions
in all markets. Arrow et al. (1961) developed the general properties of
the CES production function.
Winters (1984) criticized these assumptions on the use of the CES
functional form. Winters accepted the initial assumptions of separability
among commodities (e.g., food and machinery), while within each commodity
group the domestic and foreigner suppliers were treated as non-separable.
However, the adoption of the CES made them homothetic and separable over
all pairs of sources. Winters concluded that "the separability of
domestic and foreign supplies essentially slipped in by the back door,...,
rather than as a necessary consequence of two-stage budgeting". Winters'
empirical results rejected the assumptions of homotheticity and
separability after testing them using the AIDS model (Deaton and
Muellbauer, 1980).
Alston et al. (1990) has recently criticized Armington's approach.
His research shows that the assumptions of separability and homotheticity
with trade data for the cotton and wheat markets were also empirically

67
rejected using the AIDS model. They also recognized the problems with the
AIDS model.
The restrictiveness of the assumptions were recognized earlier by
Resnick and Truman (1973). They relaxed several assumptions of
Armington's model, especially the one that the elasticities of
substitution need to be constant and identical between all pairs of
suppliers to each market. They specified a multi-stage decision process
instead of Armington's two-stage procedure. Again, total imports were
determined first and then imports from a sequence of successively smaller
geographic regions were determined.
Artus and Rhomberg (1973) also recognized the problem with the
assumptions and replaced the CES index function. They used the constant
ratios of elasticities of substitution and homogeneous (CRESH) index
functions developed by Mukerji (1963) and Hanoch (1971).
Sparks (1987), following Artus and Rhomberg's work, used the
constant ratio of elasticity of substitution (CRES) index which makes the
model somewhat less restrictive. This assumption implies that the
elasticity of substitution for all the products in a market or region i
vary by a constant proportion, but the substitutability between products
need not be the same. This assumption increases the flexibility of the
model but also increases the computational complexity. The model was
applied to a highly aggregated commodity (vegetables). In this case, the
basic assumption of Armington's model, goods distinguished by place of
production, seems less applicable given that the aggregated commodity will
be composed of several goods. The model explained that trade flows can

68
reflect differences due to commodity composition as well as differences
due to country of origin.
Trade-flow and market-share models represent a major improvement
over the other models developed to study international trade, since they
can more readily depict observed trade flows. The assumption that
products are differentiated by country of origin and prices may vary
between regions for reasons other than transportation costs and trade
barriers is intuitively appealing. Furthermore, Armington's
simplification by the introduction of an import quantity index function
is, in many cases, a necessary condition to operationalize the model and
obtain as much information as possible from the trade flows. As will be
shown later, the Armington model provides several practical solutions for
dealing with a large number of equations and parameters.
Trade Models: The Orange Industry
The fresh orange industry has been studied many times, usually in
the context of national markets. A few studies have been developed in the
international trade of fresh oranges. In addition, none of the research
developed so far considered a complete world trade model for this
particular good. Most of the studies have been either partial or
descriptive. One of the earliest international trade documents is a
descriptive study developed by the U.S. Department of Commerce (1940),
which showed citrus world production and trade statistics and trends.
Before the 1950s, little demand estimation for citrus fruits
existed. More attention has now been devoted to this economic area by the

69
Florida Agricultural Experiment Station and by the Florida Department of
Citrus (FDOC). As reported by Chapman (1963), the first major step in
this area was the work on experimental pricing techniques applied to the
orange demand analysis developed by Godwin and Powell during the 1950s.
Chapman (1963) and Godwin et al. (1965) developed a study on demand and
substitution relationships for California and Florida Indian River and
Interior Valencia fresh orange market. Their research was basically
concentrated in the U.S. market and focused on questions regarding own-
price elasticities and cross-price elasticities between the three regions'
in the Grand Rapids, Michigan market.
Dean and Collins (1967, 1968) studied the effects of the European
Community (EC) tariff policies in a model of world trade for fresh
oranges. Their paper included a summary of world production, consumption,
and trade of fresh oranges. Projections of orange production and
consumption, estimates of transportation costs, possible future tariffs,
and income and price elasticities of demand in the EC for 22 regions were
also included. The price elasticities were estimated at the import demand
level, i.e., at the location of consumption, but before retail margins
were added to the wholesale price. Transportation costs as well as
tariffs and any special import taxes were included in determining the
wholesale price level. Using a transportation model analysis, the impact
of possible future tariff policies in the EC was procured on producer and
consumer prices in each of the major countries and on trade flows.
Finally, using the results obtained in the different tariff scenarios, the
welfare effect on consumers and producers was also captured. The major
implication of this document and the ones by Chapman (1963) and Godwin et

70
al. (1965) is that it is possible to argue that consumers actually see
products of the same kind coming from different regions as non-perfect
substitutes.
Weisenborn et al. (1970) estimated the price-quantity relationships
at the processor or packer FOB level in foodstore, institutional, and
export market channels for Florida oranges and orange products. The
products included fresh and processed oranges. As reported by Weisenborn
et al., virtually no previous demand analysis had been completed for the
institutional and export sectors at that time.
Prato (1970) used the concept of separability to separate food from
non-food items. Once the demand equation was defined for only food items,
he showed that the correlation between first differences in the prices of
orange products and first differences in the prices of each of the other
food items were not significantly different from zero. Therefore,
individual demand equations for fresh and processed oranges without the
introduction of other food item prices could be defined. As reported by
Prato, research findings appear reasonable when compared with estimates
derived using other and more conventional approaches.
Tang (1977) studied the world demand for United States fresh
grapefruit in four markets: the United States, Japan, Europe, and Canada.
In his research, Tang identified and measured the effects of the different
factors that affect domestic and export demand in order to determine the
optimal allocation of United States fresh grapefruit to the domestic and
export markets. The results were used to simulate the grapefruit industry

71
to ascertain its performance to changes in the major factors. The system
of equations was estimated using a seemingly unrelated regression (SUR)
model.
Nelson and Robinson (1978) developed a model to analyze retail and
wholesale fresh navel orange demand under marketing order policy. Two
important issues were raised in this study: apples and bananas could be
used as substitute products for fresh oranges; and the demands for fresh
and processed oranges are independent. The first issue was raised
previously by Matthews, Womack, and Huang (1974) with encouraging results.
Prato (1969) found that demands for fresh oranges and concentrate are
independent, at least in the winter season.
Ward (1981) applied time-varying parameters (TVP) to analyze the
welfare impact and economic forecast based on a better understanding of
the economics of the EC fresh orange industry. This study was especially
important at that time given the plans of enlargement of the EC to include
Greece, Spain and Portugal. To support the use of TVP, Ward argued that,
given the evolution of the EC and its related regulations, it is possible
to hypothesize that some adjustments in the demand parameters are likely
to have occurred over the decades since the early 1960s. He also
estimated the model using Ordinary Least Squares (OLS) and the results
were compared. It was clear that the use of TVP performed better that the
simple OLS estimation technique.
McCabe (1982) estimated a model to determine the characteristics of
fresh citrus consumers. The major objective was to ascertain how
demographic and household characteristics affect purchase decisions and to
determine its relationship with product prices.

72
Wardowski et al. (1986) recently edited a book that includes a
descriptive analysis of world production practices and trends and a long
term view of fresh citrus trade. An interesting discussion on trade flow
and market share of imports and exports was presented. Global trade for
the 1980s was projected, based on the assumption that trade will growth
one third less rapidly than in the previous decade, given special
assumptions for each citrus product. Individual country/region
projections were based on historical trends in per capita availability,
where such trends were evident or trends on total imports were estimated.
The use of trends in projecting import demand was based on the assumption
that future levels of economic factors that determine demand will follow
historic trends. No demand estimation was pursued to determine trade flow
and market share in this study.
Lee and Fairchild (1988) used a SUR technique to study the
relationship between exchange rates and foreign demand for United States
fresh grapefruit. The results showed that exchange rates played a major
role when studying export demand relationships and the United States fresh
grapefruit has more than one export market, with markets responding
differently to price changes. These results will be used later to define
the model to be estimated.
Lee et al. (1990), using the absolute version of the Rotherdam
model, studied the Japanese citrus products market. The study used fresh
bananas and pineapples as substitutes for fresh citrus products. One of
the major conclusions of the study was that United States fresh grapefruit
exports compete against imports of bananas and pineapples for Japanese
import dollars. In the case of fresh oranges, the results were not

73
consistent with the expected signs; especially in the case of pineapples,
which turned out to be a complement for fresh oranges, an unexplainable
result as reported by Lee et al. This article and the one by Nelson and
Robinson (1978) have interesting insights that will be considered later in
order to define the best substitute products for fresh oranges.
Even though the present study will not deal directly with the
processed orange industry, a few comments on the literature reviewed will
be made. Priscott (1969) developed a model to estimate the demand for
citrus products (juices) in the European market. One of his major
findings was that there is substantial substitution among products of
specified countries, reinforcing once again the need to differentiate
products by place of origin. Weisenborn et al. (1970) used the theory of
price discrimination to determine the optimal market allocation of Florida
orange production for maximum net returns. To solve the problem of price
discrimination, quadratic programming and calculus with LaGrangean
multiplier techniques were used. Malick (1980) used a simultaneous
equation model of the Florida retail orange-juice marketing system to
forecast changes in the FOB price and retail movement of frozen
concentrated orange juice (FCOJ). Irias (1981) developed an econometric
model to study international trade of FCOJ among three regions, the United
States, Brazil, and Europe. Margoluis (1982) developed a model to
estimate implicit prices for juice and drink characteristics using hedonic
price functions. Ting (1982) developed a model to test the existence of
asymmetric price response in the irreversible demand functions for citrus
juice products.

74
Most of the work has been concentrated on the United States domestic
market analysis and in specific econometric models designed to explain one
or more elements of the international trade flow matrix and markets. The
studies are usually related to the United States product behavior in
Canada, Europe, and Japan. In most cases, the estimation has been pursued
using single-equation estimation and, in a few cases, using SUR. The
fresh orange industry has not been studied in a full simultaneous spatial
equilibrium world trade model modified to take into account that products
are differentiated by country of origin and therefore are not perfect
substitutes. The results presented in many of the articles and books
reviewed regarding trade of different commodities, and specifically fresh
and processed oranges, strongly support the conclusion that fresh citrus
coming from different countries (or regions) are perceived as different
products by consumers. The main objective of the present study will be to
develop and estimate a modified spatial equilibrium world trade model for
the fresh orange industry. The model will be used to analyze the impact
of different trade policies and economic factors affecting the demand for
fresh oranges in different regions of the world.

CHAPTER 4
WORLD FRESH ORANGE TRADE MODEL
Introduction
With international trade models it is frequently assumed that goods
of a given kind supplied by different (national) sellers to a single
country are perfect substitutes in the final market. With this
assumption, consumers differentiate goods only by kind, and there is no
evident difference between products of the same kind supplied from
different sellers. It also implies that the elasticities of substitution
between suppliers are infinite, and that the corresponding price ratios
are constant (Armington, 1969a).
In general, fruits, and in particular fresh oranges, are expected to
be differentiated by place of origin. There are several varieties of
oranges, and regions have soil and climatic conditions favoring the
production of only a few varieties. Production seasons are highly
variable among regions and yield products at different times of the year.
For example, while the Northern Hemisphere countries harvest their oranges
from November to June, the Southern Hemisphere countries harvest their
fruit from June through October. In addition, product coming from
different regions even at the same time period could be perceived to have
distinctive quality features by the final consumer.
75

76
Under these circumstances, the theoretical model defined in this
section will be based on Armington's model of international trade
(Armington, 1969a). As previously mentioned, it is a modified spatial
equilibrium model that takes into account the concept that commodities are
differentiated not only by kind but by exporting region. Armington
distinguished commodities from products. For example, the term commodity
refers to a specific good such as fresh oranges, cotton or rice, or an
aggregated good such as fruits, meats or vegetables. On the other hand,
a product is a commodity exported from one region to another; i.e., fresh
oranges coming to France from the U.S. is a different product than fresh
oranges coming from Spain to the same country of destination.
The first basic assumption underlying this model is that consumers'
utility is weakly separable; therefore, the decision process may be viewed
as occurring in two stages. The first decision stage is to determined the
total level of consumption for each commodity known as "market demands".
This decision is usually based upon commodity prices, income levels,
substitute commodity prices, and other relevant economic variables. The
second step is to decide where to buy the product; i.e., given that the
total consumption level for each commodity has been determined, an
allocation among the different suppliers has to be made. These are known
as "product demands". The distribution among suppliers is based on the
commodity's total market demand and relative product prices.
The second basic assumption in Armington's model is that the
quantity index function used to represent quantities imported from the
regions of origin is linear and homogeneous. This assumption implies that

77
each region's market share of a commodity is influenced by changes in the
size of that market, even when relative product prices remain unchanged.
In the present study, 11 regions were defined. The regions were
selected consistently with the world orange industry and with particular
similarities among the countries included in a region. The regions were
the United States (US), Canada (CAN), Latin America (LA), Mediterranean-EC
(MED-EC), EC, rest of Western Europe (RWE), Middle East/North Africa
(ME/NA), rest of Africa (RAF), Far East (FE), Oceania (OCE), and Communist
Bloc (COMMB).
In the next section, a complete world fresh orange trade model is
specified. Demand and supply sides are included with equilibrium
conditions and price linkages set forth.
Fresh Orange Trade Model
Demand Side
The model was based on the two assumptions mentioned above. Two
stage budgeting is implied. Marginal rates of substitution between two
goods in a commodity group were assumed to be independent of goods in
other groups. In the orange industry, the rate at which consumers
substitute fresh oranges produced in one country for those produced in
another country does not depend on their purchases of other kinds of
fruits or other commodities. The first level of the two stage budgeting
is the consumers' decision to allocate their total income among the

78
different commodity groups available in the region. A percentage of that
income is allocated to the total market demand for fresh oranges.
In the general case, the utility function for consumers in region i
given n commodities is:
(4.1) Ut = U(Xt)
where
xi = (Xm,Xli2l ^nii>Xni2> is bhe total bundle
of commodities for region i, and m the total number of regions
or countries considered.
The first subscript for Xkij represents the commodity (n
commodities), the second represents the region of destination (m regions),
and the last subscript denotes the region of origin (m regions). Given
the assumption of weakly separable utility function or independence among
commodities in different groups and following Solow (1955-56) and
Armington (1969a), it is possible to write this utility function for
region i as follows:
(4.2) Ui = U' (Xu ,X2i Xni )
where
Xki. 4k (Xkil >Xki2> >Xkim) fr k=l,2,...,n.
Xki is the total market demand for commodity k in region i. The
dot represents the sum over all j's or regions of origin
including the domestic region i. The ^k represents certain
quantity index function of the product demands Xkij which
represent the demand for commodity k in region i coming from
region j where j-l,...,m.
Consumers maximize the utility function (4.2) subject to the budget
constraint given by

79
(4.3) INC, = 4 Zj (PkiJ XkiJ) = 5* (Pki. Xki.), k-1,2 n and
j-1,2,...,m
where
INCi is total expenditure (or income) for all commodities in region
i,
Pkij is the price for commodity k coming from region j in region i
(or products price) ,
Pki is the average price of commodity k in region i for
k-1,2,...,n,
Zjj Ej is the sum over all k (commodities) for all j (regions).
The resultant "market demand" equation for commodity k in region i
is a function of total income or expenditure, commodity k price, other
commodities' prices, and other relevant variables:
(4.4) Xki = Xki.(INCi,Pli.,P2i. Pki. Pni,>Zi)
where
ZA represents other variables of interest.
An interesting result is that total market demand (Xki ) is a
function of only the average import price for this commodity group and the
average price of substitutes and not of the individual product prices
(Pkij)-
Total market demand is then separated by exporting region in the
second level of the two stage budgeting process to obtain the "product
demand" equations. In this case, consumers minimize the cost of
purchasing Xki (total market demand for commodity k in region i). That
is, consumers minimize total expenditure (INCi) subject to the following
constraint:
(4.5) Xki = /k (Xkil,Xki2 Xkim)

80
to obtain the specific product demands. The function /k is assumed to be
a linear and homogenous function of the product demands xkij to ensure that
Pki is independent of Xki Pki is only a function of Pkil, Pki2> > Pkim-
The assumption that the quantity index functions are linear and
homogeneous is the second restriction (the first being the assumption of
independence) that has been placed on U. The product demands generated
under these assumptions are functions of the total market demand level
(Xki ) and the individual product prices (PkiJ) and is given by
(4.6) Xkij Xkij(Xki ,Pkil,Pki2 Plcira)
This relationship clearly states that the allocation of imports
among regions of origin depends on total market demand and relative prices
of the products in the market.
The total market demand equation for the world's fresh orange trade
model in any particular region is defined following the theoretical
framework developed above. It is possible to write the market demand
equation for fresh oranges as independent of other goods consumed in the
same region. The model will be dealing with only one good (fresh oranges)
hence the subscript "k" is no longer necessary. Let X^ represent fresh
oranges exported from region j to region i.
The market demand equation for fresh oranges is expected to be a
function of the average market price, income, population, and the price of
substitute products. The average market price should be obtained by
taking into consideration the local product price and the price of imports
including any tariff or preferential treatment. In the case of substitute
products, it has been necessary to define what is really a substitute for
fresh oranges. Several alternatives were considered, including an

81
aggregated commodity representing all other fruits, an aggregated
commodity representing all other goods, and an aggregated commodity
representing bananas and apples. The latter alternative was selected
based on the characteristics of consumption for fresh oranges which makes
bananas and apples better substitutes than the other aggregated goods.
Nelson and Robinson (1978) reported that Matthews, Womack, and Huang
(1974) used bananas and apples as substitute products for fresh oranges
with encouraging results, even though they were in some cases less
significant than the ones obtained using other alternatives. In a recent
paper, Lee et al. (1990) also used bananas as a substitute product for
U.S. citrus in Japan. However, in the latter study pineapples instead of
apples were used as the second substitute product.
The general form of the market demand equation for fresh oranges is
the following:1
(4.7) Xi. f(Pi-,INC,POP,PRS)
where
f represents some functional relationship between XA and the
variables on the right hand side,
Xt is the total market demand for fresh oranges in region i,
P is the real average market price of fresh oranges in region i,
INCi is the real income level in region i,
POPi is the population level in region i,
PRSi is the real average market price for the aggregated commodity
based on bananas and apples or other measure of substitutes in
region i.
Single letter notation represents endogenous variables while three letters depict exogenous
variables. The sign associated with each variable represents the hypothesized behavioral
relationship between the exogenous variables and the dependent variable.

82
The second level of two stage budgeting is to allocate total market
demand by supplying region. It requires the definition of a "product
demand" equation which represents the demand in region i for fresh oranges
coming from region j. The product demand functions consider Armington's
demand theory of products differentiated by place of origin. Prices for
products in commodity markets other than fresh oranges have no effect
except through the size of the markets. The function ik in equation 4.5
is assumed to be a linear and homogenous quantity index function of the
product demands Xkij to ensure that Pki is independent of Xki The average
market price Pki is a function of Pkil, Pki2, > pkim- Equation 4.6 defined
the product demands to be a function of the market size and all product
prices. Since Pki is a function of all product prices product demands and
market shares can be reduced to depend on relative prices and total market
demand or market size. The relative price for each product demand is
given by the ratio of the product price to the average fresh orange price
in the market. The product demand functions are
(4.8) XtJ = h' (Pi'jPi+.,X^/")
where
h' represents some functional relationship among variables,
XAj is the demand in region i for fresh oranges coming from region
j .
Py is the price in region i for fresh oranges coming from region
j
The actual relationship is given by
(4.9) X4j = MP^/P^X//-)
where
h represents some functional relationship among variables.

83
Given equations 4.8 and 4.9, it is possible to define each region's
market share equation as follows:
(4.10) S -Xij/Xi. = zCPy/P^Xi'-)
where
z represents some functional relationship among variables,
Sjj is the market share of fresh oranges for region j in region i.
Supply Side
Total orange production (PRD j) is defined to include oranges
supplied to the fresh market (PRD^) and oranges utilized for processing
into orange juices (PRDZj). The orange industry requires several years to
introduce new trees and new supplies in the market and high levels of
investment to built a new processing plant. It is reasonable to assume
that orange production and its utilization levels do not adjust so fast as
to be considered part of a simultaneous demand side decision model of
international trade. Therefore, total production and, in particular,
fresh orange utilization (PRD^) is considered exogenous in this model.
The general equations representing this condition are the following:
(4.11) PRDu = PRD j PRDzj
and
(4.12) PRD2j = Aj*PRD j
PRD j is total orange production in region j,
PRDXj is total fresh orange utilization in region j,
PRDZj is total processed orange utilization in region j,
where

84
Aj is the percentage of total orange production utilized in the
processed industry in region j and is assumed to be exogenous.
Export Supply Equations
Exporters will respond to export prices by adjusting their level of
exports accordingly. Given changes in total production and fresh
utilization, exports will also tend to adjust accordingly.
Export supply equations are consequently assumed to be a function of
the average export price from region j (average Free On Board price = F j)
and total fresh orange utilization (PRDXj) in the region of origin. The
export supply equation for fresh oranges is the following:
(4.13) X.j = Si Xij = v(F+j,PRDij) for i*j
where
The summations represent total exports of fresh oranges from region
j to all other regions,
v represents some functional relationships between variables,
F j represents the average export price of fresh oranges from
region j to all other regions.
The demand equations for local product will follow from the
difference between total fresh utilization (PRDij) plus the change in
inventories (when applicable) and the export supply from region j Demand
for domestically produced product is:
(4.14) Xjj = PRDu + A INVj X.j
Given that fresh oranges can be stored only for short periods of
time, it is assumed that inventory levels are zero. Accordingly, the
change in inventories will be zero and equation 4.15 will be given by
(4.15) Xjj = PRDj X.j

85
Again, Xjj is the amount of product produced domestically and remains
in the same region. Total market demand is Xj where Xjj is a subset of
XJ.-
Equilibrium Conditions
The equilibrium conditions required to have a closed model for the
fresh orange industry include three basic identities. The total market
demand in region i must equal the supply of products to that region
(4.16) Xi. = Xj Xu
Total market demand in region j (Xj ) must equal fresh utilization
(PRDjj) plus imports (XA iflj X^) minus exports (Xj ij4J Xi) as follows:
(4.17) Xj PRDij + Xt ifij Xji Xj Xld
Finally, total production of oranges must equal the total production
used fresh plus the total production used processed
(4.18) PRD j = PRDu + PRD2j
Price Linkage Equations
Total market and product demand as well as export supply are
functions of different but closely related set of prices. Total market
demand is a function of the average market price for a particular
commodity. This price is associated with the local price and the
individual product prices (Py) Each product demand is a function of its
own product price and indirectly a function of the individual product
prices (Pij) through the average market price (Pi.). Export supply is

86
associated to a similar set of prices through the FOB (Free On Board)
export price.
In this section, the price linkage equations among regions are
presented. In each region there is an export price that corresponds to
each region. This price is the FOB export price and will be denoted F^.
Accordingly, the average export price for fresh oranges (F j) from region
j to all regions is defined as follows:
(4.19) F j = tSt ^(Fy XtJ)] / i,j Xij]
The numerator in equations (4.19) represents the total export value
from region j to all other regions and the denominator represents the
total quantity exported. The use of i*j is because no data are available
for within-region export price (Fjj) and the equations represent the
weighted average export price, which should not include the local price.
Since Fjj is not available for all regions and will be used in the
following calculations, it will be assumed to equal F j.
The CIF price is the price of a product in the port of final
destination. The represent the CIF price of fresh oranges coming from
region j to region i. These prices do not include any trade barriers and
are a function of the Fi price. Changes in the FOB export prices (F^)
are not expected to have the same impact on the CIF import prices (C^)
across regions. This follows from the assumption that certain market
structures could exist and prevent perfect transmission of prices from the
regions of origin to the regions of destination. Spoilage or product
deterioration during the transportation process from region j to region i
could be different than from region r to region i. Therefore, the general
relationship between CIF (C^) and FOB (F^) prices might not be linear.

87
In addition, Cjj is assumed to be a function of a trend variable to capture
technological changes over the years and an energy price index which take
into consideration the price of energy over time. The equation for the Cjj
price is
(4.20) Ci;i
= q(F1+j,TRD+/',PEN+) for i*j
where
q
represents some functional relationships among variables.
cij
is the price for fresh oranges in the destination port (region
i) coming from region j for i^j,
TRD
is a trend variable to capture technical improvement (-) or
decay (+) over time,
PEN
is the energy price index.
The market price of fresh oranges coming from region j to a
destination region i is not given by the Cjj price directly. CIF prices
should be increased or decreased by the effect of trade barriers or
preferential treatments in the final market i. These barriers could be
found to be represented by percentages of the import price or absolute
value tariffs which have to be added (barrier) or subtracted (preference)
to obtain the real final market price. Therefore, the final market price
for fresh oranges from region j to region i is given by
(4.21) Pij
- Cjj (1 + TABij) + TAXjj
where
TABjj is a percentage that represents the tariffs (positive) or
preferential treatment (negative) effects on the price for
fresh oranges coming from region j to region i.
TAXij is an absolute value term that represents a tariff per unit of
product (positive) or a direct preferential treatment
(negative) that affects the final price in market i of fresh
oranges coming from region j .
Given that TABjj and TAXjj are zero, Pjj will be equal to Cjj and thus
equal to F j.

88
The present model will not consider other type of trade barriers.
It is not clear from the data whether quotas have been limiting; however,
they could have been in some periods within certain years, especially in
the case of Japan. Quotas are imposed only in a few countries of the
world. In most cases, they are in place just for some months of the year.
The fresh orange trade model utilized annual data, so it will not capture
seasonal barriers. If quotas are included in the model, inequality
restrictions must be required. The empirical estimation of the
econometric simultaneous system will be unnecessarily complicated given
the presence of the inequality restrictions.
Given the assumptions and conditions set forth above, the average
market prices for fresh oranges in region i will be given by
(4.22) Pi. = [ZjCPij Xij)] / Xi.
The world fresh orange trade model presented above determines the
equilibrium prices, trade flows, fresh utilization, and total demand and
export supply for fresh oranges for all regions simultaneously.
CRES Model Restrictions
In the present study, 11 regions representing the world's countries
have been specified. Product demand equation (4.9) will be too
complicated to be of practical use, given the number of parameters to be
estimated. As noted earlier, Armington (1969a) made two important
assumptions regarding the substitutability between different products in
order to simplify and make the model applicable for empirical analysis.
The elasticities of substitution in each market were constant, and the

89
elasticity of substitution between any two products competing in a market
is the same as any other pair of products competing in the same market.
Following his assumptions, Armington used a constant elasticity of
substitution (CES) index of the quantities imported from the regions of
origin.
While Armington argues that product coming from different (national)
sellers could be differentiated by place of origin, the adoption of the
CES specification implies that the elasticities of substitution are
constant and equal between all pairs of exporting regions in each market.
The cross-price elasticities between all pairs of regions need not be
estimated, since they can be obtained from the price elasticities and the
elasticities of substitution (Learner and Stern, 1970). The approach
followed by Artus and Rhomberg (1973) and later by Sparks (1987) is
applicable in the world fresh orange trade model developed in this study.
Artus and Rhomberg used the constant ratio of elasticity of substitution
and homothetic (CRESH) index and Sparks used the constant ratio of
elasticity of substitution (CRES) index which makes the model somewhat
less restrictive. The CRES assumption implies that, even though
elasticities of substitution will vary proportionally to maintain the
ratios fixed, they are allowed to be different between any two pairs of
products competing in the same market.
Given the general form of the market demand equations for fresh
oranges:
(4.23) Xi. = /*(Xu,Xi2 Xim)
and assuming that n is a CRES index function, the market demand equation
has the following form:

90
(4.24) Xt. [EJ(bij*XijQij)](1/Qi )
It can be shown (see Appendix B) that the product demand equation
derived from this total market demand is the following:
(4.25) X Uai./(u*b11))<1/<1J'1)>l*[ .pc,
The market share equation is
(4.26) Sij = Xij/Xi.
[ (ai./(Qij*bij)) ]*[ (pij/pi.) ]
*[xi.(0l-"lj)/(aij'1)]
Based on Hanoch (1971), the following terms are defined to obtain a
simple relationship that includes the Allen-Uzawa (Allen, 1938; Uzawa,
1962) partial elasticity of substitution:
(4.27) WlJ = 1/(1- aqj) as an identity,
(4.28) ViJ [Pij Xj] / [SjPj Xy)] as the value share of fresh
oranges coming from region j to region i,
(4.29) Oy [wit ¡j] / [EjCVy Wjj)] as the partial elasticity of
substitution for fresh oranges.
As expected, this elasticity varies only by a constant ratio which
is l/ISjiV^ )].
The remaining equations in the model are not affected by the
assumption of the CRES index. Hence, the complete system to be estimated
is the following:
[Market Demand]
(4.30) Xt. [/30i 1 [pi.(/3li)] [INCi^ai)] [POP^3^] [PRs/^]

91
(4.31) Xxj [0iJ] [(Pi/Pi.)(lij)] [(Xi.)(52ij)
where
6 SUj = l/(airl) = Uij
2ij (i.-D/Cay-l)
(4.32) Xjj = PRDXj X j
and
[Product Demand]
[Demand for Domestic Product]
- PRD.J PRD2j
[Fresh Utilization]
Aj*PRD j
[Processed Utilization]
17oj] t(F.j)(7lj)] [(PRDXj)(72j)
] [Export Supply]
XU
[Equilibrium Condition]
PRDij + i5tj XJt Sj XAj
[Equilibrium Condition]
PRDij + PR^2j
[Equilibrium Condition]
[Si ijtj(Fij Xij)] / [X, m XtJ]
[FOB Average Export Price]
(4.40) Cy = [jr0iJ] [ (FJ)(?riij) ] [ (TRD)
(4.41) Pi;j Cu (1 + TABi;i) + TAXij
(4.42) Pi. = [SjiPij Xi;j)] / Xi.
[CIF import price]
] [(PEN)(,r3ij)]
[Market Price]
[Average Market Price]
The model is clearly formed by several identities and behavioral
equations. The identities need not be estimated. To estimate the rest of
the model, equations (4.30), (4.31), (4.35), and (4.40) can be transformed
to linear equations by applying logs to both sides of the equations. The
parameters which represent the partial elasticities can be read directly
from the estimation with the exception of the intercept.

92
Ordinary least squares estimated parameters are biased for this
simultaneous equation model. Another estimation problem in the model is
that, even though it is linear in the parameters, it is intrinsically
nonlinear in the variables, given that after transforming equation (4.30)
total market demand will be in the log linear form while it appears
without the log linear form in equation (4.36). As a consequence,
nonlinear two stage least square procedure was used. The specific
estimation steps used and results will be given in the next chapter.
Model Implications
Modeling the changes of world trade flows of the orange industry by
identifying international trade linkages among the major trading regions
and recognizing current and emerging problems in the industry is the major
objective of the present study. Estimation of the world trade model
described above will generate consistent estimates of the parameters of
the market demands, product demands, export supply, and CIF import price
equations for the major trading regions in the industry. Analysis of the
estimated parameters will provide information to help understand the
reasons for changes in market shares and facilitate longer term forecasts
and policy analyses.
The parameters of the market demands measure the strength of the
influence of the average price of fresh oranges in a given region, as well
as the intensity of income and population levels, and substitute commodity
prices. Using price elasticities, it is possible to predict responses in
the different markets to changes in supply prices. Income and population

93
elasticities give an idea of possible changes in consumption and trade
patterns and substitute product price elasticities give information about
the strength of substitution with respect to other commodities. The
estimated parameters also yield a measure of the substitutability among
products of the sam kind coming from different regions in a given region.
On the supply side, the parameters measure the strength of the
relationship between export supply, the average export price and the total
fresh utilization for a particular region. The relationship among the
import price (CIF), export price (FOB), the trend, and the energy price
index (PEN) between regions was measured.
The system will be used to perform sensitivity analyses over several
scenarios. External shocks to the different exogenous variables are used
to illustrate the impact on fresh orange trading levels and patterns
across regions. Each scenario is described in detail at the appropriate
point in Chapter 6.
Trade Data Base
To quantify the fresh orange trade model, considerable international
trade information is required, including interregional trade flows and
values. Trade data on fresh oranges for all countries in the defined
regions was needed. Data were used on trade flows from every country to
all the other countries reported in quantities (metric tons) and monetary
value (U.S. dollars). With these data, it will be possible to obtain by
aggregation total export and import quantities and unit prices for each
defined region. The unit export price (FOB) from each region was obtained

94
by using the total amount of dollars and the total quantities shipped from
the region. The CIF prices use the data for the port of destination.
The period of study includes annual data from 1966 to 1986.
The data mentioned above were obtained from the United Nations
Commodity Trade Statistic Tapes (1987). These data have several problems,
ranging from reporting errors to different reporting systems and coding
mistakes. Some of the problems found were mixed classifications, missing
data, CIF import prices less than their associated FOB export prices, and
total reported exports different from total imports in the same year. It
took at least four full months and the use of several SAS (1982) data
management procedures to get the data into a useable form. When possible,
data were validated against other sources such Food and Agriculture
Organization (FAO) and International Monetary Fund (IMF) Trade Statistics.
However, the UN trade data are the only source available containing the
information regarding trade flows among the countries included in the
analysis. The Standard International Trade Classification code
corresponding to fresh oranges (SITC = 05711) was used.
Using the annual UN trade data has two important drawbacks that
should be noted. Intrayear seasonality was not captured, since data are
given in annual observations. Quality and varietal differences are not
captured, since data are collected for aggregated fresh oranges in each
country.
UN trade data quantities are given in metric tons and monetary
values in thousands of U.S. dollars. Since unit price information in real
terms were needed, regional CPI (Consumer Price Index) were used to
deflate the dollar values in each region. The U.S. CPI is an alternative

95
variable which could have been used as in most international trade models
(for example see Sparks, 1987). In those cases, the exchange rates are
not explicitly included in the model. They are only implicitly included
since all value units are expressed in U.S. dollars. That use implies the
assumption of purchasing power parity in all regions. It assumes that the
exchange rates in each country will perfectly reflect the differences in
inflation rates relative to the United States. Given that in practice
that is not true (Dornbusch, 1988; Lessard, 1985), the CPI's were
estimated for each region using a procedure suggested by Edwards and Ng
(1985) that relates exchange-rate indices with inflation rates for each
country. The results for each country were aggregated into the regions
using a weighted average based on trade levels. The details of the
procedure utilized is included in Appendix C. The raw data required to
develop this calculations are the exchange and inflation rates per country
obtained from the country section of the IMF International Financial
Statistics Supplements (various issues).
Income and population levels by country were needed. The Gross
Domestic Product (GDP) in current market prices for each country was used
as a proxy for income. These data were obtained from the IMF
International Financial Statistics Supplements (various issues) and are
expressed in billions of U.S. dollars. Population data by country were
obtained from Food and Agriculture Organization (FAO) Production Yearbook
(various issues). Income levels and population were aggregated according
to the regions defined. The regional CPI's were used to deflate the
GDP's.

96
The total production levels by country were obtained in the FAO
Production Yearbook (various issues). Given data limitations, the
information used in the first five years of the analysis included
tangerine production. However, the percentage of tangerines in total
production was not important. The allocation to the fresh and processed
markets for all regions was not available in published documents. To
obtain the information, various documents were used including United
States Department of Agriculture-Foreign Agricultural Service (USDA-FAS)
Attache Citrus Annual Reports and Supplements (various issues), the
Horticultural Products Review (various issues), and the Citrus Reference
Book (1988, 1990) by the Florida Department of Citrus. Also direct
consultation with several governmental offices in Washington was used. It
is important to note that the final data developed for orange utilization
for this research are not available in any other source in the detail
collected. Appendix D shows the final orange utilization data used for
estimation. Information regarding inventory levels was not needed, given
that orange utilization is treated as exogenous and the study is limited
to perishable fresh oranges.
Energy prices (PEN) were obtained from the Commodity Prices' section
of the IMF International Financial Statistics Supplements (various
issues). Fuel prices used corresponded to crude prices from Saudi Arabia
and are expressed in U.S. dollars per barrel.
Since local market prices were not available for all countries
substitute product prices for fresh oranges were defined using average
unit prices for bananas and apples for each country. Weighted average
unit prices were obtained by dividing total import values over total

97
import quantities. The data were obtained from the FAO Trade Yearbook
(various issues) and aggregated for regions.
Trade barriers and preferential treatment data were collected from
several sources. USDA-FAS Attache Citrus Annual Reports and Supplements
(various issues) often provide a reasonable source for identifying trade
barriers including tariffs. Other documents consulted included the USDA-
FAS publication on U.S. Import Duties (1973), The Florida Citrus Mutual
Report (August 1977) Citrus in Japan (1978), the Bulletin International
des Douanes (various issues), U.S. Exports: Harmonized Schedule B.
Commodity by Country (various issues), Customs Tariff Schedules of Japan
(1980), Sarris (1984), Baker and Mori (1985), and the Tariff Schedules of
the United States-Annotated (1983, 1984, 1985). Tariff schedules obtained
by country were weighted by countries' volume of trade to determine the
tariff schedules for each region. Appendix E shows the final tariff data
used for estimation.

CHAPTER 5
ECONOMETRIC PROCEDURE AND EMPIRICAL RESULTS
Introduction
This chapter will be divided into two main sections. The first
covers the estimation procedure and the associated econometric issues.
The second discusses the empirical results and its major implications in
terms of the fresh orange trade model developed. A general conclusion is
given at the end of the chapter.
Econometric and Estimation Procedure
The fresh orange trade model under study is based on 13
relationships for each one of the 11 regions considered. Nine of those
relationships are identities and therefore do not need to be estimated.
The rest are behavioral relationships that must be estimated. The
equations to be estimated are total market demands, export supplies,
product demands, and CIF price linkage equations. Each region has one
total market demand equation, one export supply equation, and ten product
demand and CIF price linkage equations, one for each partner region. The
total number of equations in the model including identities added to 440
and the number of equations to estimate totals 242.
98

99
Since some of the endogenous variables appear both in natural and in
the logarithmic form in the different equations the system is nonlinear.
It is simultaneous because the endogeneous variables are jointly
dependent.
A basic assumption of the Ordinary Least Square (OLS) model is that
the right-hand side variables are independent of the error term. This
implies that the expected value of X (exogenous variable) and n (error
term) is zero. In the case where an endogenous variable appears on the
right-hand side, this assumption is no longer valid. If a simultaneous
system is estimated using OLS, then the parameters obtained will be
inconsistent. Therefore, the model has to be estimated using a
simultaneous system estimation technique.
The specific method of estimation partially depends on the
identification problem. This means whether numerical estimates of the
parameters of the structural equations can be obtained from the estimated
reduced form coefficients. The reduced form of a model is obtained when
the endogenous variables are expressed as functions of all exogenous
variables. If this can be done, the particular equation is identified.
If not, the equation under consideration is not identified or is
underidentified. An exactly identified equation implies that unique
numerical values of the structural parameters can be obtained. An
overidentified equation implies that more than one numerical value can be
obtained for some of the parameters. Fisher (1976) and Brown (1983) give
a complete description and interpretation of the alternative approaches to
the identification problem for linear and nonlinear system models in
econometrics.

100
The rules for identification are the so-called order and rank
conditions. Let's say that M is the number of endogenous variables and K
the number of exogenous variables in a given model. Additionally, m is
the number of endogenous variables and k the number of exogenous variables
in a given equation of the same model. The order condition says that, in
a model of M simultaneous equations, an equation is identified if it
excludes at least M 1 variables (both endogenous and exogenous)
appearing in the model. If it excludes exactly M 1 variables, the
equation is just identified. If it excludes more than M 1 variables, it
is overidentified. The order condition is a necessary but not sufficient
condition for identification.
The rank condition is both a necessary and sufficient condition for
identification. It says that, given a model containing M equations in M
endogenous variables, an equation is identified if and only if at least
one nonzero determinant of order (M 1)*(M 1) can be constructed. The
determinant has to be created from the coefficients of the variables (both
endogenous and exogenous) excluded from the equation but included in the
other equations of the model.
In the case of the fresh orange industry, the trade model is
overidentified. The total number of variables in the model is 871, 310
exogenous variables (K) and 561 endogenous variables (M). There is only
one way for the estimated equations to be underidentified given the order
condition. This will be the case when an equation includes more than 311
variables in the right-hand side, which is clearly not the case. The rank
condition is also satisfied, but details will not be presented here.

101
If a model is just identified, Indirect Least Squares (ILS) could be
used to obtained the structural coefficients from the OLS estimates of the
reduced form coefficients. Given that the equations are overidentified,
the use of ILS will provide multiple estimates for the parameters in each
case. Therefore, it will be necessary to use a system estimation
technique that provides only one estimate per parameter.
The method used for estimation was nonlinear two stage least squares
(NL2SLS), which is a simultaneous limited information method. The first
stage of this technique is to determine a set of instruments to be used
instead of the endogenous variables that appear in the right-hand side of
the original equations. These instruments should not be correlated with
the error term but should be highly correlated with the endogenous
variable to be substituted. The use of these instruments will assure that
the parameters obtained are consistent. The traditional method to obtain
these instruments is by regressing the endogenous variables on all the
exogenous variables included in the system using OLS. The instruments
obtained will be the predicted values or estimated mean values of the
original endogenous variables conditional upon the fixed exogenous
variables. The second stage consists in using the instruments obtained to
substitute the endogenous variables appearing in the right-hand side of
the original equations. Once the endogenous variables have been
substituted, the model can be estimated using OLS.
The NL2SLS method does not take into consideration the correlation
among the errors across the equations in the model as full information
methods do. Accordingly, the use of NL2SLS is based on the assumption
that there is no evidence of the existence of an external factor that

102
could affect all the equations in the model. Maddala (1971) supports this
assumption for large econometric models based on two essential points
where conventional methods pose problems. One is when the unrestricted
reduced form is not estimable because the number of predetermined
variables in the system is larger than the number of observations. The
second is when one uses system methods where the covariance matrix of the
residuals can not be computed again because of too few degrees of freedom.
The fresh orange trade model developed here fits partially Maddala's
classification, given that the number of equations to estimate is 242 and
the number of observations available is 21.
There are several additional benefits in using NL2SLS over the full
information methods in this particular trade model. Specification errors
are common in large econometric models, and data problems are also
expected, given that the model deals with trade data. If there is any
specification error in one of the equations of the model, the use of
NL2SLS will prevent the error from affecting the rest of the estimated
results. On the other hand, full information methods are sensitive to
small changes in specification and/or data (Goldstein and Khan, 1976).
Using NL2SLS clearly simplifies the estimation procedure, given that it
can be applied to an individual equation without directly taking into
consideration any other equation(s) in the system. In addition, full
information methods require for practical implementation sharpness of
identification of the whole model, otherwise it will interfere with the
estimation (Klein, 1969) Finally, research has been inconclusive about
the performance of the full information methods when compared to the

103
limited information methods such as the NL2SLS procedure (Goldfeld and
Quandt, 1968).
The statistical justification of the NL2SLS is of the large-sample
type, which implies that the estimated standard errors in the second-stage
regressions are not completely reliable. Therefore, a rule of thumb is
that a parameter with a "t" statistic greater than one can be considered
significant (Gujarati, 1988).
The procedure to obtain the instruments consists in regressing the
right-hand side endogenous variables on all the exogenous variables in the
model. Given that in the trade model presented here the number of
exogenous variables is large and the period of study includes only 21
years, there will not be enough degrees of freedom to perform the
estimation needed to obtained the instruments.
One way to solve this problem is the use of the principal components
approach developed by Kloek and Mennes (1960) and strongly supported later
by Amemiya (1966). This procedure is capable of reducing the information
to a subset represented by a subspace of the K-dimensional exogenous
variable space. Six or less principal components usually contained most
of the variation of the exogenous variables included in the model and
therefore provided the necessary information to perform the estimation.
The subset of principal components is used instead of the exogenous
variables to obtain the predicted values of the right-hand side endogenous
variables or instruments. This procedure has been widely used with good
results. A few examples are Jones and Ward (1989), Klein (1969), and
Fisher (1965).

104
In certain cases, it is not necessary and is even better not to use
all the exogenous variables of the model to obtain the principal
components (Amemiya, 1966). The researcher may wish to remove from the
set of exogenous variables those that are not clearly exogenous or those
which contribute little to the explanation of the endogenous variables.
For the fresh orange trade model, a subset represented by the exogenous
variables that clearly contribute to explain the variations of the
endogenous variables was included in the first stage. The principal
component procedure and final estimation as well as the list of variables
included are shown in Appendix F.
The specific relationship between the endogenous variables and the
principal components selected for the first stage of the NL2SLS procedure
depends on the problem on hand. Previous research has shown that in a
model in which nonlinearities in the variables appear, the specification
should be also nonlinear (Goldfeld and Quandt, 1972).
In the first stage of the NL2SLS procedure, it is possible to use
different functional forms for each relationship among endogenous
variables and selected principal components (Johnston, 1984; Goldfeld and
Quandt, 1968; Kloek and Mennes, 1960). The use of different
specifications gives the model the necessary flexibility to obtain the
best results from the first and second stages. In other words, any
alternative nonlinear specification use will provide consistent estimates.
Two nonlinear specifications were used. The first one follows a
second-degree polynomial which is considered a good approximation
following Goldfeld and Quandt's (1972) findings. The second specification
follows the Gompertz equation, which is another nonlinear function that

105
produces an S curve that starts from a lower asymptote and rises to a
higher one. Kotler (1971) reports a detailed theoretical discussion on
the Gompertz function. Ward and Forker (1990) recently used the Gompertz
function in a similar application with good results in the beef industry.
The polynomial and Gompertz specifications were used to obtained the first
stage of the different equations. The final decision about the type of
functional form use was based on the performance of the different
equations in the second stage.
To make the method work, it will be necessary to make another
important decision. That is to determine the functional form of the
endogenous variables to be used as dependent variables in the first stage
of the NL2SLS procedure. For the linear case the problem does not arise
since variables are the same in every section of the model. When dealing
with a model that includes nonlinearities in the variables, at least two
different alternatives should be considered.
Suppose that y is an endogenous variable that appears as "y" and as
the "logarithm of y" in two different equations of the model. If the
"logarithm of y" appears in the right-hand side of one of the equations of
the model, two alternatives are possible. First, use the nonlinear form
as a dependent variable in the first stage, i.e., log y. This implies one
obtains the predicted value of the "logarithm of y" as an instrument for
the second stage. Second, use the linear form of y, i.e., y. This
implies one obtains the predicted value of y and uses as an instrument the
logarithm of the predicted value of y. The second alternative does not
follow the rationale of NL2SLS, given that what is needed is the predicted
value of the actual variable appearing in the right-hand side of the

106
equation. It is also known that the expectation of a function is
generally unequal to the function of the expectation. Therefore, the
second alternative is inappropriate (Goldfeld and Quandt, 1968; Goldfeld
and Quandt, 1972). The first method was used to obtain the necessary
instruments for the second stage.
The methodology utilized to estimate the model closely follows the
nonlinear two stage least squares method proposed by Kelejian (1971) and
supported later by Goldfeld and Quandt (1972) and Amemiya (1974).
However, the final procedure used introduced different specifications for
the equations in the first stage.
Time Series Processor (TSP) International PC Version (TSP User's
Guide and Reference Manual, 1983) procedures were used to estimate the
model. The final program used for the estimation of the model is included
in Appendix G.
The data used cover from 1966 to 1986 and have been recorded from
many sources as described in Chapter 4. Trade data were insufficient in
terms of degrees of freedom to perform an adequate estimation for some
equations. In those cases, a special TSP procedure was utilized to select
and estimate only those equations for which trade took place. Equations
with less than six trade observations did not provide enough information
for a reliable estimation. In that event, the equation was not estimated
and its parameters were set to zero. Data were read in from Lotus files.

107
Empirical Results and Implications
This section of the chapter will discuss the empirical results
obtained from the estimation of the fresh orange trade model. The section
will be divided into four parts. The first part will show the empirical
results for each estimated equation. The second part will present figures
showing the actual and fitted values of some of the model's most important
equations. The third part will include statistics to evaluate the model's
performance in terms of fit, specification, and simulation ability. The
fourth part will present a discussion about the parameters obtained.
Theoretical economic expectations about signs and magnitudes as well as
the implications of the associated "t" statistics will be addressed. The
connection of the different findings regarding trade patterns described in
Chapter 2 will be addressed throughout the discussion.
Empirical Results
Tables 5.1 to 5.24 show the empirical results of the estimation of
the four behavioral equations -- total market demands, export supplies,
product demands, and CIF price linkage equations -- for the 11 regions.
Tables 5.1 and 5.2 present the results for total market demand and export
supply equations. The regions are United States (US), Canada (CAN), Latin
America (LA), Mediterranean-EC (MED-EC), EC, rest of Western Europe (RWE),
Middle East/North Africa (ME/NA), rest of Africa (RAF), Far East (FE) ,
Oceania (Oceania), and the Communist Bloc (COMMB).

Table 5.1 Total Market Demand Equations
Region
Intercept
(+/-)
Market
Price
<-)
Real
GDP
( + )
Subs
Population
( + )
. Product
Price
(+)
eoBs
@RSQ
@DW
@FST
UTHEIL
21 Years Average
Z of Total World
Demand Imports
us
PARAM. VALUE
STD.ERROR
t STATISTIC
-14.555
7.330
-1.986
-2.182
0.310
-7.030
-0.549
0.406
-1.352
2.119
0.771
2.749
0.640
0.186
3.436
21
0.77
2.52
13.16
0.028458
7.117
1.200
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
25.927
10.893
2.380
-0.569
0.453
-1.256
0.693
0.386
1.798
-1.792
1.237
-1.449
0.038
0.325
0.117
21
0.23
1.94
1.22
0.040324
0.747
4.806
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
1.307
7.079
0.185
-0.057
0.379
-0.151
0.091
0.224
0.407
1.044
0.493
2.116
0.134
0.248
0.539
21
0.78
1.83
14.22
0.042904
28.590
0.064
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-66.000
54.799
-1.204
-0.781
0.752
-1.039
-0.641
0.803
-0.798
7.179
4.946
1.452
-0.200
0.474
-0.421
21
0.37
3.27
2.38
0.085052
10.591
0.069
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
118.270
60.693
1.949
-1.182
0.363
-3.255
0.674
0.544
1.239
-9.114
5.279
-1.726
0.259
0.303
0.856
21
0.49
2.52
3.90
0.031024
9.915
63.427
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
17.852
21.072
0.847
-1.028
0.560
-1.833
-0.010
0.282
-0.036
-0.717
2.053
-0.349
0.124
0.286
0.432
21
0.40
1.73
2.67
0.033662
1.637
10.572
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
19.492
8.751
2.228
-1.138
0.767
-1.483
0.411
0.157
2.626
-0.554
0.800
-0.692
-0.476
0.290
-1.643
21
0.95
1.98
80.91
0.031827
11.507
1.654
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
5.648
5.390
1.048
-0.250
0.904
-0.276
-0.092
0.288
-0.318
0.563
0.467
1.205
0.097
0.286
0.338
21
0.27
1.21
1.49
0.061481
2.456
0.164
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
-5.327
6.888
-0.773
0.181
0.086
2.099
0.151
0.201
0.749
1.446
0.567
2.550
-0.165
0.073
-2.252
21
0.95
2.40
79.89
0.023174
23.015
4.290
OCE
PARAM.VALUE
STD.ERROR
t STATISTIC
13.838
7.696
1.798
-1.655
0.591
-2.800
-0.335
0.405
-0.827
-0.427
0.955
-0.447
0.182
0.172
1.061
21
0.35
1.49
2.20
0.042669
0.808
0.202
COTOffl
PARAM.VALUE
STD.ERROR
t STATISTIC
-0.126
44.299
-0.003
-1.216
0.553
-2.197
1.176
0.553
2.127
0.505
3.681
0.137
-0.523
0.500
-1.046
21
0.86
1.42
23.60
0.054849
3.616
13.552
Total
100.000
100.000

109
Table 5.2 Export Supply Equations
21 Years
FOB Fresh Average X of
Region
Intercept
(+/-)
Price
(+)
Production
(+)
@OBS
@RSQ
@DW @FST
UTHEIL
Total World
Exports
US
PARAM.VALUE
STD.ERROR
t STATISTIC
-3.549
7.807
-0.455
-1.679
1.662
-1.010
0.914
0.491
1.862
21
0.20
1.71
2.19
0.106359
8.895
CAN
PARAM.VALUE
STD.ERROR
t STATISTIC
8.834
1.754
5.038
2.245
0.849
2.643
1.000
0.000
0.000
21
0.28
2.68
3.49
0.636455
0.004
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
1.416
5.935
0.239
-0.155
0.425
-0.364
0.639
0.340
1.879
21
0.17
1.45
1.89
0.13531
3.919
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
4.459
3.808
1.171
-0.009
0.303
-0.031
0.646
0.234
2.755
21
0.31
2.49
4.02
0.079362
43.996
EC
PARAM.VALUE
STD. ERROR
t STATISTIC
1.825
0.847
2.155
-0.275
1.470
-0.187
0.692
0.290
2.389
21
0.82
1.41
41.36
0.172107
0.248
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
1.693
1.320
1.283
-2.650
0.634
-4.182
1.000
0.000
0.000
21
0.49
1.24
8.75
0.355627
0.052
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
-0.663
5.754
-0.115
1.424
0.429
3.317
1.187
0.439
2.704
21
0.43
1.11
6.72
0.050222
34.953
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
7.157
5.097
1.404
-0.191
0.711
-0.269
0.357
0.356
1.002
21
0.06
0.55
0.53
0.077483
6.242
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
5.870
8.745
0.671
-1.244
0.345
-3.602
0.154
0.593
0.259
21
0.67
1.30
17.98
0.196982
1.089
OCE
PARAM.VALUE
STD.ERROR
t STATISTIC
1.823
10.532
0.173
-2.261
0.941
-2.402
0.303
0.921
0.329
21
0.33
0.78
4.51
0.165562
0.479
COMMB
PARAM.VALUE
STD.ERROR
t STATISTIC
-23.264
3.764
-6.181
-1.792
1.424
-1.258
2.144
0.220
9.738
21
0.85
1.52
51.55
0.165291
0.123
Total
100.000

Ill
Table 5.4 Canada Product Demands
Partner
Region
Intercept
( + \-)
Relative
Price
(-)
Total
Market
Demand
H+>
@OBS
@RSQ
@DW
@FST
UTHEIL
21 Year Average
Z Z Total
Total Market
Imports Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
-A.777
13.596
-0.351
-0.567
1.195
-0.474
1.372
1.122
1.223
21
0.30
2.06
3.80
0.057894
79.674
79.674
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
232.670
90.774
2.563
-0.470
1.635
-0.288
-18.571
7.440
-2.496
21
0.33
1.15
4.38
0.383285
1.230
1.230
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
293.400
127.050
2.309
3.087
1.329
2.323
-23.719
10.453
-2.269
21
0.32
1.00
4.30
0.785846
1.078
1.078
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-157.700
105.150
-1.500
-1.564
0.881
-1.775
13.259
8.652
1.532
9
0.35
0.65
1.60
0.3838
0.010
0.010
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
ME/NA
PARAM.VALUE
STD.ERROR
t STATISTIC
64.112
45.461
1.410
-0.185
0.695
-0.266
-4.531
3.729
-1.215
21
0.12
0.65
1.21
0.286252
5.480
5.480
RAF
PARAM.VALUE
STD. ERROR
t STATISTIC
70.472
36.261
1.944
0.366
0.771
0.474
-5.021
2.965
-1.694
12
0.46
3.31
3.87
0.086752
3. AAA
3. AAA
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
-28.588
9.581
-2.984
-0.924
0.359
-2.571
3.187
0.793
4.017
21
0.50
1.74
8.89
0.064944
8.388
8.388
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
-168.840
47.128
-3.583
-2.546
1.264
-2.014
14.513
3.874
3.746
21
0.52
0.87
9.69
0.228376
0.694
0.694
COMMB
PARAM.VALUE
1
0.002
0.002
STD.ERROR
t STATISTIC
Total 100.000100.000

110
Table 5.3 United States Product Demands
Partner
Region
Intercept
( + \-)
Relative
Price
(-)
Total
Market
Demand
<-\+>
OBS
RSQ
DW
FST
UTHEIL
21 Year Average
Z Z Total
Total Market
Imports Demand
CAN
PARAM.VALUE
STD. ERROR
t STATISTIC
75.967
84.489
0.899
1.436
1.118
1.285
-5.097
5.869
-0.868
13
0.17
1.42
1.04
0.39015
0.035
0.001
LA
PARAM.VALUE
STD.ERROR
t STATISTIC
-36.637
40.069
-0.914
1.051
0.370
2.840
3.298
2.785
1.184
21
0.31
1.68
4.06
0.197823
83.619
2.194
MED-EC
PARAM.VALUE
STD.ERROR
t STATISTIC
-388.180
153.100
-2.535
-0.857
1.473
-0.582
27.351
10.626
2.574
21
0.28
1.03
3.48
0.70272
2.790
0.073
EC
PARAM.VALUE
STD.ERROR
t STATISTIC
-7.761
68.958
-0.113
-1.214
0.364
-3.334
0.815
4.778
0.171
19
0.44
2.17
6.20
0.335905
0.036
0.001
RWE
PARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
ME/NA
PARAM.VALUE
STD.ERROR
t STATISTIC
-159.400
104.410
-1.527
-5.564
1.607
-3.464
11.692
7.248
1.613
21
0.43
1.90
6.77
0.265217
11.705
0.307
RAF
PARAM.VALUE
STD.ERROR
t STATISTIC
3
0.040
0.001
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
-8.263
14.097
-0.586
7.510
1.763
4.261
0.578
1.002
0.577
21
0.54
1.29
10.70
0.179367
1.671
0.044
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
-427.920
122.020
-3.507
-1.512
0.726
-2.082
30.060
8.461
3.553
10
0.75
0.78
10.45
0.288411
0.104
0.003
COM
PARAM.VALUE
0
0.000
0.000
SID.ERROR
t STATISTIC
100.000 2.624
Total

112
Table 5.5 Latin America Product Demands
Partner
Region
Intercept
(+\->
Relative
Price
(-)
Total
Market
Demand
(-H)
SOBS
SRSQ
SDW
SFST
UTHEIL
21 Year Average
1 Z Total
Total Market
Imports Demand
us
PARAM.VALUE
STD. ERROR
t STATISTIC
22.120
11.172
1.980
1.923
1.093
1.759
-1.060
0.711
-1.490
21
0.21
0.75
2.41
0.275535
88.135
0.031
CAN
PARAM. VALUE
STD. ERROR
t STATISTIC
-A.856
42.909
-0.113
-2.161
1.179
-1.833
0.634
2.754
0.230
7
0.48
0.09
1.87
0.417781
0.344
0.000
MED-EC
PARAM. VALUE
STD. ERROR
t STATISTIC
-75.104
38.319
-1.960
-2.047
2.615
-0.783
5.098
2.258
2.258
8
0.80
1.07
9.93
0.402184
0.796
0.000
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-18.663
23.599
-0.791
-0.106
1.368
-0.077
1.490
1.397
1.067
21
0.12
1.01
1.19
0.352092
5.706
0.002
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
1
0.041
0.000
ME/NA
PARAM.VALUE
STD.ERROR
t STATISTIC
50.881
69.445
0.733
-2.980
2.725
-1.093
-2.791
4.322
-0.646
11
0.14
1.67
0.65
0.786243
2.175
0.001
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
A
0.090
0.000
FE
PARAM. VALUE
STD. ERROR
t STATISTIC
6
0.262
0.000
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
2
0.086
0.000
COMMB
PARAM. VALUE
STD.ERROR
t STATISTIC
-213.950
96.979
-2.206
1.673
4.521
0.370
13.613
6.109
2.228
8
0.51
0.28
2.60
0.479401
2.365
0.001
100.000 0.035
Total

113
Table 5.6 Mediterranean-EC Product Demands
Total 21 Year Average
Relative Market X Z Total
Partner Intercept Price Demand Total Market
Region (+\-) (-) (-\+) SOBS @RSQ @DW @FST UTHEIL Imports Demand
US
PARAM. VALUE
STD.ERROR
t STATISTIC
5
0.085
0.000
CAN
PARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-117.140
113.960
-1.028
1.777
4.482
0.397
8.218
7.737
1.062
10
0.27
1.67
1.28
0.736229
22.012
0.022
EC
PARAM. VALUE
STD. ERROR
t STATISTIC
-155.000
49.751
-3.115
-4.717
1.020
-4.625
11.146
3.328
3.349
15
0.82
2.25
27.61
0.287941
47.784
0.048
RWE
PARAM. VALUE
STD. ERROR
t STATISTIC
-185.810
35.174
-5.283
-0.771
0.646
-1.193
12.805
2.376
5.388
12
0.76
1.18
14.53
0.354966
1.747
0.002
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
-29.011
44.476
-0.652
-0.869
0.851
-1.022
2.365
3.003
0.788
16
0.10
1.86
0.73
0.451283
17.668
0.018
RAF
PARAM.VALUE
STD.ERROR
t STATISTIC
5
9.973
0.010
FE
PARAM. VALUE
STD. ERROR
t STATISTIC
2
0.310
0.000
OCE
PARAM. VALUE
STD. ERROR
t STATISTIC
1
0.042
0.000
COMMB
PARAM.VALUE
STD.ERROR
t STATISTIC
3
0.372
0.000
Total
100.000
0.100

114
Table 5.7 EC Product Demands
Partner
Region
Intercept
( + \-)
Relative
Pries
(-)
Total
Market
Demand
(-\ + )
@OBS
@RSQ
@DW
@FST
UTHEIL
21 Year Average
Z Z Total
Total Market
Imports Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
70.236
47.110
1.491
-4.896
1.594
-3.072
-3.938
3.191
-1.234
21
0.36
2.03
5.12
0.279023
1.821
1.812
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
60.184
114.670
0.525
-1.616
2.825
-0.572
-3.856
7.778
-0.496
14
0.04
1.34
0.25
0.726798
0.002
0.002
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-27.456
19.627
-1.399
0.199
0.699
0.285
2.625
1.331
1.972
21
0.19
0.74
2.16
0.189179
3.278
3.265
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-13.094
9.152
-1.431
0.755
1.191
0.634
1.846
0.620
2.976
21
0.42
1.45
6.61
0.075885
54.192
53.955
RWE
PARAM. VALUE
STD. ERROR
t STATISTIC
93.332
41.206
2.265
-2.261
1.084
-2.085
-5.828
2.774
-2.101
21
0.23
1.44
2.66
0.251156
0.030
0.029
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
11.630
8.549
1.361
1.868
0.783
2.387
0.144
0.582
0.247
21
0.27
0.56
3.30
0.065822
32.690
32.548
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
12.939
9.060
1.428
-0.712
0.537
-1.326
-0.042
0.616
-0.068
21
0.10
0.57
0.95
0.070983
7.691
7.656
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
-342.120
193.120
-1.772
4.105
2.517
1.631
23.361
13.018
1.795
21
0.15
1.60
1.61
0.626029
0.007
0.007
OCE
PARAM.VALUE
STD.ERROR
t STATISTIC
-151.980
65.557
-2.318
5.085
2.766
1.838
10.671
4.410
2.420
21
0.27
2.48
3.28
0.286791
0.108
0.106
COMMB
PARAM. VALUE
STD.ERROR
t STATISTIC
-183.660
86.698
-2.118
-9.158
2.519
-3.636
12.889
5.882
2.191
20
0.57
0.71
11.08
0.44278
0.181
0.180
Total 100.000 99.560

115
Table 5.8 Rest of Western Europe Product Demands
Partner
Region
Intercept
(+\->
Relative
Price
(-)
Total
Market
Demand
(*\ + )
@OBS
@RSQ
@DW
@FST
UTHEIL
21 Year Average
% Z Total
Total Market
Imports Demand
us
FARAM. VALUE
STD.ERROR
t STATISTIC
151.270
53.762
2.814
-8.876
3.255
-2.727
-10.850
4.102
-2.645
21
0.33
2.50
4.44
0.248259
0.941
0.941
CAN
FARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM.VALUE
STD.ERROR
t STATISTIC
-7.246
14.239
-0.509
-0.653
0.513
-1.274
1.200
1.100
1.091
21
0.13
1.29
1.36
0.143369
1.022
1.022
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-8.745
10.036
-0.871
-2.700
1.061
-2.544
1.612
0.776
2.078
21
0.28
1.24
3.41
0.086744
43.412
43.412
EC
FARAM. VALUE
STD.ERROR
t STATISTIC
-24.543
14.097
-1.741
-2.654
0.333
-7.974
2.660
1.084
2.454
21
0.85
2.05
50.81
0.160084
1.501
1.501
ME/NA
PARAM.VALUE
STD.ERROR
t STATISTIC
9.311
5.635
1.652
-2.666
0.504
-5.289
0.212
0.437
0.486
21
0.73
1.35
23.85
0.047351
47.547
47.547
RAF
PARAM.VALUE
STD. ERROR
t STATISTIC
-0.171
6.415
-0.027
0.622
0.513
1.212
0.782
0.494
1.582
21
0.15
1.22
1.55
0.062721
5.327
5.327
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
58.197
129.350
0.450
-0.754
1.827
-0.413
-4.150
9.953
-0.417
17
0.01
2.32
0.10
0.740192
0.039
0.039
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
-91.473
42.710
-2.142
1.792
1.181
1.518
7.456
3.296
2.262
21
0.28
0.89
3.51
0.330177
0.163
0.163
COMMB
PARAM.VALUE
STD.ERROR
t STATISTIC
27.034
33.816
0.799
5.388
1.662
3.243
-1.672
2.613
-0.640
18
0.42
1.98
5.50
0.280677
0.048
0.048
Total 100.000100.000

116
Table 5.9 Middle East/North Africa Product Demands
Partner
Region
Intercept
(+\-)
Relative
Price
(-)
Total
Market
Demand
C-\+)
SOBS
0RSQ
@DW
@FST
UTHEIL
21 Year Average
Z Z Total
Total Market
Imports Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
-16.023
29.119
-0.550
-2.293
1.337
-1.715
1.551
1.900
0.816
18
0.34
1.20
3.83
0.875418
1.493
0.033
CAN
PARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM.VALUE
STD.ERROR
t STATISTIC
-93.477
57.540
-1.625
1.807
2.541
0.711
6.749
3.825
1.764
13
0.29
2.05
2.03
0.439828
14.926
0.334
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-47.015
22.077
-2.130
0.202
1.152
0.175
3.655
1.493
2.448
18
0.32
1.17
3.60
0.610091
8.371
0.187
EC
PARAM.VALUE
STD.ERROR
t STATISTIC
-30.816
18.524
-1.664
-2.826
0.944
-2.994
2.576
1.243
2.072
16
0.48
2.79
6.05
0.662988
1.122
0.025
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
-46.630
36.562
-1.275
-0.857
1.450
-0.591
3.596
2.453
1.466
11
0.24
1.48
1.24
0.51107
1.849
0.041
RAF
PARAM.VALUE
STD.ERROR
t STATISTIC
-73.883
19.749
-3.741
-1.071
3.696
-0.290
5.605
1.389
4.036
17
0.60
0.72
10.58
0.356634
28.481
0.637
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
-39.437
12.263
-3.216
-0.670
0.468
-1.431
3.294
0.824
3.998
21
0.57
0.92
11.83
0.365248
39.474
0.883
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
-62.556
10.678
-5.858
-1.459
0.770
-1.895
4.743
0.710
6.678
20
0.77
1.41
27.91
0.350599
4.230
0.095
COMMB
PARAM.VALUE
STD.ERROR
t STATISTIC
-119.410
80.654
-1.481
-3.702
1.415
-2.616
8.263
5.376
1.537
7
0.64
1.45
3.55
0.594771
0.054
0.001
Total
100.000
2.237

117
Table 5.10 Rest of Africa Product Demands
Partner
Region
Intercept
C + \-)
Relative
Price
(*)
Total
Market
Demand
C-\+)
@OBS
@RSQ
@DW
@FST
UTHEIL
21 Year Average
Z Z Total
Total Market
Imports Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
51.498
37.908
1.359
-1.816
0.990
-1.835
-3.544
2.797
-1.267
17
0.20
0.75
1.75
0.444633
0.183
0.002
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
3.924
36.723
0.107
-3.669
0.629
-5.833
0.185
2.751
0.067
21
0.66
1.68
17.09
0.440843
5.890
0.061
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-66.464
31.232
-2.769
-0.367
1.006
-0.365
6.938
2.298
3.019
21
0.48
1.16
8.33
0.354209
7.812
0.081
EC
PARAM.VALUE
STD.ERROR
t STATISTIC
-7.827
8.884
-0.881
0.166
0.476
0.349
1.081
0.648
1.667
21
0.14
2.12
1.47
0.123771
13.859
0.144
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
1
0.047
0.000
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
4.764
6.913
0.689
0.654
0.363
1.799
0.251
0.513
0.490
21
0.16
2.41
1.72
0.092017
68.176
0.708
FE
PARAM. VALUE
STD. ERROR
t STATISTIC
97.502
38.647
2.523
-2.498
0.572
-4.371
-6.961
2.869
-2.426
13
0.66
1.25
9.56
0.367393
0.149
0.002
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
24.628
26.746
0.921
-5.716
1.193
-4.790
-1.179
1.998
-0.590
21
0.56
1.39
11.48
0.507281
3.830
0.040
COMMB
PARAM. VALUE
STD.ERROR
t STATISTIC
3
0.054
0.001
Total
100.000
1.039

118
Table 5.11 Far East Product Demands
Partner
Region
Intercept
( + \-)
Relative
Price
(-)
Total
Market
Demand
C-\ + )
@OBS
@Rsq
1 @DW
@FST
UTHEIL
21 Year Average
1 1 Total
Total Market
Imports Demand
us
PARAM.VALUE
STD.ERROR
t STATISTIC
-30.208
3.885
-7.776
0.570
0.198
2.887
2.666
0.254
10.485
21
0.93
2.25124.47
0.073596
81.384
2.361
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
-51.580
47.757
-1.080
-2.940
1.502
-1.958
3.662
3.098
1.182
9
0.40
1.01
1.97
0.755161
0.043
0.001
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-43.310
31.301
-1.384
-3.517
0.951
-3.700
3.155
2.018
1.564
20
0.46
2.34
7.21
0.59917
0.305
0.009
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-24.850
12.777
-1.945
0.438
0.815
0.538
2.012
0.833
2.415
21
0.45
1.71
7.28
0.198234
0.540
0.016
EC
PARAM.VALUE
STD.ERROR
t STATISTIC
-58.658
23.280
-2.520
-3.013
1.389
-2.169
4.084
1.527
2.674
20
0.32
1.12
4.09
0.693573
0.022
0.001
RWE
PARAM. VALUE
STD. ERROR
t STATISTIC
41.784
18.245
2.290
1.481
0.759
1.950
-2.612
1.188
-2.198
13
0.34
0.92
2.58
0.284931
0.003
0.000
ME/NA
PARAM. VALUE
STD. ERROR
t STATISTIC
10.014
7.069
1.417
-0.782
0.278
-2.813
-0.026
0.457
-0.058
21
0.46
1.62
7.74
0.14595
7.575
0.220
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
-33.078
44.401
-0.745
-2.827
1.686
-1.677
2.688
2.874
0.935
21
0.15
0.69
1.60
0.397977
3.529
0.102
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
-8.334
10.302
-0.809
-0.125
0.520
-0.240
1.129
0.668
1.690
21
0.16
0.76
1.67
0.226227
6.597
0.191
COMMB
PARAM. VALUE
STD.ERROR
t STATISTIC
-96.284
54.672
-1.761
-0.488
1.179
-0.414
6.273
3.523
1.781
9
0.42
2.56
2.16
0.3465
0.002
0.000
Total
100.000
2.901

119
Table 5.12 Oceania Product Demands
Partner
Region
Intercept
(+\->
Relative
Price
(-)
Total
Market
Demand
(-\ + )
SOBS
@RSQ
@DW
@FST
UTHEIL
21 Year Average
Z Z Total
Total Market
Imports Demand
us
PARAM. VALUE
STD. ERROR
t STATISTIC
-23.433
18.241
-1.285
-1.235
0.793
-1.558
2.663
1.494
1.782
21
0.21
0.53
2.45
0.268945
85.922
3.342
CAN
PARAM.VALUE
STD. ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM. VALUE
STD. ERROR
t STATISTIC
56.815
27.151
2.093
3.978
2.907
1.368
-4.127
2.208
-1.869
7
0.72
2.09
5.19
0.110958
4.504
0.175
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-69.670
71.041
-0.981
5.089
2.053
2.479
5.972
5.829
1.025
7
0.67
0.52
4.13
0.389675
1.868
0.073
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
5
0.027
0.001
EWE
PARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
197.030
55.499
3.550
-6.590
1.154
-5.710
-15.478
4.520
-3.424
13
0.77
1.25
16.68
0.375281
4.878
0.190
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
6
2.626
0.102
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
-22.286
83.976
-0.265
-1.408
4.445
-0.317
2.104
6.744
0.312
7
0.08
0.11
0.18
0.821817
0.175
0.007
COMMB
PARAM.VALUE
0
0.000
0.000
STD.ERROR
t STATISTIC
100.000 3.890
Total

120
Table 5.13 Communist Bloc Product Demands
Partner
Region
Intercept
(+\-)
Relative
Price
(-)
Total
Market
Demand
(-\ + )
@OBS
@RSQ
@DW
@FST
UTHEIL
21 Year Average
Z Z Total
Total Market
Imports Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
3.212
32.681
0.098
-9.518
3.602
-2.642
0.484
2.424
0.200
9
0.57
1.14
4.01
0.481618
0.486
0.283
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-37.344
12.928
-2.889
-3.402
1.060
-3.208
3.375
0.936
3.606
20
0.44
1.98
6.79
0.22123
2.764
1.612
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
8.301
2.802
2.963
0.143
0.523
0.273
0.280
0.204
1.371
21
0.11
2.10
1.11
0.106237
35.192
20.526
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-5.284
7.567
-0.698
-2.190
1.531
-1.430
0.925
0.553
1.674
21
0.20
1.50
2.24
0.293654
0.069
0.040
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
-53.550
21.610
-2.478
-0.333
1.278
-0.260
4.159
1.545
2.691
15
0.43
0.32
4.50
0.620833
0.013
0.008
ME/NA
PARAM.VALUE
STD.ERROR
t STATISTIC
-3.883
3.999
-0.971
-1.549
0.371
-4.175
1.237
0.296
4.177
21
0.55
1.85
10.97
0.116997
61.439
35.836
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
3
0.003
0.002
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
-105.470
68.491
-1.540
8.172
5.707
1.432
7.847
4.871
1.611
9
0.30
0.92
1.30
0.576348
0.032
0.018
OCE
PARAM. VALUE
STD. ERROR
t STATISTIC
4
0.002
0.001
Total
100.000
58.326

Table 5.14 United States C1F Price Linkage Equations
Partner
Region
Intercept
(+\->
FOB
Price
( + )
Year
Trend
<-\+)
Energy
Index
Price
( + )
@OBS
@RSQ
@DW
@FST
21 Years
Z of
Total
UTHEIL Imports
Average
X Total
Market
Demand
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
-17.809
15.161
-1.175
0.596
0.263
2.271
4.148
3.520
1.178
-0.116
0.186
-0.624
21
0.68
2.20
11.92
0.191764
0.035
0.001
LA
PARAM.VALUE
STD.ERROR
t STATISTIC
20.197
10.327
1.956
1.098
0.498
2.203
-4.576
2.169
-2.110
0.182
0.068
2.682
21
0.77
2.13
19.20
0.079479
83.619
2.194
MED-EC
PARAM. VALUE
STD. ERROR
t STATISTIC
6.593
7.346
0.898
0.884
0.315
2.804
-1.500
1.641
-0.914
0.163
0.109
1.493
21
0.93
2.36
72.38
0.048339
2.790
0.073
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
1.660
10.364
0.160
0.987
0.209
4.728
-0.272
2.442
-0.111
0.047
0.146
0.320
21
0.68
1.37
11.84
0.134116
0.036
0.001
RWE
PARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
-10.247
8.365
-1.225
0.736
0.331
2.223
2.403
1.953
1.230
0.032
0.160
0.202
21
0.83
2.34
27.35
0.100822
11.705
0.307
RAF
PARAM.VALUE
STD.ERROR
t STATISTIC
3
0.040
0.001
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
0.589
8.295
0.071
1.131
0.350
3.230
-0.067
1.882
-0.036
-0.043
0.064
-0.670
21
0.94
1.61
87.50
0.051425
1.671
0.044
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
-6.967
16.075
-0.433
1.001
0.490
2.042
1.749
3.827
0.457
-0.084
0.332
-0.254
21
0.59
2.47
8.22
0.334228
0.104
0.003
COMMB
PARAM. VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
Total
100.000
2.624

Table 5.15 Canada CIF Price Linkage Equations
Energy 21 Years Average
Partner
Region
Intercept
<+\->
FOB
Price
(+)
Year
Trend
<-\+)
Index
Price
( + )
@OBS
@RSQ
gDW
@FST
UTHEIL
Z of
Total
Imports
Z Total
Market
Demand
us
PARAM.VALUE
STD.ERROR
t STATISTIC
-15.806
15.849
-0.997
0.215
0.781
0.275
3.462
3.465
0.999
0.018
0.089
0.205
21
0.81
1.82
24.55
0.09676
79.673
79.673
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-18.427
9.063
-2.033
0.109
0.261
0.418
3.902
2.046
1.907
0.154
0.105
1.476
21
0.87
1.95
39.33
0.123418
1.230
1.230
MED-EC
PARAM.VALUE
STD. ERROR
t STATISTIC
-4.863
9.817
-0.495
0.850
0.327
2.598
1.098
2.324
0.473
0.094
0.198
0.476
21
0.82
2.54
26.13
0.154421
1.078
1.078
EC
PARAM. VALUE
STD. ERROR
t STATISTIC
16.953
13.066
1.298
0.985
0.332
2.969
-4.019
2.995
-1.342
0.307
0.150
2.047
21
0.87
2.12
37.59
0.215998
0.010
0.010
RWE
PARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
0.160
8.159
0.020
0.831
0.286
2.904
-0.090
1.872
-0.048
0.095
0.137
0.698
21
0.86
2.73
34.64
0.141848
5.480
5.480
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
-0.845
7.786
-0.109
1.058
0.503
2.103
0.260
1.719
0.151
-0.011
0.192
-0.057
21
0.88
2.16
42.52
0.123417
3.444
3.444
FE
PARAM. VALUE
STD. ERROR
t STATISTIC
-1.558
10.818
-0.144
0.966
0.468
2.063
0.388
2.445
0.159
0.003
0.087
0.037
21
0.89
1.18
44 A0
0.093371
8.388
8.388
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
3.626
7.381
0.491
0.363
0.393
0.925
-1.048
1.670
-0.628
0.278
0.095
2.922
21
0.82
1.56
26.01
0.089569
0.694
0.694
COM1B
PARAM. VALUE
1
0.002
0.002
STD. ERROR
t STATISTIC
Total 100.000100.000

Table 5.16 Latin America CIF Price Linkage Equations
Energy 21 Years Average
Partner
Region
Intercept
( + \-)
FOB
Price
( + )
Year
Trend
(-\ + )
Index
Price
( + )
@OBS
@RSQ
@DW
@FST
UTHEIL
Z of
Total
Imports
Z Total
Market
Demand
us
PARAM.VALUE
STD.ERROR
t STATISTIC
29.787
20.140
1.479
2.960
1.321
2.241
-6.093
4.196
-1.452
-0.246
0.267
-0.920
21
0.87
2.22
37.70
0.145829
88.135
0.031
CAN
PARAM.VALUE
STD.ERROR
t STATISTIC
9.739
18.950
0.514
0.689
0.275
2.510
-2.441
4.389
-0.556
0.407
0.209
1.950
21
0.80
2.11
21.99
0.129447
0.344
0.000
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-0.883
11.891
-0.074
0.726
0.338
2.151
0.128
2.759
0.047
0.170
0.183
0.928
21
0.76
2.03
18.38
0.181972
0.796
0.000
EC
PARAM.VALUE
STD. ERROR
t STATISTIC
-3.003
6.741
-0.445
0.975
0.322
3.025
0.775
1.555
0.498
0.004
0.115
0.034
21
0.86
2.58
33.91
0.094185
5.706
0.002
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
1
0.041
0.000
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
-8.814
28.433
-0.310
0.649
0.535
1.214
2.005
6.582
0.305
0.032
0.326
0.098
21
0.39
2.30
3.65
0.248192
2.175
0.001
RAF
PARAM.VALUE
STD.ERROR
t STATISTIC
4
0.090
0.000
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
6
0.262
0.000
OCE
PARAM.VALUE
STD.ERROR
t STATISTIC
2
0.086
0.000
COMffl
PARAM. VALUE
STD.ERROR
t STATISTIC
-7.965
16.448
-0.484
0.408
0.485
0.841
1.640
3.701
0.443
0.186
0.139
1.338
21
0.78
2.20
19.89
0.179599
2.365
0.001
100.000 0.035
Total

Table 5.17 Mediterranean-EC CIF Price Linkage Equations
Partner
Region
Intercept
( + \-)
FOB
Price
< + )
Year
Trend
(-\ + )
Energy
Index
Price
( + >
gOBS
gRSQ
§DW
gFST
21 Years
Z of
Total
UTHEIL Imports
Average
Z Total
Market
Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
5
0.085
0.000
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
0.000
0.000
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-26.756
15.945
-1.678
0.090
0.464
0.193
5.846
3.578
1.634
0.109
0.177
0.618
21
0.78
1.84
20.03
0.193848
22.019
0.022
EC
PARAM.VALUE
STD.ERROR
t STATISTIC
-11.072
9.503
-1.165
0.222
0.522
0.426
2.425
2.211
1.097
0.059
0.166
0.357
21
0.57
2.54
7.66
0.190169
47.784
0.048
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
-17.337
13.007
-1.333
1.088
0.338
3.219
4.199
2.987
1.406
-0.289
0.134
-2.155
21
0.84
2.72
29.74
0.092578
1.747
0.002
ME/NA
PARAM. VALUE
STD. ERROR
t STATISTIC
27.404
20.916
1.310
1.192
0.442
2.698
-6.370
4.805
-1.326
0.366
0.249
1.473
21
0.64
2.52
10.20
0.315843
17.668
0.018
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
5
9.973
0.010
FE
PARAM. VALUE
STD. ERROR
t STATISTIC
2
0.310
0.000
OCE
PARAM.VALUE
STD.ERROR
t STATISTIC
1
0.042
0.000
COMMB
PARAM. VALUE
STD.ERROR
t STATISTIC
3
0.372
0.000
Total
100.000
0.100

Table 5.18 EC CIF Price Linkage Equations
Energy 21 Years Average
Partner
Region
Intercept
( + \-)
FOB
Price
( + )
Year
Trend
(-\+)
Index
Price
( + )
SOBS
@RSQ
@DW
@FST
UTHEIL
Z of
Total
Imports
Z Total
Market
Demand
us
PARAM.VALUE
STD.ERROR
t STATISTIC
-2.268
8.547
-0.265
0.556
0.264
2.103
0.399
1.945
0.205
0.128
0.082
1.557
21
0.85
2.10
31.16
0.087405
1.821
1.812
CAN
PARAM.VALUE
STD.ERROR
t STATISTIC
5.290
13.454
0.393
1.021
0.468
2.183
-1.167
3.070
-0.380
0.028
0.193
0.144
21
0.62
2.62
9.32
0.193149
0.002
0.002
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-6.349
7.108
-0.893
0.210
0.211
0.995
1.167
1.581
0.738
0.176
0.074
2.370
21
0.90
1.92
52.34
0.073432
3.278
3.265
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
6.695
3.022
2.216
1.065
0.1A 7
7.241
-1.468
0.671
-2.190
-0.011
0.051
-0.220
21
0.97
1.97
184.49
0.042842
54.192
53.955
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
2.582
5.782
0.447
1.274
0.356
3.576
-0.405
1.349
-0.300
-0.121
0.123
-0.981
21
0.74
2.20
15.87
0.09092
0.030
0.029
ME/NA
PARAM.VALUE
STD.ERROR
t STATISTIC
-2.449
3.136
-0.781
1.063
0.233
4.557
0.671
0.705
0.952
-0.030
0.064
-0.467
21
0.94
2.10
89.99
0.045453
32.690
32.548
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
-0.561
4.433
-0.127
0.579
0.230
2.523
0.005
0.986
0.005
0.102
0.068
1.490
21
0.93
2.77
71.77
0.056549
7.691
7.656
FE
PARAM.VALUE
STD.ERROR
t STATISTIC
2.336
8.875
0.263
0.935
0.239
3.915
-0.479
2.110
-0.227
-0.033
0.157
-0.212
21
0.66
2.21
10.97
0.126091
0.007
0.007
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
0.820
5.792
0.142
0.886
0.386
2.295
-0.145
1.406
-0.103
-0.016
0.168
-0.092
21
0.78
1.83
20.58
0.075877
0.108
0.106
COMMB
PARAM.VALUE
STD. ERROR
t STATISTIC
2.951
4.699
0.628
0.895
0.244
3.671
-0.682
1.051
-0.649
0.042
0.072
0.590
21
0.88
2.34
40.39
0.05886
0.181
0.180
Total 100.000 99.560

Table 5.19 Rest of Western Europe CIF Price Linkage Equations
Energy 21 Years Average
Partner
Region
Intercept
( + \-)
FOB
Price
( + )
Year
Trend
(-\ + )
Index
Price
( + )
@OBS
6RSQ
@DW
@FST
UTHEIL
Z of
Total
Imports
Z Total
Market
Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
-0.318
6.950
-0.046
0.632
0.267
2.365
0.014
1.561
0.009
0.123
0.069
1.788
21
0.90
1.36
52.20
0.081064
0.941
0.941
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-17.386
7.065
-2.461
0.040
0.215
0.185
3.696
1.544
2.394
0.154
0.046
3.377
21
0.97
1.81
207.61
0.044011
1.022
1.022
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-4.745
2.502
-1.897
0.863
0.122
7.095
1.132
0.569
1.991
-0.004
0.044
-0.103
21
0.98
1.30
242.65
0.040091
43.412
43.412
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
8.083
4.781
1.691
1.301
0.296
4.395
-1.738
1.112
-1.563
-0.012
0.092
-0.129
21
0.81
2.81
24.78
0.060969
1.501
1.501
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
-4.677
1.952
-2.395
1.074
0.132
8.156
1.217
0.432
2.819
-0.047
0.035
-1.334
21
0.98
1.80
368.62
0.029
47.547
47.547
RAF
PARAM.VALUE
STD.ERROR
t STATISTIC
-2.233
2.433
-0.918
0.650
0.135
4.807
0.429
0.559
0.768
0.122
0.049
2.458
21
0.98
2.27
225.44
0.032975
5.327
5.327
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
-6.418
10.699
-0.600
0.831
0.291
2.854
1.539
2.536
0.607
-0.024
0.188
-0.126
21
0.68
2.76
11.94
0.149779
0.039
0.039
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
7 759
4.459
1.740
0.644
0.245
2.627
-1.860
1.029
-1.808
0.238
0.087
2.724
21
0.91
2.53
58.78
0.050678
0.163
0.163
COMMB
PARAM.VALUE
STD.ERROR
t STATISTIC
-3.146
5.382
-0.584
0.606
0.159
3.820
0.627
1.218
0.515
0.089
0.051
1.732
21
0.94
2.71
86.65
0.06096
0.048
0.048
Total 100.000100.000

Table 5.20 Middle East/North Africa CIF Price Linkage Equations
Partner
Region
Intercept
c+\->
FOB
Price
( + )
Year
Trend
(*\+)
Energy-
Index
Price
< + )
@OBS
@RSQ
@DW
@FST
21 Years
Z of
Total
UTHEIL Imports
Average
Z Total
Market
Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
0.181
27.620
0.007
0.983
0.488
2.014
0.033
6.534
0.005
-0.001
0.412
-0.003
21
0.21
2.44
1.53
0.789752
1.493
0.033
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM. VALUE
STD. ERROR
t STATISTIC
-2.352
10.625
-0.221
0.739
0.366
2.019
0.507
2.334
0.217
0.048
0.112
0.426
21
0.88
2.34
42.08
0.101175
14.926
0.334
MED-EC
PARAM.VALUE
STD.ERROR
t STATISTIC
12.745
13.267
0.961
0.789
0.408
1.935
-3.046
3.130
-0.973
0.251
0.256
0.980
21
0.64
2.73
9.89
0.461457
8.371
0.187
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-10.257
13.997
-0.733
0.872
0.300
2.905
2.450
3.240
0.756
-0.088
0.160
-0.551
21
0.71
2.40
13.99
0.198066
1.122
0.025
RWE
PARAM.VALUE
STD.ERROR
t STATISTIC
15.652
21.853
0.716
1.167
0.483
2.414
-3.583
5.008
-0.716
0.205
0.217
0.943
21
0.61
2.84
8.85
0.308409
1.849
0.041
RAF
PARAM.VALUE
STD.ERROR
t STATISTIC
-4.077
4.985
-0.818
0.433
0.177
2.450
0.746
1.155
0.646
0.157
0.081
1.940
21
0.90
2.29
52.72
0.070873
28.481
0.637
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
10.152
15.306
0.663
1.243
0.365
3.407
-2.223
3.667
-0.606
0.164
0.244
0.675
21
0.50
3.11
5.69
0.289771
39.474
0.883
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
15.765
12.024
1.311
1.004
0.467
2.147
-3.679
2.807
-1.311
0.259
0.184
1.406
21
0.52
2.34
6.23
0.222937
4.230
0.095
cot*
PARAM. VALUE
STD.ERROR
t STATISTIC
9.986
22.097
0.452
1.407
0.599
2.347
-2.049
4.981
-0.411
-0.073
0.200
-0.365
21
0.71
2.05
13.87
0.264828
0.054
0.001
100.000 2.237
Total

Table 5.21 Rest of Africa CIF Price Linkage Equations
Partner
Region
Intercept
( + \->
FOB
Price
( + )
Year
Trend
<-\ + )
Energy
Index
Price
(+)
@OBS
@RSQ
@DW
@FST
21 Years
Z of
Total
UTHEIL Imports
Average
Z Total
Market
Demand
US
PARAM.VALUE
STD.ERROR
t STATISTIC
-1.372
10.261
-0.134
1.033
0.299
3.456
0.478
2.414
0.198
-0.065
0.163
-0.397
21
0.67
1.95
11.26
0.193658
0.183
0.002
CAN
PARAM. VALUE
STD. ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-2.396
7.206
-0.333
1.026
0.135
7.611
0.715
1.685
0.424
-0.055
0.101
-0.544
21
0.88
1.77
42.74
0.109284
5.890
0.061
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
-25. m
13.378
-1.880
0.136
0.482
0.283
5.798
3.038
1.909
-0.116
0.128
-0.908
21
0.67
2.83
11.74
0.148992
7.812
0.081
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
0.483
2.724
0.177
1.087
0.191
5.700
0.066
0.612
0.107
-0.072
0.055
-1.312
21
0.96
1.10
149.29
0.032802
13.859
0.144
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
1
0.047
0.000
ME
PARAM. VALUE
STD.ERROR
t STATISTIC
-11.651
5.737
-2.031
0.578
0.276
2.099
2.698
1.302
2.073
-0.064
0.098
-0.652
21
0.85
1.75
32.86
0.086437
68.176
0.708
FE
PARAM. VALUE
STD.ERROR
t STATISTIC
-10.019
12.852
-0.780
1.390
0.504
2.760
2.607
3.031
0.860
-0.222
0.202
-1.099
21
0.43
2.50
4.31
0.200405
0.149
0.002
OCE
PARAM.VALUE
STD.ERROR
t STATISTIC
-8.542
9.153
-0.933
1.239
0.371
3.340
2.251
2.233
1.008
-0.205
0.216
-0.946
21
0.75
2.24
16.85
0.207496
3.830
0.040
COM
PARAM. VALUE
STD.ERROR
t STATISTIC
3
0.054
0.001
Total
100.000
1.039
128

Table 5.22 Far East CIF Price Linkage Equations
Partner
Region
Intercept
( + \)
FOB
Price
(+)
Year
Trend
<-\ + >
Energy
Index
Price
< + )
@OBS
@RSQ
@DW
0FST
21 Years
Z of
Total
UTHEIL Imports
Average
Z Total
Market
Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
-0.165
6.523
-0.025
0.792
0.306
2.588
0.086
1.434
0.060
-0.016
0.062
-0.253
21
0.93
2.51
70.94
0.054252
81.384
2.361
CAN
PARAM. VALUE
STD. ERROR
t STATISTIC
11.208
11.286
0.993
0.792
0.184
4.293
-2.603
2.605
-0.999
0.159
0.117
1.357
21
0.83
2.13
27.28
0.113137
0.043
0.001
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-4.972
9.880
-0.503
0.957
0.378
2.533
1.269
2.257
0.563
-0.005
0.152
-0.034
21
0.79
2.11
21.80
0.175802
0.305
0.009
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
3.120
7.214
0.433
0.917
0.405
2.267
-0.660
1.535
-0.430
0.011
0.057
0.200
21
0.94
2.81
95.22
0.041532
0.540
0.016
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
11.625
13.630
0.853
0.985
0.362
2.723
-2.652
3.143
-0.844
0.123
0.144
0.852
21
0.60
2.28
8.38
0.12125
0.022
0.001
RWE
PARAM. VALUE
STD.ERROR
t STATISTIC
-4.253
11.484
-0.370
0.725
0.358
2.025
0.997
2.591
0.385
-0.002
0.114
-0.014
21
0.81
3.49
23.88
0.136008
0.003
0.000
ME/NA
PARAM.VALUE
STD.ERROR
t STATISTIC
14.364
13.506
1.064
2.523
0.757
3.333
-2.509
2.862
-0.877
-0.155
0.106
-1.460
21
0.87
2.35
38.49
0.13241
7.575
0.220
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
-4.255
9.721
-0.438
0.600
0.377
1.591
0.916
2.123
0.432
0.062
0.064
0.962
21
0.96
2.58
139.78
0.04477
3.529
0.102
OCE
PARAM.VALUE
STD.ERROR
t STATISTIC
-3.673
3.560
-1.032
0.499
0.211
2.359
0.753
0.782
0.963
0.112
0.044
2.518
21
0.96
2.16
123.62
0.036083
6.597
0.191
COMMB
PARAM. VALUE
STD.ERROR
t STATISTIC
0.290
4.576
0.063
0.612
0.244
2.512
-0.116
1.042
-0.112
0.077
0.091
0.848
21
0.88
2.19
42.24
0.079808
0.002
0.000
Total
100.000
2.901

Table 5.23 Oceania CIF Price Linkage Equations
Partner
Region
Intercept
( + \-)
FOB
Price
(+)
Year
Trend
<-\ + >
Energy
Index
Price
< + )
gOBS
gRSQ
gDW
gFST
21 Years
Z of
Total
UTHEIL Imports
Average
Z Total
Market
Demand
us
PARAM. VALUE
STD.ERROR
t STATISTIC
5.489
5.354
-1.025
0.630
0.176
3.572
1.268
1.218
1.042
-0.011
0.056
-0.206
21
0.89
2.17
48.21
0.060483
85.922
3.342
CAN
PARAM. VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
LA
PARAM. VALUE
STD.ERROR
t STATISTIC
-3.085
7.077
-0.436
0.290
0.324
0.896
0.478
1.644
0.291
0.169
0.110
1.540
21
0.68
1.68
12.30
0.101043
4.504
0.175
MED-EC
PARAM.VALUE
STD.ERROR
t STATISTIC
-7.934
11.732
-0.676
0.131
0.442
0.296
1.526
2.616
0.583
0.257
0.149
1.732
21
0.80
2.43
22.42
0.106035
1.868
0.073
EC
PARAM.VALUE
STD.ERROR
t STATISTIC
5
0.027
0.001
RWE
PARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
2.771
12.795
0.217
0.616
0.290
2.125
-0.720
3.091
-0.233
0.158
0.262
0.604
21
0.65
1.95
10.50
0.444433
4.878
0.190
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
6
2.626
0.102
FE
PARAM.VALUE
STD.ERROR
t STATISTIC
0.127
14.163
0.009
0.867
0.468
1.854
-0.005
3.320
-0.001
0.075
0.211
0.356
21
0.46
2.10
4.89
0.311537
0.175
0.007
OCE
PARAM.VALUE
STD.ERROR
t STATISTIC
0
0.000
0.000
Total
100.000
3.890

Table 5.24 Communist Bloc CIF Price Linkage Equations
Partner
Region
Intercept
< + \->
FOB
Price
( + )
Year
Trend
C-\ + )
Energy
Index
Price
(+)
@OBS
@RSQ
@DW
@FST
21 Years
Z of
Total
UTHEIL Imports
Average
Z Total
Market
Demand
US
PARAM.VALUE
STD.ERROR
t STATISTIC
7.433
16.852
0.441
1.244
0.354
3.513
-1.620
3.849
-0.421
0.058
0.142
0.407
21
0.86
2.04
34.89
0.125012
0.486
0.283
CAN
PARAM.VALUE
STD.ERROR
t STATISTIC
0.000
0.000
LA
PARAM.VALUE
STD.ERROR
t STATISTIC
13.521
10.845
1.247
1.583
0.435
3.636
-2.826
2.357
-1.199
0.004
0.117
0.035
21
0.92
1.82
61.78
0.113207
2.764
1.612
MED-EC
PARAM. VALUE
STD.ERROR
t STATISTIC
0.927
3.190
0.291
1.247
0.191
6.517
-0.078
0.720
-0.108
-0.031
0.062
-0.497
21
0.96
1.44
131.18
0.040714
35.192
20.526
EC
PARAM. VALUE
STD.ERROR
t STATISTIC
0.447
8.423
0.053
1.219
0.582
2.093
0.001
1.863
0.000
-0.023
0.126
-0.184
21
0.82
1.78
25.32
0.109253
0.069
0.040
RWE
PARAM.VALUE
STD.ERROR
t STATISTIC
8.066
16.971
0.475
1.100
0.538
2.046
-1.873
3.972
-0.472
0.178
0.235
0.756
21
0.32
1.49
2.72
0.235897
0.013
0.008
ME/NA
PARAM. VALUE
STD.ERROR
t STATISTIC
3.903
7.141
0.546
1.937
0.363
5.342
-0.406
1.544
-0.263
-0.194
0.079
-2.455
21
0.96
1.97
124.96
0.083128
61.439
35.836
RAF
PARAM. VALUE
STD.ERROR
t STATISTIC
3
0.003
0.002
FE
PARAM.VALUE
STD.ERROR
t STATISTIC
15.493
10.569
1.466
1.389
0.276
5.036
-3.520
2.428
-1.450
0.232
0.118
1.971
21
0.82
2.63
26.68
0.086843
0.032
0.018
OCE
PARAM. VALUE
STD.ERROR
t STATISTIC
4
0.002
0.001
Total
100.000
58.326
131

132
The first column of Table 5.1 and 5.2 shows the name of the regions.
The second column explains the meaning of the values appearing under each
variable. The first value corresponds to the parameter, the second to the
standard error, and the third to the "t" statistic. The following columns
display the name of the variables included in the estimation, the
statistics of the estimation, and percentages showing the relative
importance of each region in fresh orange world consumption and trade.
Table 5.1 includes an intercept and four variables, average market
price, real GDP, population level, and substitute product price. Table
5.2 includes an intercept and two variables, FOB average export price, and
total fresh production. The following five columns in both tables show
the number of observations and statistics used to evaluate the general
performance of the model for each equation. The statistics included are
the R square, the Durbin Watson, the F Test, and Theil's inequality
coefficient. The last two columns in Table 5.1 display percentages that
show the importance of each region with respect to total world demand and
total world imports, respectively. The last column in Table 5.2 presents
a percentage that shows the importance of each region with respect to
total world exports of fresh oranges.
Tables 5.3 to 5.13 show the results for the product demand
equations. Each table corresponds to one region with a maximum of ten
estimated equations. The definition of a product for Armington's model in
any of the regions refers to the same type of good but differentiated by
country or region of origin. For example, a fresh orange from the United
States is assumed to be perceived differently by EC consumers than a fresh
orange from the Mediterranean-EC.

133
Given that the model includes a total of 11 regions, there will be
at least ten product demand equations per region. Each product demand
will represent the region's demand for fresh oranges originating in the
other ten regions. Each table represents one final market or importing
region. The names of the partners or regions of origin are shown in the
first column. The second column explains the values appearing under each
variable. The values are the same presented in Tables 5.1 and 5.2; i.e.,
parameter values, standard errors, and "t" statistics. The following
columns display an intercept and the name of the variables included in the
product demand equations for all regions. The variables are relative
prices and total market demand. The following five columns show the
number of observations and the same statistics used for total market
demand and export supply equations. The last two columns show the
relative importance of each partner region's exports to total imports and
total market demand in the final market region, respectively. Estimated
product demand equations are less than ten in some regions due to
insufficient data points.
Tables 5.14 to 5.24 show the results for the CIF price linkage
equations. Each table refers to one region with a maximum of ten
estimated equations. CIF price linkage equations link every region's
import price with the FOB export price for the rest of the regions in the
model. Therefore, there are unique CIF and FOB prices for every trade
flow. Consequently, as in the case for product demands, each region is
associated with the other ten regions through an equation.
The basic structure of Tables 5.14 to 5.24 is the same as the one
presented for Tables 5.3 to 5.13, and hence it will not be repeated here.

134
The differences are the number and type of variables included in the
estimation of each set of equations. The variables for the CIF price
linkage equations are the FOB export price, a yearly trend and an index
price for energy. Given insufficient data points in some regions, the
number of estimated equations is less than ten.
Empirical Results: Graphical Analysis
The following analysis provides a way to evaluate the ability of the
model to predict the original data. Figures for all total market demand
and export supply equations as well as selected figures from relevant
product demand equations are presented and evaluated. The graphical
analysis presented here are complemented with a statistical and economic
analysis of the empirical results in order to have a better criterion to
judge the overall performance of the model. The statistical and economic
analyses are presented in subsequent parts of this section of the chapter.
Figures 5.1 to 5.11 show the actual and fitted values of each
region's total market demand equation. Figures 5.12 to 5.22 present the
actual and fitted values for each one of the export supply equations.
Figures 5.23 to 5.49 show the actual and fitted values of selected product
demand equations. The product demand equations were selected on the basis
of trade relevance to the fresh orange trade model. The figure units
include thousands or millions of metric tons over time. Notice that the
units of measurement vary throughout the different figures. It is
important to take these differences into consideration when evaluating the
model's prediction ability.

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.1. Total Market Demand for Fresh Oranges in the United States.
135

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.2. Total Market Demand for Fresh Oranges in Canada.
136

METRIC TONS (Millions)
YEAR
ACTUAL I FITTED
Figure 5.3. Total Market Demand for Fresh Oranges in Latin America.

METRIC TONS (Thousands)
YEAR
ACTUAL FITTED
Figure 5.4. Total Market Demand for Fresh Oranges in the Mediterranean-EC.
138

METRIC TONS (Thousands)
YEAR
ACTUAL t FITTED
Figure 5.5. Total Market Demand for Fresh Oranges in the EC.
139

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.6. Total Market Demand for Fresh Oranges in the rest of Western Europe.
140

METRIC TONS (Thousands)
YEAR
ACTUAL FITTED
Figure 5.7. Total Market Demand for Fresh Oranges in the Middle East/North Africa.
141

METRIC TONS (Thousands)
YEAR
ACTUAL -+- FITTED
Figure 5.8. Total Market Demand for Fresh Oranges in the rest of Africa.
142

METRIC TONS (Millions)
10
8
0
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
YEAR
ACTUAL
FITTED
Figure 5.9. Total Market Demand for Fresh Oranges in the Far East.

METRIC TONS (Thousands)
Figure 5.10. Total Market Demand for Fresh Oranges in Oceania.
144

METRIC TONS (Thousands)
YEAR
ACTUAL -H FITTED
Figure 5.11. Total Market Demand for Fresh Oranges in the Communist Bloc.
-P'
Ln

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.12. Total Export Supply of Fresh Oranges from the United States.
146

METRIC TONS
YEAR
ACTUAL H- FITTED
Figure 5.13. Total Export Supply of Fresh Oranges from Canada.

METRIC TONS (Thousands)
YEAR
ACTUAL -H FITTED
Figure 5.14. Total Export Supply of Fresh Oranges from Latin America.
148

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.15. Total Export Supply of Fresh Oranges from the Mediterranean-EC.
VO

METRIC TONS (Thousands)
YEAR
ACTUAL -H FITTED
Figure 5.16. Total Export Supply of Fresh Oranges from EC.
150

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.17. Total Export Supply of Fresh Oranges from the rest of Western Europe.
151

METRIC TONS (Thousands)
YEAR
ACTUAL -+- FITTED
Figure 5.18. Total Export Supply of Fresh Oranges from Middle East/North Africa.
152

METRIC TONS (Thousands)
YEAR
ACTUAL -+- FITTED
Figure 5.19. Total Export Supply of Fresh Oranges from the rest of Africa.
153

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.20. Total Export Supply of Fresh Oranges from the Far East.
154

METRIC TONS (Thousands)
YEAR
ACTUAL FITTED
Figure 5.21. Total Export Supply of Fresh Oranges from Oceania.
155

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.22. Total Export Supply of Fresh Oranges from the Communist Bloc.
156

157
Total market demand
Figures 5.1 to 5.11 show that the model captures the trend for the
regions included in the model. The model has the ability to predict if a
region has a growing total market demand or if the demand is decreasing
over time. The results also show that the model predicts most turning
points with few exceptions. The exceptions are Latin America,
Mediterranean-EC, and the rest of Africa. These regions have in common
that almost 100% of their total consumption comes from local production.
The trade model developed here is mainly concerned with trade flows among
regions.
Latin America is the largest producer of oranges and the largest
consumer of fresh oranges within the reporting regions. This region, in
particular Brazil, has developed a fast-growing orange-producing and
processing industry in the last 20 years. These conditions generate a
special effect that might have not been reflected by the model. The model
assumption of independence between fresh and processed consumption could
have been too restrictive for this region. In fact, it is possible to
argue that consumption could have been dependent on how much of the orange
production was processed. This argument, however, will not necessarily
apply to other regions since they either have little Figure 5.1-5.22
orange production or have a slower growing orange-producing and processing
industry.
Mediterranean-EC major turning points are generated by the model up
to 1982. The model failed to capture the changes in demand that occurred
after that year. Figure 5.4 shows that the changes in demand after 1982
are very unusual and probably related to changes that occurred that year

158
and the years after. Those changes culminated in 1986 with the admission
of Spain and Portugal to the EC.
The rest of Africa consumption pattern has been very irregular over
time and the model has been unsuccessful in reflecting the major turning
points. This region is formed by one large producer and exporter (South
Africa) and many countries that usually consume only what they produce.
Imports in this region are very small as compared to total market demand.
Turning points in this region's total market demand are probably related
to exogenous changes in local production of oranges and therefore are not
predicted by the model.
Export supply
Figures 5.12 to 5.22 show that the model generates the trend of the
export supply equations for every region in the model. The model has the
ability to predict if a region has a growing export supply or if the
export supply is decreasing over time. The results also show that the
model does not capture the turning points as well as it did for total
market demand equations with some exceptions. The model does reflect
most of the turning points for the United States, Mediterranean-EC, and
Middle East/North Africa. These regions represented 88% of total world
exports between 1966 and 1986.
Product demand
The third group of figures show the actual and fitted values for
selected product demand equations. Total trade for selected regions
represented over 90% of total world trade in the 21-year period

159
considered. Total imports per region relative to total world imports are
shown in Table 5.1. Imports from each partner relative to total imports
in a given region are shown in Tables 5.3 to 5.13. The figures will be
examined and discussed on a region-by-region basis.
United States. United States total imports represented 1.2% of
total world imports in the period studied. Figures 5.23 and 5.24 show the
demand for Latin America and Middle East/North Africa products in the
United States, respectively. The figures show that, in both cases, the
model predicts the trend as well as some of the major turning points. The
demand for the Latin America product had an unusual peak during 1982 that
was not generated by the model. The demand for the Middle East/North
Africa product has been irregular. However, the model captures the trend
and most turning points.
Canada. Canada imported 4.8% of total world imports in the 21-year
period considered. Figures 5.25 to 5.27 display the demand for the United
States, Middle East/North Africa, and Far East products in Canada,
respectively. The model shows the ability to predict the trend in every
case. The demand for the Far East product is replicated quite well. The
model is not predicting well some turning points for the United States and
Middle East/North Africa products.
Latin America. Latin America imports represented 0.06% of total
world imports for the period considered. Figures 5.28 and 5.29 exhibit
the demand for the United States and EC products in Latin America,
respectively. The estimated product demands reflect the trends over time.

160
The demand for United States product is replicated quite well. The model
is not predicting some turning points in the case of the EC product
demand.
Mediterranean-EC. Mediterranean-EC represented 0.07% of total world
imports from 1966 to 1986. Figures 5.30 and 5.31 show the demand for
Latin America and EC products in the Mediterranean-EC, respectively. The
model generates the trend in both cases. Major turning points for the EC
product demand are also captured by the model. Even though the EC is not
a major producer of oranges, it does have some production and trade with
other regions of the world. Some Latin America product demand's turning
points are not reflected by the model. However, trade between Latin
America and the Mediterranean-EC was negligible until 1980. This could be
a partial explanation for the failure of the model in replicating the data
in this particular case.
EC. EC represented 63.4% of total world imports in the period
considered. Figures 5.32 to 5.34 show the demand for Mediterranean-EC,
Middle East/North Africa, and the rest of Africa products in the EC,
respectively. The model predicts the trend in every case. Figures 5.32
indicate that it also generates most turning points for the case of the
Mediterranean-EC. The demand for the Mediterranean-EC product in the EC
represented 35% of total world trade and 55% of EC's total imports.
Figures 5.33 and 5.34 show a good general fit, but some turning points are
not captured by the model.
Rest of Western Europe. Rest of Western Europe imports represented
10.6% of total world imports in the 21-year period considered. Figures
5.35 to 5.37 present the demand for Mediterranean-EC, Middle East/North

161
Africa, and the rest of Africa products in the rest of Western Europe
respectively. The three product demands show that the model reflects the
trend. The best fit is obtained by the demand for Middle East/North
Africa product for which turning points are predicted by the model. The
demand for Mediterranean-EC product shows that the model generates only
some turning points, and for the rest of Africa product shows that just a
few turning points are captured.
Middle East/North Africa. Middle East/North Africa imports
represented 1.65% of total world imports from 1966 to 1986. The region
has been growing rapidly in terms of total market demand and trade in the
last 15 years. Figures 5.38 to 5.40 display the demand for Latin America,
rest of Africa, and Far East products in the Middle East/North Africa
respectively. The model reflects the trend of the product demands in
every case, but it is not predicting some turning points in each equation.
Rest of Africa. Rest of Africa imports represented 0.16% of total
world imports in the period considered. Figures 5.41 and 5.42 exhibit the
demand for EC and Middle East/North Africa products in the rest of Africa,
respectively. The model generates the trend in both cases. The figures
indicate that several turning points in each product demand are not
captured by the model.
Far East. Far East imports represented 4.3% of total world imports
in the 21-year period studied. This market has been growing fast in the
last two decades. Figures 5.43 to 5.45 show the demand for the United
States, Middle East/North Africa, and Oceania products in the Far East,
respectively. Figure 5.43 shows that the model closely reflects the
demand for the United States product in the Far East. Figures 5.44 and

162
5.45 show that the model predicts the trend for the demands of Middle
East/North Africa and Oceania products in the Far East. However, in the
last two cases the model is not generating some of the turning points in
each equation.
Oceania. Oceania represented .21 of total world imports in the
period studied. Figures 5.46 and 5.47 display the demand for the United
States and Middle East/North Africa products in Oceania, respectively.
Figure 5.46 indicates that the model is capturing the trend and major
turning points for the United States quite well. On the other hand,
Figure 5.47 shows an irregular trade that went to zero in 1983. In this
case, the model reflects the general trend but fails to predict some of
the turning points.
Communist Bloc. Communist Bloc imports represented 13.6% of total
world imports between 1966 and 1986. Figures 5.48 and 5.49 present the
demand for Mediterranean-EC and Middle East/North Africa products in the
Communist Bloc. Both figures show that the model has the ability to
predict the trend. Figure 5.48 indicates that some turning points for the
demand of the Mediterranean-EC product are not generated by the model.
However, Figure 5.49 indicates that most turning points for the product
coming from the Middle East/North Africa are captured by the model.
Conclusion: graphical analysis
The graphical analysis presented indicates that the model has the
ability to predict the trends for nearly all equations. Some turning
points were not, however, captured by the model. Over 60% of the

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.23. United States Imports of Fresh Oranges from Latin America (Product Demand 13).
163

METRIC TONS (Thousands)
YEAR
ACTUAL -H FITTED
Figure 5.24. United States Imports of Fresh Oranges from the Middle East/North Africa (Product Demand 1_7).
164

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.25. Canada Imports of Fresh Oranges from the United States (Product Demand 21).
165

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.26. Canada Imports of Fresh Oranges from the Middle East/North Africa (Product Demand 27).
166

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.27. Canada Imports of Fresh Oranges from the Far East (Product Demand 29).
o>

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.28. Latin America Imports of Fresh Oranges from the United States (Product Demand 31).
168

METRIC TONS
YEAR
ACTUAL I FITTED
Figure 5.29. Latin America Imports of Fresh Oranges from the EC (Product Demand 35).
169

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.30. Mediterranean-EC Imports of Fresh Oranges from Latin America (Product Demand 43).
170

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.31. Mediterranean-EC Imports of Fresh Oranges from the EC (Product Demand 45).
171

METRIC TONS (Thousands)
YEAR
ACTUAL 1 FITTED
Figure 5.32. EC Imports of Fresh Oranges from the Mediterranean-EC (Product Demand 54).
172

METRIC TONS (Thousands)
YEAR
- ACTUAL I FITTED
Figure 5.33. EC Imports of Fresh Oranges from the Middle East/North Africa (Product Demand 57).
173

METRIC TONS (Thousands)
YEAR
ACTUAL FITTED
Figure 5.34. EC Imports of Fresh Oranges from the rest of Africa (Product Demand 58).
174

METRIC TONS (Thousands)
YEAR
ACTUAL -+ FITTED
Figure 5.35. Rest of Western Europe Imports of Fresh Oranges from Mediterranean-EC (Product Demand 64).
-vj
Ln

METRIC TONS (Thousands)
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
YEAR
ACTUAL -+- FITTED
Figure 5.36. Rest of Western Europe Imports of Fresh Oranges from the Middle East/North Africa
(Product Demand 6_7).
176

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.37. Rest of Western Europe Imports of Fresh Oranges from the rest of Africa (Product Demand 6_8).
177

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.38. Middle East/North Africa Imports of Fresh Oranges from Latin America (Product Demand 73).
178

METRIC TONS (Thousands)
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
YEAR
ACTUAL 1 FITTED
Figure 5.39. Middle East/North Africa Imports of Fresh Oranges from the rest of Africa (Product Demand 78).
179

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.40. Middle East/North Africa Imports of Fresh Oranges from the Far East (Product Demand 79).
180

METRIC TONS
YEAR
ACTUAL FITTED
Figure 5.41. Rest of Africa Imports of Fresh Oranges from the EC (Product Demand 85).
181

METRIC TONS (Thousands)
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
YEAR
ACTUAL I FITTED
Figure 5.42. Rest of Africa Imports of Fresh Oranges from the Middle East/North Africa (Product Demand 87).
182

METRIC TONS (Thousands)
YEAR
ACTUAL FITTED
Figure 5.43. Far East Imports of Fresh Oranges from the United States (Product Demand 91).
183

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.44. Far East Imports of Fresh Oranges from the Middle East/North Africa (Product Demand 97).
184

METRIC TONS (Thousands)
YEAR
ACTUAL t FITTED
Figure 5.45. Far East Imports of Fresh Oranges from Oceania (Product Demand 910).
185

METRIC TONS (Thousands)
YEAR
ACTUAL I FITTED
Figure 5.46. Oceania Imports of Fresh Oranges from the United States (Product Demand 101).
186

METRIC TONS
YEAR
ACTUAL FITTED
Figure 5.47. Oceania Imports from the Middle East/North Africa (Product Demand 107).
187

Figure
METRIC TONS (Thousands)
YEAR
ACTUAL + FITTED
5.48. Communist Bloc Imports of Fresh Oranges from the Mediterranean-EC (Product Demand 114).
188

METRIC TONS (Thousands)
YEAR
ACTUAL H- FITTED
Figure 5.49. Communist Bloc Imports of Fresh Oranges from the Middle East/North Africa (Product Demand 117). $

190
equations predicted most turning points. A quantitative statistic that
measures the model's ability to predict or simulate major turning points
will be introduced in the next section. The statistic will provide an
additional criterion to judge the model performance for predictions.
Model predictions were better for total market demand equations than
for export supply and product demand equations. Figures 5.1 to 5.11 show
that the model reflects the trends and most turning points for the total
market demand equations. The best predictions in the supply side
corresponded to three regions whose exports combined represented 88% of
total world exports from 1966 to 1986. In addition the main problems
found in product demands were not related to significant trade flows, but
were more related to regions with minimal trade flows.
Empirical Results: Statistical Analysis
In this section, the results will be analyzed using the statistics
obtained from the estimation of the model. The criteria to evaluate an
equation in multi-equation models are similar to the criteria used to
evaluate single-equation regression models, even if a multi-equation
estimation procedure was used. A nonlinear two-stage least squares was
used to estimate the system of equations and each equation has an
associated set of statistics.
Five statistics were selected to be included in the discussion over
the performance of the fresh orange trade model. The statistics selected
are the R square (@RSQ), the Durbin Watson (@DW), the F test (@FST), the
Theil's Inequality Coefficient (UTHEIL), and the "t" statistic. The

191
following discussion will cover four of the five statistics. The "t"
statistic will be evaluated in the fourth part of this section of the
chapter together with the economic analysis of the parameter signs and
magnitudes.
The @DW statistic is generally used to determined the existence and
type of serial correlation in time series. If data are given on a yearly
basis, as is the case for the fresh orange trade model, evidence of serial
correlation would probably be related to model misspecification. If the
@DW is close to two, then there is no evidence of misspecification.
A criterion that is used to evaluate a simulation model is the fit
of the individual variables in a simulation context. It is expected that
the results of a historical simulation match the behavior of the real
world rather closely. It is therefore interesting to perform a historical
simulation and examine how closely each endogenous variable tracks the
historical data. This is especially important when the model is
nonlinear, given the weakness of the @RSQ and @FST in those cases.
Theil's Inequality Coefficient (UTHEIL) is a useful simulation statistic
related to the RMS (Root-Mean-Square) simulation error and applied to the
evaluation of historical simulations or ex post forecasts. The UTHEIL
will give an idea on how well the model captures the turning points of the
estimated equations. If the value of UTHEIL is zero, then the predicted
value is equal to the actual value and there is a perfect fit. If UTHEIL
is equal to one, then the predicted performance of the model is no better
than a random estimate.
Tables 5.1 to 5.24 present the empirical results of the estimated
fresh orange trade model and includes the major statistics discussed

192
above. Total market demand and export supply statistics are also
highlighted in Table 5.25 to facilitate more detailed discussions. Tables
H.l to H.4 in Appendix H present the statistics for the product demand and
CIF price linkage equations region by region.
Total market demand
Table 5.25 presents the statistics for total market demand and
export supply equations. Two sets of six columns are included in Table
5.25. The first column describes the region name, the second the number
of observations, the third the @RSQ, the fourth the @DW, the fifth the
@FST, and the sixth the UTHEIL coefficient. The number of equations
considered in each table is 11, one per region.
The number of observations used to estimate total market demand
equations is 21 in all cases. There were sufficient data points for every
variable in the whole range covered by the model. The period considered
for the estimation was from 1966 to 1986.
The UTHEIL in all cases is far below .5 for total market demand
equations. In general the model is reflecting the major turning points of
the historical data. This is definitely an important result given the
nonlinear nature of the model.
Export supply
Table 5.25 shows the export supply equations statistics. As
mentioned above, a total of 11 equations are reported. Each region has
one export supply equation associated with it.

Table 5.25 Total Market Demand and Export Supply Equations Statistics
Region
Total
Market
Demand
Total
Export Supply
@0BS
@RSQ
@DW
@FST
UTHEIL
@OBS
@RSQ
@DW
@FST
UTHEIL
US
21
0.77
2.52
13.16
0.028457
21
0.20
1.71
2.19
0.106359
CAN
21
0.23
1.94
1.22
0.040323
21
0.28
2.68
3.49
0.636455
LA
21
0.78
1.83
14.22
0.042903
21
0.17
1.45
1.89
0.135310
MED-EC
21
0.37
3.27
2.38
0.085051
21
0.31
2.49
4.02
0.079362
EC
21
0.49
2.52
3.90
0.031023
21
0.82
1.41
41.36
0.172106
RWE
21
0.40
1.73
2.67
0.033662
21
0.49
1.24
8.75
0.355627
ME/NA
21
0.95
1.98
80.91
0.031826
21
0.43
1.11
6.72
0.050221
RAF
21
0.27
1.21
1.49
0.061481
21
0.06
0.55
0.53
0.077483
FE
21
0.95
2.40
79.89
0.023174
21
0.67
1.30
17.98
0.196981
OCE
21
0.35
1.49
2.20
0.042669
21
0.33
0.78
4.51
0.165562
COMMB
21
0.86
1.42
23.60
0.054849
21
0.85
1.52
51.55
0.165290
193

194
The UTHEIL indicates that all export supply equations have a
coefficient far below .5 except for Canada. Overall, the model is
predicting the major turning points of the historical data. Canada is a
net importer and has no production of oranges; therefore, the results will
not have an important impact on the fresh orange trade model.
Product demands and CIF price linkage equations
Tables H.l to H.4 in Appendix H present the statistics for the
product demand and CIF price linkage equations. Detailed region-by-region
discussions about those statistics will not be included here.
Conclusion: statistical analysis
The statistical analysis shows that the model is capturing the major
variations of the different dependent variables for each total market
demand equation. Problems found were usually related to regions that will
not affect the major driving issues of the fresh orange trade model.
The analysis also shows that the model is reflecting the major
variations of the different dependent variables for total market demands
better than it does for export supply equations. However, the results
show that export supply equations for major world exporters are well
captured by the model.
The reported statistics show that the model is predicting the major
variations of the product demands better for which relevant trade took
place. Important trade flows like Canadian imports from the United
States, EC imports from the Mediterranean-EC and the Middle East/North

195
Africa, rest of Western Europe imports from the Middle East/North Africa
and Mediterranean-EC, Far East imports from the United States, and
Communist Bloc imports from the Middle East/North Africa and
Mediterranean-EC; seem to be captured by the model. The CIF price linkage
equations show better results than the ones obtained by the product
demands. The model did not represent the rest of Africa data very well.
It is important to notice that, in almost every case, the UTHEIL
coefficient was in acceptable ranges, indicating that major turning points
in the data were captured by the model.
Empirical Results: Economic Analysis
In this section theoretical economic expectations and implications
about signs and magnitudes and the "t" statistics associated with
estimated coefficients are considered.
The probability distribution of estimators for small sample sizes in
a system of simultaneous equations is unknown except for a few highly
special cases (Judge et al., 1985). Therefore, the procedure utilized to
estimate the fresh orange trade model implies that the estimated
parameters are consistent but biased. Therefore, the "t" statistics
obtained using NL2SLS can be used only to give some idea about the
accuracy of the estimated parameters. Gujarati (1988) suggests that a "t"
value in absolute terms greater than one would imply that the parameter is
probably significant in a model such as the one used in this study. This
guideline will be used in the following discussion.

196
In order to facilitate the presentation and discussion of the
results, a new set of tables will be introduced. Table 5.26 shows the
elasticities obtained from the total market demand and the export supply
equations. Tables 5.27 and 5.28 present the relative price and total
market demand elasticities obtained from the product demands. Tables
5.29, 5.30, and 5.31 present the FOB export price, the year trend, and the
index price for energy elasticities obtained from the CIF price linkage
equations. The new tables will be analyzed separately. More important
regions will be emphasized during the discussion.
Total market demand
Table 5.26 presents two sets of results, one for total market
demands and one for export supply equations. Total market demand measures
the demand of a single region for fresh oranges. The variables considered
in the estimation of this section of the model are the average market
price, real GDP, population, and substitute product price. The results
are presented in a matrix where the columns represent the regions and the
rows the different variables included in the model.
The economic expectations about the sign and magnitude of the
different elasticities vary depending on the variable analyzed. Based on
economic theory, the sign for the average market price elasticities is
expected to be negative; i.e., as the average market price for fresh
oranges goes up, it is expected that their consumption decreases. On the
other hand, the signs for income, population, and substitute product price
elasticities are expected to be positive. Consumption of fresh oranges is
expected to increase as disposable income and the number of consumers in

197
Table 5.26 Market Demand and Export Supply Equations Elasticities.
REGION
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
COMMB
"MARKET
DEMANDS"
"MARKET
PRICE ELASTICITIES"
PVa
-2.182
-0.569
-0.057
-0.781
-1.182
-1.028
-1.138
-0.250
0.181
-1.655
-1.216
TSb
-7.030
-1.256
-0.151
-1.039
-3.255
-1.833
-1.483
-0.276
2.099
-2.800
-2.197
"INCOME
ELASTICITIES"
PV
-0.549
0.693
0.091
-0.641
0.674
-0.010
0.411
-0.092
0.151
-0.335
1.176
TS
-1.352
1.798
0.407
-0.798
1.239
-0.036
2.626
-0.318
0.749
-0.827
2.127
"POPULATION ELASTICITIES"
PV
2.119
-1.792
1.044
7.179
-9.114
-0.717
-0.554
0.563
1.446
-0.427
0.505
TS
2.749
-1.449
2.116
1.452
-1.726
-0.349
-0.692
1.205
2.550
-0.447
0.137
"SUBSTITUTE PRODUCT PRICE
ELASTICITIES"
PV
0.640
0.038
0.134
-0.200
0.259
0.124
-0.476
0.097
-0.165
0.182
-0.523
TS
3.436
0.117
0.539
-0.421
0.856
0.432
-1.643
0.338
-2.252
1.061
-1.046
"EXPORT
SUPPLIES"
"FOB EXPORT PRICE ELASTICITIES"
PV
-1.679
2.245
-0.155
-0.009
-0.275
-2.650
1.424
-0.191
-1.244
-2.261
-1.792
TS
-1.010
2.643
-0.364
-0.031
-0.187
-4.182
3.317
-0.269
-3.602
-2.402
-1.258
"FRESH PRODUCTION ELASTICITIES"
PV
0.914
1.000
0.639
0.646
0.692
1.000
1.187
0.357
0.154
0.303
2.144
TS
1.862
0.000
1.879
2.755
2.389
0.000
2.704
1.002
0.259
0.329
9.738
aParameter values.
bt Statistic.
cCanada and rest of Western Europe are not orange producers.

198
the region increase. The consumption of fresh oranges is also expected to
increase as the prices of substitute products like bananas and apples go
up.
Price elasticities are negative except for the total market demand
in the Far East. The Far East is a fast-growing market (see Figure 5.9)
with a lot of interest in high-quality fruit. This is especially true for
Japan. Most fruit consumed in these markets come from local production.
However, 80% of their imports are high-grade fruit from the United States.
These conditions of the market and the characteristics of the consumers in
this part of the world may partially explain why the price elasticity
could be positive.
The absolute value of the "t" statistics for the price elasticities
are greater than one except for Latin America and rest of Africa. Based
on the characteristics of production, exports, and consumption in these
two regions, consumption is probably marginal and highly driven by how
much fruit is processed and/or exported. The price level under these
circumstances is probably not very important since marginal fruit has to
be consumed anyway. Market price elasticities are elastic (greater than
one in absolute value) except for Canada and Mediterranean-EC. This
indicates that the demand for fresh oranges in most regions is highly
responsive to changes in the market price. Even though Canada and
Mediterranean-EC have elasticities lower than one, they are between .57
and .78 which are relatively high numbers.
Income elasticities are reported in the second row. Six out of 11
regions have the correct positive sign. Positive income elasticities for
Canada, EC, Middle East/North Africa, and Communist Bloc are significant.

199
The Far East and Latin America have positive income elasticities, but the
"t" statistics indicate that they are not significant.
The rest of the regions have negative income elasticities, which is
an unexpected result. The result could be related to trending population
and income which implies high potential correlation between the two
variables. However, only one of the negative elasticities is
significantly different from zero in terms of the "t" statistic. That is
the case for the United States. The "t" statistic for the income
elasticity is far below the other "t" statistics obtained for the rest of
the variables included in the total market demand equation for the United
States. This condition indicates that the income elasticity is less
significant than the rest of the parameters in the model. Income
elasticities for Latin America and the rest of Africa are not significant.
Demand in these regions is marginal, and apparently income does not seem
to be a major demand driver.
Positive and significant income elasticities are less than one with
the exception of the Communist Bloc. That is, income elasticities are
inelastic (smaller than one in absolute value) in most cases. Communist
Bloc income elasticity is slightly greater than one. Given the usual
market controls of centrally planned economies, a slightly elastic income
elasticity may be expected. Consumers in this region will buy fresh
fruits and, in particular, fresh oranges in larger proportions than
changes in their income whenever they have the opportunity. Also,
reported import value is less likely to be reflected in consumer prices
because of price controls.

200
Population elasticities are reported in the third row. Six out of
11 regions have the correct positive elasticities. The United States,
Latin America, Mediterranean-EC, rest of Africa, and the Far East have
positive and significant population elasticities. The Communist Bloc has
a positive population elasticity, but it is not significant. The
discussion maintained above regarding Latin America and rest of Africa is
again confirmed with the results obtained. Population, and not price and
income, is the major driver of the marginal demands in these regions.
Price and income elasticities are insignificant for these regions.
The rest of the regions have negative population elasticities but
only the ones from Canada and the EC are apparently significantly
different from zero. This is an unexpected result, which implies that as
population increases, consumption of fresh oranges decreases. Unexpected
signs may be resulting from data or specification errors that are more
likely to occur in large models as used in this study.
The magnitudes of positive population elasticities range from .56
for the rest of Africa to 7.2 for Mediterranean-EC. The rest of them are
between one and 2.12. The results indicate highly elastic population
elasticities in most cases.
Substitute product price elasticities are shown in row four. Seven
out of the 11 regions have the correct positive elasticities, but only
United States and Oceania are significantly different from zero indicating
inelastic substitute product price elasticities. Consumption of fresh
oranges increases less than proportionally to increases in the price index

201
used for bananas and apples. A recent study of the U.S. apple industry
also failed to find any substitution between apples and oranges (Ward,
1991).
Four regions have negative substitute product price elasticities.
Three of them have significant parameters. These regions are Middle
East/North Africa, Far East, and Communist Bloc. A negative substitute
product price indicates that, as the price index for bananas and apples go
down (up), the consumption for fresh oranges go up (down). This is an
unexpected result, which implies complementarity instead of
substitutability. In can be argued that, for Middle East/North Africa and
Communist Bloc, bananas and apples are not the best substitute products
for fresh oranges. These substitutes were selected from models in the
literature that were mainly applied to developed markets. In the Far East
the substitution may exist, but given the market characteristics and
consumption patterns mentioned above, the results are not as expected.
The rest of the regions with positive or negative substitute product
price elasticities are not significant. This suggests that the price
index for bananas and apples has little effect on the demand for fresh
oranges in those regions.
Export supply
Export supply equations are also presented in Table 5.26. The
columns show the regions. The rows show the two variables included in the
estimation. The variables included in the export supply equations are FOB
export price and fresh production. The first part of the following

202
discussion will cover general issues about the results; subsequently, the
major exporting regions will be addressed separately.
Economic expectations about the sign and magnitude of the different
elasticities vary depending on the variable analyzed. Based on economic
theory, the sign for the FOB export price elasticities is expected to be
positive. If the FOB export price for fresh oranges in a given region
increases, it is expected that exports from that region increase. Fresh
production elasticities are expected to be positive. If fresh production
goes up, it is expected that exports go up.
Two positive FOB export price elasticities were obtained, one for
the Middle East/North Africa, which is a major net exporter, and one for
Canada, which is a net importer. The rest of the regions have negative
FOB export price elasticities. The elasticity obtained for the Middle
East/North Africa is 1.42. This indicates that FOB export price for this
region is highly elastic. A change in the FOB export price will generate
a larger-than-proportional change in fresh export supply.
Out of the nine regions with negative elasticities, there are five
with significant parameters, three of them with strong ones. However, the
regions with strong negative signs are not major exporters. Rest of
Western Europe and Far East are net importers. Oceania exports
represented only .5% of total world exports. The results indicate that
the FOB export price is not a major factor for world fresh supply.
Fresh production is the major driver of exports. Two regions,
Canada and rest of Western Europe, have zero parameters indicating that
these regions have zero local fresh production. The rest of the regions
have correct positive fresh production elasticities. Seven out of nine

203
have significant parameters. The two regions with insignificant
parameters are the Far East, which is a net importer, and Oceania, which
is a minor exporter.
The results show that the export supply behavior for major world
exporters is good. Mediterranean-EC, with exports accounting for 44% of
total world exports, has a well-behaved export supply equation with
insignificant FOB export price elasticity but a strong positive fresh
production elasticity. The Middle East/North Africa, accounting for 35%
of total world exports, has a well-behaved export supply equation with
strong positive FOB export price and fresh production elasticities. The
United States and Latin America, accounting for 12.8% of total world
exports, have well-behaved export supply equations. Both regions have
strong positive fresh production elasticities. The United States has
negative but weak FOB export price elasticity, and Latin America's price
elasticity is insignificant.
Product demand
In previous sections, most of the equations were analyzed
graphically and in terms of fit, performance, and simulation ability. In
order to avoid unnecessary repetition, the emphasis of the following
discussion will be on important trade flows. The fresh orange trade model
developed here is basically interested in understanding the demand factors
that make regions shift their imports from one source to another.
Decisions about relevance have to be made first by selecting major world
importers and then by identifying their major suppliers. Relevant regions

204
Table 5.27 Product Demands Relative Price Elasticities3
Region ib
REGION jb
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
COM
US
-0.567c
-0.474d
1.923
1.759
-4.896
-3.072
-8.876
-2.727
-2.293
-1.715
-1.816
-1.835
0.570
2.887
-1.235
-1.558
-9.518
-2.642
CAN
1.436
1.285
-2.161
-1.833
-1.616
-0.572
-2.940
-1.958
LA
1.051
2.840
-0.470
-0.288
1.777
0.397
0.199
0.285
-0.653
-1.274
1.807
0.711
-3.669
-5.833
-3.517
-3.700
3.978
1.368
-3.402
-3.208
MED-EC
-0.857
-0.582
3.087
2.323
-2.047
-0.783
0.755
0.634
-2.700
-2.544
0.202
0.175
-0.367
-0.365
0.438
0.538
5.089
2.479
0.143
0.273
EC
-1.214
-3.334
-1.564
-1.775
-0.106
-0.077
-4.717
-4.625
-2.654
-7.974
-2.826
-2.994
0.166
0.349
-3.013
-2.169
-2.190
-1.430
RWE
-0.771
-1.193
-2.261
-2.085
-0.857
-0.591
1.481
1.950
-0.333
-0.260
ME/NA
-5.564
-3.464
-0.185
-0.266
-2.980
-1.093
-0.869
-1.022
1.868
2.387
-2.666
-5.289
0.654
1.799
-0.782
-2.813
-6.590
-5.710
-1.549
-4.175
RAF
0.366
0.474
-0.712
-1.326
0.622
1.212
-1.071
-0.290
-2.827
-1.677
FE
7.510
4.261
-0.924
-2.571
4.105
1.631
-0.754
-0.413
-0.670
-1.431
-2.498
-4.371
-1.408
-0.317
8.172
1.432
OCE
-1.512
-2.082
-2.546
-2.014
5.085
1.838
1.792
1.518
-1.459
-1.895
-5.716
-4.790
-0.125
-0.240
COM
1.673
0.370
-9.158
-3.636
5.388
3.243
-3.702
-2.616
-0.488
-0.414
aProduct demand equals X^j and Relative price equals P^j/P^.
^Region i across the top is the region importing from region j down the column.
cThe first line in each region represents parameter values.
dThe second line in each region represents the t Statistic.

205
Table 5.28 Product Demands Total Market Demand Elasticities3
Region ib
REGION jb
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
COMMB
US
1.372c
1.223d
-1.060
-1.490
-3.938
-1.234
-10.850
-2.645
1.551
0.816
-3.544
-1.267
2.666
10.485
2.663
1.782
0.484
0.200
CAN
-5.097
-0.868
0.634
0.230
-3.856
-0.496
3.662
1.182
LA
3.298
1.184
-18.571
-2.496
8.218
1.062
2.625
1.972
1.200
1.091
6.749
1.764
0.185
0.067
3.155
1.564
-4.127
-1.869
3.375
3.606
MED-EC
27.351
2.574
-23.719
-2.269
5.098
2.258
1.846
2.976
1.612
2.078
3.655
2.448
6.938
3.019
2.012
2.415
5.972
1.025
0.280
1.371
EC
0.815
0.171
13.259
1.532
1.490
1.067
11.146
3.349
2.660
2.454
2.576
2.072
1.081
1.667
4.084
2.674
0.925
1.674
RWE
12.805
5.388
-5.828
-2.101
3.596
1.466
-2.612
-2.198
4.159
2.691
ME/NA
11.692
1.613
-4.531
-1.215
-2.791
-0.646
2.365
0.788
0.144
0.247
0.212
0.486
0.251
0.490
-0.026
-0.058
-15.478
-3.424
1.237
4.177
RAF
-5.021
-1.694
-0.042
-0.068
0.782
1.582
5.605
4.036
2.688
0.935
FE
0.578
0.577
3.187
4.017
23.361
1.795
-4.150
-0.417
3.294
3.998
-6.961
-2.426
2.104
0.312
7.847
1.611
OCE
30.060
3.553
14.513
3.746
10.671
2.420
7.456
2.262
4.743
6.678
-1.179
-0.590
1.129
1.690
COMffl
13.613
2.228
12.889
2.191
-1.672
-0.640
8.263
1.537
6.273
1.781
aProduct demand equals Xj and total market demand equals X.
^Region i across the top is the region importing from region j down the column.
cThe first line in each region represents parameter values.
dThe second line in each region represents the t Statistic.

206
will be addressed separately and both relative price and total market
demand variables will be analyzed in each case. Latin America,
Mediterranean-EC, rest of Africa, and Oceania are net exporters; their
imports represented only .5% of total world imports during the 21-year
period considered. These regions will be analyzed briefly following the
discussion about leading importers.
Tables 5.27 and 5.28 show the estimated parameters and their
associated "t" statistics for the product demands. Product demands
measure the demand in a given region for fresh oranges coming from another
region. There will be one product demand in each region for each one of
the partner regions. Since the model has a total of 11 regions, the
number of estimated product demands should be 110 (11 regions with ten
partners each).
The variables considered in the estimation of this section of the
model are relative price and total market demand. The relative price
variable refers to the price of the imported product (for example, the
price of fresh oranges from Latin America in the EC) relative to the
average market price (in the EC). If the relative price variable
increases, imported product price is going up faster than the average
market price. In that case, demand for that product should decrease in
the final market. This example implies that the exporting region will be
losing part of its market share in the final market. Therefore, the
expected sign for relative price elasticities is negative.
The other variable included in the model is total market demand.
This variable measures apparent consumption or total size of the market
for fresh oranges in a given region. If market size increases

207
(decreases), it is possible that a given product demand could increase or
decrease. The resultant sign will depend on consumer preference about
where to buy their product when the size of the market is increasing or
decreasing. For example, if total market size is growing, the demand
increase could be satisfied by increasing local product consumption or by
shifting between any of the ten supplying regions. Therefore, the
expected sign could be positive or negative.
Tables 5.27 and 5.28 have the same structure. The results are
presented in a matrix where both columns and rows represent regions. The
columns represent the final market region or importer and the rows the
partner regions or exporters. For example, the second column represents
Canadian imports. The first entry (second column first row) is the result
for the relative price elasticity of the Canadian demand for United States
fresh oranges (product). Some of the product demand equations were not
estimated, given insufficient trade between some regions. Therefore, some
table cells are empty.
Table 5.27 present a total of 82 estimated relative price
elasticities. Fifty five of them show the correct negative sign. The
remaining 27 are positive, but ten of them are not significant. Table
5.28 shows the results for the total market demand elasticities. Signs
are mixed as expected, and 62 of the 82 estimates are significantly
different from zero.
United States. United States imports represented 1.2% of total
world imports during the 21-year period studied. Major imports to the
United States came from Latin America, Middle East/North Africa, and
Mediterranean-EC. The parameter for Latin America is positive and

208
significant. Given the results in terms of fit and performance discussed
above, this is an unexpected result. Given the large production capacity
of the United States and that imports from Latin America represent only
2.6% of U.S. total market demand, possibly imports are occurring only when
domestic production is insufficient to supply the fresh market. The
results show that the parameters are negative for the Middle East/North
Africa and Mediterranean-EC. However, the parameter is not significant
for the Mediterranean-EC. The parameter for the Middle East/North Africa
product is -5.56, which indicates a highly elastic relative price
elasticity. A small change in relative prices will have an important
effect in the demand for the product.
Total market demand elasticities for United States' three major
partners are positive, highly elastic, and significant. Elasticities are
3.3 for Latin America, 11.7 for Middle East/North Africa, and 27.3 for
Mediterranean-EC. Results indicate that the demand for the product of
these regions in the United States is very sensitive to changes in the
size of the market. Therefore, a small increase in the United States
total market demand results in a large increase in the demand for the
different products. Given the differences in elasticity magnitudes, any
change in the United States market size implies a different change for
each product demand. Mediterranean-EC product demand will change faster
than demands for Latin America and Middle East/North Africa products in
the United States.
Canada. Canada imports represented 4.8% of total world imports in
the period studied. Relevant Canadian partners are the United States, Far
East, and Middle East/North Africa. Relative price elasticities for these

209
regions are negative. However, only one is highly significant. This is
the case for the Far East product. The magnitudes in the three cases
indicates that relative price elasticities for this region are inelastic.
An increase in the relative import price implies a less than proportional
decrease in demand for the product.
To understand some of the implication of the empirical results, an
example will be developed. Suppose that there is no local production and
that only two suppliers exist for Canada: the Far East (-.92) and Oceania
(-2.55) (see Table 5.27). The elasticities obtained imply that an equal
change in the relative price variables will have different effects in each
product demand. A similar increase in relative prices will cause a shift
from consuming Oceania product to consuming relatively more product from
the Far East.
Total market demand elasticities for Canada's three major partners
are elastic and significant. The elasticity for the Middle East/North
Africa product is negative, while the ones for the United States and the
Far East are positive. The magnitudes of these elasticities indicate that
the demand for these products in Canada is highly sensitive to small
changes in the size of the market. The direction of change for the Middle
East/North Africa is different from the other regions. For example, if
Canadian market size grows consumers will shift from Middle East/North
Africa product to the United States or Far East products.
EC. EC imports represented 63.4% of total world imports in the 21-
year period studied. Major EC partners are Mediterranean-EC, Middle
East/North Africa, rest of Africa, Latin America, and United States.
Product demands for the United States and rest of Africa have significant

210
negative elasticities. The magnitudes for the rest of Africa and the
United States indicate an inelastic relative price elasticity for the rest
of Africa and an elastic one for the United States in the EC.
The other three regions have positive elasticities but only the one
from the Middle East/North Africa is significant. EC's major partner has
been Mediterranean-EC. This partnership has been growing fast and trade
has been shifting from the Middle East/North Africa to the Mediterranean-
EC through the years. It is possible to argue that EC imports from the
Middle East/North Africa are marginal in the sense that they are needed
only to complement fruit purchases from Mediterranean-EC. This suggests
that the fruit is imported when prices are going up due to the lack of
sufficient fruit in the market. This conclusion could partially explain
the positive sign, but is truly a conjecture not based on actual data.
Total market demand elasticities for three of EC's five major
partners are elastic and significant. United States product elasticity is
-3.94; Latin America is 2.63; and Mediterranean-EC is 1.85. Middle
East/North Africa and rest of Africa have insignificant parameters. The
magnitudes of significant elasticities indicate that demand for the
product of the United States, Latin America, and Mediterranean-EC in the
EC is very sensitive to changes in the size of the market. However, as in
the Canadian case, the direction and magnitude of change are different for
each partner region. For example, if EC's market size grows, consumers
will shift from United States product to the Mediterranean-EC or Latin
America products.
Rest of Western Europe. Rest of Western Europe imports represented
10.6% of total world imports in the period studied. Major partners are

211
Middle East/North Africa, Mediterranean-EC, and rest of Africa. The
results indicate that two of the three relative price elasticities are
negative and significant. Rest of Africa shows a positive elasticity, but
the "t" statistic is low and therefore not significant. Middle East/North
Africa and Mediterranean-EC elasticities are elastic. Product demands are
highly sensitive to changes in relative prices. The same is true for
relatively smaller partners such as Latin America and the United States.
Total market demand elasticities for rest of Western Europe's three
major partners are positive, and two of them are significant. The
magnitudes show an elastic total market demand elasticity for
Mediterranean-EC and an inelastic one for the rest of Africa. Results
imply that, if market size grows the demand for Mediterranean-EC product
will grow in a higher proportion than the demand for rest of Africa
product. Middle East/North Africa total market demand elasticity is not
significantly different from zero.
Middle East/North Africa. Middle East/North Africa is a net
exporting region; however, imports have been growing fast lately and
represented 1.65% of total world imports during the 21-year period. Major
partners are Far East, rest of Africa, Latin America, Mediterranean-EC,
and Oceania. Three of the five product demands have negative relative
price elasticities, two of which are significant. Negative and
significant elasticities were obtained by product demands from the Far
East and Oceania. The rest of Africa has a negative but insignificant
elasticity. Latin America and Mediterranean-EC have positive
elasticities, but they are also not significant. The results show that
the elasticity for the Far East is inelastic, while that for Oceania is

212
elastic. Other smaller partners have product demand elasticities which
are negative, highly elastic, and significantly different from zero in
this region.
Total market demand elasticities for Middle East/North Africa's five
major partners are positive, highly elastic, and significant.
Elasticities are 6.75 for Latin America, 3.65 for Mediterranean EC, 5.61
for rest of Africa, 3.29 for Far East, and 4.74 for Oceania. The
magnitudes indicate that product demands are highly sensitive to change in
the size of the market. Figure 5.7 implies a fast-growing tendency for
this market in the last few years. Given the results, Middle East/North
Africa have apparently been willing to import whatever is necessary to
supply their needs. Given that this region is a net exporter it is also
possible that part of the imported fruit had been used for reexports
and/or processing. The differences in magnitudes imply that consumers
would prefer to import certain products before others while supplies last.
Far East. Far East imports represented 4.3% of total world imports
in the period studied. Major Far East partners are the United States,
Middle East/North Africa, Oceania, and rest of Africa. Three of the
product demands have negative relative price elasticities, two of which
are significant. Negative and significant elasticities were obtained for
the Middle East/North Africa and the rest of Africa. The first one is
inelastic and the second one elastic. Oceania has a negative and
inelastic elasticity, but it is not significantly different from zero.
United States product demand shows a positive relative price elasticity.
This would indicate that consumers in the Far East are willing to consume
more from the United States, even when its relative price is going up.

213
This result confirms the discussion maintained in previous sections about
the characteristics of the markets and consumers in the Far East with
regards to fast growth and interest in quality and high-grade fruit. This
is especially true for trade between the United States and the Far East
markets.
Total market demand elasticities for Far East's four major partners
are positive except for the case of the Middle East/North Africa.
Elasticities are 2.66 for the United States, 2.69 for the rest of Africa,
and 1.3 for Oceania. Middle East/North Africa elasticity is negative but
not significantly different from zero. Elasticity magnitudes indicate
that product demands from the different sources are very sensitive to
changes in the size of the market. Since the Far East market has been
growing fast in the last 21 years, shifts from one region to another are
common and will probably continue in the future.
Communist Bloc. The Communist Bloc imports represented 13.6% of the
world's total imports in the period considered. Major partners are Middle
East/North Africa, Mediterranean-EC, and Latin America. Two of them have
significant negative relative price elasticities. Mediterranean-EC has a
positive but not significant elasticity. Middle East/North Africa and
Latin America results show product demands with highly elastic relative
price elasticities.
Total market demand elasticities for Communist Bloc's three major
partners are positive and significant. Elasticities are 3.38 for Latin
America, .28 for Mediterranean-EC, and 1.24 for the Middle East/North
Africa. The results show an elastic response with respect to total market
demand for Latin America and Middle East/North Africa products and an

214
inelastic one for Mediterranean-EC. Given an increase in the Communist
Bloc market size, consumers will consume relatively more from Latin
America and the Middle East/North Africa than from Mediterranean-EC.
Latin America. Mediterranean-EC, rest of Africa, and Oceania. Latin
America, Mediterranean-EC, rest of Africa, and Oceania are net exporters.
Their imports represented only .5% of total world imports during the 21-
year period considered. Latin America major partners are the United
States and EC. The results for relative price elasticities indicate that
in Latin America the product demand for the United States has a positive
elasticity and, for the EC is negative but not significant.
Mediterranean-EC major partners are Latin America, EC, and Middle
East/North Africa. Demands for EC and Middle East/North Africa products
have negative and significant elasticities. Relative price elasticity for
Latin America product demand is positive but not significant.
Rest of Africa major partners are EC, Middle East/North Africa, and
Oceania. Elasticities for EC and Middle East/North Africa products are
positive. However, the elasticity from the EC is not significant. The
demand for the Oceania product is negative and significant.
Oceania major partners are the United States, Latin America, and the
Middle East/North Africa. Elasticities for the United States and Middle
East/North Africa products are negative and significant. Latin America
product has a positive and significant elasticity.
Relative price elasticities turned out to be positive and
significant in three cases. They are the demands for United States
product in Latin America, Middle East/North Africa product in rest of
Africa, and Latin America product in Oceania. As mentioned before, for

215
Latin America and rest of Africa it can be argued that imports are
required only on special occasions, probably related to insufficient local
production or high quality needs. These conditions may partially explain
the positive signs. Given the number of observations for the Latin
America product demand in Oceania, the results in this case are probably
related to insufficient information.
Total market demand elasticities for low import regions indicate
that, in most cases, the parameters are significant. The exceptions are
for Middle East/North Africa and Oceania products in the rest of Africa
and for the Middle East/North Africa product in the Mediterranean-EC.
CIF price linkage equations
Tables 5.29 to 5.31 show the estimated parameters and their
associated "t" statistics for CIF price linkage equations. These
equations measure the linkage between the CIF (Cost-Insurance-Freight)
import price in the final market and the FOB (Free-On-Board) export price
in the exporting region. There will be one CIF price linkage equation in
each region for each one of the partner regions. Since the model has a
total of 11 regions, the number of estimated CIF price linkage equations
should be 110 (11 regions with ten partners each).
The variables considered in the estimation of this section of the
model are FOB export price, a year trend, and an index price for energy.
The FOB export price variable refers to the Free-On-Board price in the
exporting region, i.e., the price of fresh oranges in the port from which
the export will be made. If the FOB export price in the exporting region
increases, it is expected that the CIF import price in the final market

216
also increases. The magnitude of this relationship is expected to be one
in the ideal case. That will be the event when there are no
transportation costs to be added or other external factors which affect
the relationship. This is obviously not the case for the fresh orange
trade model developed here. However, the results are expected to be close
to one and have a positive sign.
The year trend variable was included in the model to capture
structural changes in the industry and the transportation-system. These
exogenous effects are not expected to be predicted by other variables in
the model. The sign of this variable could be positive or negative,
depending on the type of structural change occurring between two trading
regions.
The index price for energy was included in the model to capture
changes in transportation costs due to changes in the world price of oil.
Since transportation cost is expected to increase (decrease) with
increases (decreases) in the price of oil, then the expected sign is
positive.
Tables 5.29 to 5.31 have the same structure. The results are
presented in a matrix where both columns and rows represent regions. The
columns represent the final market region or importer and the rows the
partner regions or exporters. Some CIF price linkage equations were not
estimated, given insufficient trade between some regions. Therefore, some
entries are empty.
The discussion about the CIF price linkage equations will be
developed in general terms, i.e., the regions will not be addressed on a
one-by-one basis. The reason for this approach is that, in general, the

217
Table 5.29 CIF Price Linkage FOB Export Price Elasticities3
Region ib
REGION jb
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
COMMB
US
0.215c
0.275d
2.960
2.241
0.556
2.103
0.632
2.365
0.983
2.014
1.033
3.456
0.792
2.588
0.630
3.572
1.244
3.513
CAN
0.596
2.271
0.689
2.510
1.021
2.183
0.792
4.293
LA
1.098
2.203
0.109
0.418
0.090
0.193
0.210
0.995
0.040
0.185
0.739
2.019
1.026
7.611
0.957
2.533
0.290
0.896
1.583
3.636
MED-EC
0.884
2.804
0.850
2.598
0.726
2.151
1.065
7.241
0.863
7.095
0.789
1.935
0.136
0.283
0.917
2.267
0.131
0.296
1.247
6.517
EC
0.987
4.728
0.985
2.969
0.975
3.025
0.222
0.426
1.301
4.395
0.872
2.905
1.087
5.700
0.985
2.723
1.219
2.093
RWE
1.088
3.219
1.274
3.576
1.167
2.414
0.725
2.025
1.100
2.046
ME/NA
0.736
2.223
0.831
2.904
0.649
1.214
1.192
2.698
1.063
4.557
1.074
8.156
0.578
2.099
2.523
3.333
0.616
2.125
1.937
5.342
RAF
1.058
2.103
0.579
2.523
0.650
4.807
0.433
2.450
0.600
1.591
FE
1.131
3.230
0.966
2.063
0.935
3.915
0.831
2.854
1.243
3.407
1.390
2.760
0.867
1.854
1.389
5.036
OCE
1.001
2.042
0.363
0.925
0.886
2.295
0.644
2.627
1.004
2.147
1.239
3.340
0.499
2.359
COMffi
0.408
0.841
0.895
3.671
0.606
3.820
1.407
2.347
0.612
2.512
aCIF price linkage equals j and FOB export price equals Fj^j.
^Region i across the top is the region importing from region j down the column.
cThe first line in each region represents parameter values.
dThe second line in each region represents the t Statistic.

218
Table 5.30 CIF Price Linkage Year Trend Elasticity3
Region ib
REGION jb
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
CCKMB
US
3.462
0.999d
-6.093
-1.452
0.399
0.205
0.014
0.009
0.033
0.005
0.478
0.198
0.086
0.060
1.268
1.042
-1.620
-0.421
CAN
4.U6
1.178
-2.441
-0.556
-1.167
-0.380
-2.603
-0.999
LA
-4.576
-2.110
3.902
1.907
5.846
1.634
1.167
0.738
3.696
2.394
0.507
0.217
0.715
0.424
1.269
0.563
0.478
0.291
-2.826
-1.199
MED-EC
-1.500
-0.914
1.098
0.473
0.128
0.047
-1.468
-2.190
1.132
1.991
-3.046
-0.973
5.798
1.909
-0.660
-0.430
1.526
0.583
-0.078
-0.108
EC
-0.272
-0.111
-4.019
-1.342
0.775
0.498
2.425
1.097
-1.738
-1.563
2.450
0.756
0.066
0.107
-2.652
-0.844
0.001
0.000
RWE
4.199
1.406
-0.405
-0.300
-3.583
-0.716
0.997
0.385
-1.873
-0.472
ME/NA
2.403
1.230
-0.090
-0.048
2.005
0.305
-6.370
-1.326
0.671
0.952
1.217
2.819
2.698
2.073
-2.509
-0.877
-0.720
-0.233
-0.406
-0.263
RAF
0.260
0.151
0.005
0.005
0.429
0.768
0.746
0.646
0.916
0.432
FE
-0.067
-0.036
0.388
0.159
-0.479
-0.227
1.539
0.607
-2.223
-0.606
2.607
0.860
-0.005
-0.001
-3.520
-1.450
OCE
1.749
0.457
-1.048
-0.628
-0.145
-0.103
-1.860
-1.808
-3.679
-1.311
2.251
1.008
0.753
0.963
COMB
1.640
0.443
-0.682
-0.649
0.627
0.515
-2.049
-0.411
-0.116
-0.112
aCIF price linkage equals Cjj and Year trend equals TRD.
^Region i across the top is the region importing from region j down the column.
cThe first line in each region represents parameter values.
dThe second line in each region represents the t Statistic.

219
Table 5.31 CIF Price Linkage Index Price For Energy Elasticity*
Region ib
REGION jb
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
COM
US
0.018c
0.205d
-0.246
-0.920
0.128
1.557
0.123
1.788
-0.001
-0.003
-0.065
-0.397
-0.016
-0.253
-0.011
-0.206
0.058
0.407
CAN
-0.116
-0.624
0.407
1.950
0.028
0.144
0.159
1.357
LA
0.182
2.682
0.154
1.476
0.109
0.618
0.176
2.370
0.154
3.377
0.048
0.426
-0.055
-0.544
-0.005
-0.034
0.169
1.540
0.004
0.035
MED-EC
0.163
1.493
0.094
0.476
0.170
0.928
-0.011
-0.220
-0.004
-0.103
0.251
0.980
-0.116
-0.908
0.011
0.200
0.257
1.732
-0.031
-0.497
EC
0.047
0.320
0.307
2.047
0.004
0.034
0.059
0.357
-0.012
-0.129
-0.088
-0.551
-0.072
-1.312
0.123
0.852
-0.023
-0.184
RWE
-0.289
-2.155
-0.121
-0.981
0.205
0.943
-0.002
-0.014
0.178
0.756
ME/NA
0.032
0.202
0.095
0.698
0.032
0.098
0.366
1.473
-0.030
-0.467
-0.047
-1.334
-0.064
-0.652
-0.155
-1.460
0.158
0.604
-0.194
-2.455
RAF
-0.011
-0.057
0.102
1.490
0.122
2.458
0.157
1.940
0.062
0.962
FE
-0.043
-0.670
0.003
0.037
-0.033
-0.212
-0.024
-0.126
0.164
0.675
-0.222
-1.099
0.075
0.356
0.232
1.971
OCE
-0.084
-0.254
0.278
2.922
-0.016
-0.092
0.238
2.724
0.259
1.406
-0.205
-0.946
0.112
2.518
COM
0.186
1.338
0.042
0.590
0.089
1.732
-0.073
-0.365
0.077
0.848
^IF price linkage equals Cjj and index price for energy equals PEN.
^Region i across the top is the region importing from region j down the column.
cThe first line in each region represents parameter values.
dThe second line in each region represents the t Statistic.

220
results have similar interpretations and differ very little from the
expected signs and values.
Table 5.29 presents a total of 82 estimated FOB export price
elasticities. All of them are positive as expected. Seventy one are
significant, 30 of them elastic and the rest inelastic. Most FOB export
price elasticities are close to unity. This indicates that, for a given
change in the FOB export price in the exporting region, a similar change
will occur in the CIF import price in the importing region. This result
was expected.
Table 5.30 shows the results for the year trend variable. Signs are
mixed, as expected, and 31 of the 82 parameters are significant. All
regions have at least one CIF price linkage equation with a significant
parameter. It is interesting to notice that only elastic (greater than
one in absolute value) positive or negative elasticities are significant.
This implies that, for those CIF price linkage equations with significant
parameters, the CIF prices are changing faster than the year trend.
Industry and transportation-system structural change seems to be
unimportant for those trading regions with insignificant parameters. This
implies either that there had not been a structural change or that the
change had been negligible.
Table 5.31 presents the elasticities for the index price of energy.
The values obtained indicate that 49 have the correct positive sign and 33
have negative signs. However, only six of the negative elasticities are
significantly different from zero. The six significant elasticities are
spread among a few regions. They are rest of Western Europe product
demanded in Mediterranean-EC; Middle East/North Africa product demanded in

221
the rest of Western Europe; EC and Far East products demanded in the rest
of Africa; Middle East/North Africa product demanded in the Far East; and
Middle East/North Africa product demanded in the Communist Bloc. Given
that, in most cases, the relationship is strong and the rest of the model
is well behaved with respect to this variable, two possible alternatives
could explain this situation: a data problem, or the existence of certain
structural change still not predicted by the model.
Conclusion: economic analysis
The economic analysis shows that the model results seem more
satisfactory for total market demands than it does for export supply
equations. In most cases the model is capturing major variations of total
market demands and export supplies for leading regions in world markets.
Most total market demand elasticities were between the expected signs and
magnitudes and made sense in most cases. In the events where wrong
elasticity signs were obtained, possible explanations were given.
The results for the export supply equations indicate that the FOB
export price is apparently not a major factor for export supplies. This
is an unexpected result. The other variables included in the model are
reflecting most of the export supply variations. Fresh production is the
strongest variable in the model. Nevertheless, export supply equations
for major world exporters behaved quite well as concluded in the
statistical analysis.
Once again, the results show that major trade flow equations are
captured by the model in most cases. Product demand equations for the

222
most important trading regions have the correct signs. Magnitudes were
usually in the normal ranges.
CIF price linkage equations are definitely the best behaved in every
case. The equations responded correctly to the expected signs and
magnitudes in almost every case.
Application for Policy Purposes
The empirical results and implications have been discussed using
three complementary analyses, graphic, statistical, and economic. The
general results indicate that the fresh orange trade model developed here
captures the trend of all dependent variables; has a general good fit; is
apparently well specified; predicts most turning points; and in a majority
of cases conforms to economic expectations.
These results do not apply for every equation. However, in most
cases the best behaved equations belonged to the leading regions and trade
partners in the world's fresh orange industry. In any event, given the
size and complexity of the model, the individual equation problems are
difficult to adjust. A possible solution to obtaining better results
might be accomplished by developing a different model for every equation.
Given the number of equations in the model, this task was not possible
given the resource constraints on this study.
The results obtained from this study have a number of policy
applications, including changes in market shares, market growth, and
reallocations among markets. Drawing on the empirical results, one can
readily address many of the important policy questions. Rather than

223
dealing with all possible issues, it is probably more useful to illustrate
the application of the model with specific examples. Two issues are of
particular importance across every region. First, what would cause the
total market demand to change, and can that change be predicted? Second,
given that the market demand grows for a particular region, how will the
market be supplied (i.e., who will be the exporters?) Asked another way,
how will each region's share of the market growth change? In this brief
section, a few examples are given showing how to address these questions.
To illustrate a selected case, suppose that the economic development
in the Communist Bloc yields an increase for real income (GDP) of 5% a
year in the next five years. Drawing from the results in Table 5.1, the
income elasticity for total market demand in this region is 1.176.
Therefore, if the income growth assumption is valid, fresh orange total
market demand is expected to increase 5.88% per year, holding other
variables fixed.
Next, the issue of who will supply this demand growth can be shown
using the empirical results from the product demand equations. For
example, from the United States perspective, what will the benefit of an
increase in the Communist Bloc's total market demand to the United States?
That is, will the United States share of the Communist Bloc market grow?
Table 5.13 provides the Communist Bloc product demand for each region
exporting to this market. The first row in this table corresponds to the
United States supply. If the relative price of United States exports were
held fixed and the Communist market grew by 5.88% as assumed above, then,
with this equation, one would predict the United States exports to grow by
2.85% (e.g., 5.88*.484-2.85) The Far East, Latin America, and the Middle

224
East/North Africa will increase their exports by 46.14%, 19.85%, and 7.27%
per year, respectively. In terms of the United States market share, the
model would point to a decrease in the United States share since the
market share elasticity in this equation is -.516 (see equations 4.25 and
4.26). Thus, for this specific example, the United States loses shares in
this importing region relative to other suppliers. Clearly, other
regions' shares of this market must be increasing. The Far East, Latin
America, and the Middle East/North Africa will increase their market share
in the same period.
Given that the United States share actually declines, the model can
also be used to show how much of a price adjustment would be needed to
offset the decreasing market share. These results clearly indicate that
the United States must be more price competitive in this example in order
to prevent an erosion in their share of this market.
Now, suppose that the population of the United States grows by 1% a
year for the next five years. Table 5.1 indicates that fresh orange total
market demand will increase by 2.12% a year in the same period. Table 5.3
shows the product demands' total market demand elasticities for the United
States. The model predicts that Oceania, Mediterranean-EC, Middle
East/North Africa, and Latin America will benefit from increases in United
States total market demand holding relative prices fixed. However,
Oceania and Mediterranean-EC are predicted to have the major benefits.
Suppose that world prices increase in the same proportion for all
regions. What will happen with the product demands in the Communist Bloc?
Table 5.1 shows that total market demand in the Communist Bloc will
decrease. Table 5.13 indicates that United STates product demand in that

225
region will also decrease given the change in the Communist Bloc total
market demand. The United States product demand will also be affected
through the price changes. The final direction of the United States
product demand in the Communist Bloc will depend on the specific
percentage change in prices and the parameter values.
These same procedures can be applied to other total market demand
and to any of the product demands. The results will differ depending on
the specific elasticities for the situation being explored. Discussions
and analyses similar to the previous ones can be developed for every
region and partner region, and for the rest of the variables included in
the model. Policy decisions can be proposed or supported by the results
obtained using the model.
Conclusion
The first part of this Chapter described the different steps
followed to estimate the fresh orange trade model developed in Chapter 4.
The second part presented and analyzed the empirical results and their
major implications. Given the nature of the model, a NL2SLS procedure was
utilized to estimate the model. The parameters obtained are consistent
but biased. The empirical results and the analysis developed indicate
that the model generally behaved well. Therefore, it can be used to
predict changes in the world fresh orange industry given changes in the
different variables.
In the next chapter a sensitivity analysis procedure will be
developed. Leading world fresh orange industry participants will be
evaluated under changes in the most important variables of the model.

CHAPTER 6
ECONOMIC IMPLICATIONS FROM SENSITIVITY ANALYSIS
Introduction
Two basic objectives were laid out in Chapter 1 regarding the
development of the fresh orange trade model. The first objective was to
develop a model to understand the major driving factors affecting world
fresh orange consumption and trade. This was accomplished in the
discussion in Chapter 5. The second objective was to determine what
happens when variables in the model change. In other words, what are the
comparative static implications of the trade model.
This chapter sets forth a sensitivity-analysis procedure to evaluate
the consequences of changes in the main variables of the model. The
results obtained from applying the procedure will complement the
discussion of Chapter 5 and will add new insights into the behavior of the
model.
In this chapter, the more important responses were illustrated by
selecting the major partner regions for each region. The variables that
better explain the model were also selected to be modified by the
sensitivity analysis. The comparative static implications in each case
were assessed. To illustrate the relative responses, scale effects were
removed by indexing the variables to the base year. The chapter also
provides a graphical presentation of the results, which helped to
226

227
visualize the pattern of adjustments to specific variables. It also
facilitated the comparison of the responses among partner regions in each
region. This comparison is not easy to see when looking only at the
coefficients, especially given the size of the model and the number of
parameters estimated. None of the analysis up to this point dealt with
adjustments in the variables. It was mainly a discussion on the
coefficients sign, magnitude, and significance. Certain variables have
important policy implications that could be clarified by using the
information presented here. As an example, given the characteristics of
the fresh orange trade data, a range of 30% above and below the base year
was considered reasonable. This gives an indication of the type of
responses and their limits for the fresh orange trade model (for example,
for relative prices). Much of the information in this chapter is intended
to help the reader to have a better understanding of the full model faced
with such a large number of coefficients or elasticities.
The chapter is divided into four sections. The first section
develops and explains the procedure for the sensitivity analysis. The
second discusses the rationale utilized to select the regions, equations,
and variables to be analyzed. The third develops the sensitivity analysis
for selected regions and equations, including a detailed discussion about
the results. The fourth summarizes the major conclusions and implications
of the chapter.

228
Sensitivity Analysis Procedure
The fresh orange trade model developed is a nonlinear simultaneous
system of equations. If the reduced form of the model can be obtained,
the model can be simulated as a whole for changes in the different
exogenous variables. The reduced form of a simultaneous system of
equations is obtained when all equations are expressed with only exogenous
variables in the right-hand side. This approach implies that, for a given
change in any exogenous variable, it is possible to assess the impact in
all 561 endogenous variables. A change in any exogenous variable produces
changes in all equations. The impact comes first from the exogenous
variable itself, and then from all the endogenous variables that will be
affected through the different equations. Given the size and complexity
of the model, the reduced-form parameters are difficult to obtain; and it
is not assured that they can be found. For the fresh orange trade model
presented in this study, it was not possible to solve for the reduced form
parameters. However, it was possible to perform a comparative static
analysis equation by equation.
Sensitivity analysis can be conducted to investigate the effects of
changes in the different variables of the model. It is possible to assess
the impact on any dependent variable, using the estimated parameters and
introducing changes in selected variables. This approach implies that the
analysis will not take into consideration the rest of the model when a
variable is changed and the impact on a given equation evaluated.
However, the estimation procedure does take into consideration the rest of
the model and its nonlinear and simultaneous characteristics.

229
The procedure developed works on an equation by equation basis.
Every equation has endogenous and exogenous variables in the right-hand
side. Therefore, the variables to be changed in the sensitivity analysis
could be exogenous or endogenous. Given that four different types of
equations were estimated, the variables to be changed will vary depending
on the type of equation analyzed. A total of 242 equations were estimated
using a simultaneous system approach. Each region has one total market
demand, one export supply, and ten product demands and CIF price linkage
equations. The main objective of the sensitivity analysis was to evaluate
the impact on the dependent variable of a given equation for a given
change in one of the right-hand side variables.
Sensitivity analyses can be developed following different
approaches. The procedure has to provide the necessary tools to address
the important questions. In the fresh orange trade model, a major issue
of interest is to compare the behavior of the different regions under
different scenarios. Since consumption and trade volumes differ
dramatically among regions, to facilitate the analysis and its
interpretation, a common framework has to be built. One way to overcome
the problem is to develop an index number common to all regions through
which they can be compared.
The index chosen for the present study is based on a starting year.
The changes in the variables will depart from the base year and the
results will be evaluated and compared for the different regions. The
decision about the base year depends on the type of questions to be
addressed. As mentioned above, the objective of sensitivity analysis was
to forecast changes in the dependent variables given changes in the

230
different right-hand side variables. The last observation of the original
data set is 1986. Using 1986 as the base year, departures from 1986
provide simulated values in response to specific variable levels.
The procedure and the computer program used for the sensitivity
analysis is included in Appendix I. The first temporary data set was
developed that included seven simulated values for all of the original
variables at the 1986 level. All observations had exactly the same
values, i.e., 1986 values. The temporary data set was then modified by
using a step procedure that affected each observation. The temporary data
set was multiplied by a vector containing .7, .8, .9, 1, 1.1, 1.2, and
1.3, thus giving a completely new data set with seven simulated values
expressed as some percentage of the base. The observations of the new
data set ranged from 30% below the 1986 values in the first observation to
30% above the 1986 values for the seventh observation. Thirty percent
above and below the 1986 values was selected considering that bigger
percentage changes were unlikely to have occurred. See Sparks (1987) for
another application of this simulation approach.
Two additional data sets were needed to complete the sensitivity
analysis: the original data set with 1986 values, and a new data set
containing the values of all the estimated parameters of the model.
Since the estimated model equations were given in the log form, it
was necessary to reexpress all the equations in exponential form. The
equations have dependent variables no longer in the log form in the left-
hand side and exponential equations on the right-hand side. The right
side of the equations included the estimated parameters and the right-hand
side variables (see Appendix I). The specific values to substitute for

231
the right-hand side variables were obtained from the simulated data sets.
In order to compare the changes in the dependent variables, only one
right-hand side variable was modified at a time. The rest of the
variables remained at their original 1986 values.
The substitution of the new data sets into the equations generated
seven new dependent variable values, one for each percentage adjustment to
the base. The new simulated dependent variable values were indexed to the
1986 base value. The index helped to show the implications from changes
in the right-hand side variables in the different equations and markets.
It also helped to compare the results among the different regions. If the
index number was one, then the new dependent variable value was equal to
the 1986 value. If the index number was above or below one, the simulated
value was above or below the 1986 value, respectively.
The sensitivity analysis and resulting responses can be illustrated
with figures that show the behavior of the different dependent variables,
given changes in the right-hand side variable. The graphical approach
provides a useful interesting framework to illustrate the impact of
changes in the different equations and markets and to compare the results
among regions.
For example, Figure 6.1 shows the total market demand for major
world consumers. Total market demands are functions of average market
price, income (GDP), population, and substitute product price. The
average market price is one of the right-hand side variables. The figure
shows an index number ranging from .7 to 1.3 for the average market price
on the bottom axis. That is, the average market price has been modified
from 30% below to 30% above the 1986 price level. The left axis shows an

232
index number for total market demand. The simulated or new total market
demand values are obtained by substituting the modified average market
prices into the original equation while holding all other variable values
fixed at the 1986 levels. The exact index on the left axis will depend on
the simulated values actually obtained, given the step-wise changes in the
average market price.
Rationale for Region. Equation, and Variable Selections
The analysis of the fresh orange industry shows that total
consumption, imports, and exports are concentrated in a few regions. Some
regions are major consumers, others major importers or exporters, and most
regions have a small set of important trade partners. Considering these
conditions, it is reasonable to select a subset of regions on which to
perform the sensitivity analysis in each event. In most cases, a few
regions will represent over 90% of total world consumption or exports, and
a few regions will also account for over 90% of total supply. Applying
the sensitivity analysis to all 262 estimated equations will provide
little additional information since many relationships will have almost no
impact on the major factors affecting the world fresh orange industry.
Total market demand equations represent total consumption or demand
for fresh oranges in a given region. The model developed considers both
total domestic consumption and total trade. Therefore, total demand and
total imports per region relative to the rest of the world should be
considered to select the more important regions. Given that a small group
of regions represented major world consumers and another small but

233
different group represented major world importers, two groups were
selected for the total market demand analysis. This implies that two sets
of figures will be included in the sensitivity analysis, one set for major
world consumers and another set of figures for major world importers.
The selection of regions for the export supply analysis was based on
major world exporters. Total exports per region relative to the rest of
the world were considered to select the relevant regions. Given that a
small group of regions represented most of the world exports, one set of
figures will be presented in this case.
The selection of partner regions for the product demands and CIF
price linkage equations was based on relative volumes supplied by each
partner region. The volumes were accumulated, and at least 90% of total
supply had to be accounted for, to decide which regions to include. Each
region had a different set of partners, depending on its major trade
flows.
The final objective of any simulations or sensitivity analysis is to
find out what are the forecasted values of the main variables. The main
variables for the fresh orange trade model are total market demands,
export supplies, and product demands for important regions. Most
intermediate variables and equations in the model will change, given
changes in the right-hand side variables. However, the final impact on
the main variables of the model is the important issue.
CIF price linkage equations were developed to capture the linkage
between the FOB export price in supplying regions and the CIF import price
in the final markets. These equations were not considered in the
sensitivity analysis. Import prices in the final market were assumed to

234
be known and were changed from approximately 30% below to 30% above 1986
levels. It is also possible to determine what the necessary change in the
FOB export price, tariff, tax, or any other factor would be when the final
market import price changes over the proposed range. To perform this
analysis, the respective CIF price linkage equation and some of the model
identities had to be used. These questions are important, but they can
always be addressed at a later step considering the results and the
specific regions of interest. Therefore, the equations selected for the
sensitivity analysis were total market demands, export supplies, and
product demands.
Each equation selected is a function of a different set of
variables. A decision had to be made regarding the set of variables to be
modified for each equation. Total market demands are functions of average
market price, income (GDP), population, and substitute product price.
Changes in the average market price are related to changes in tariffs,
taxes, local prices, and FOB export and CIF import prices. Economic
theory and the empirical results of Chapter 5 indicate that average market
price and income (GDP) are the major driving factors for consumption in
most regions. Therefore, these variables were included in the sensitivity
analysis for total market demand equations.
Export supply equations are functions of the FOB average export
price and domestic fresh production. The empirical results of Chapter 5
indicate that fresh production is probably the major driving factor for
exports. However, economic theory suggests that FOB average export price
should also be a major factor. Given these conditions, both variables
were included in the analysis.

235
Product demand equations were defined as functions of the relative
price and total market demand. Relative price variables refer to the
import price of a product coming from a certain region relative to the
final market average price. The import price could change relative to the
final market average price, due to changes in tariffs, taxes, FOB export
prices, other factors included in the CIF equations, and other causes. If
the import price for a certain region increases relative to the average
market price, less consumption relative to other suppliers is expected in
the final market. Total market demand variables measure total consumption
of fresh oranges in the final market. It is a measure of the size of the
market. Economic theory and the empirical results in Chapter 5 indicate
that both variables help to determine trade flows. Relative price and
total market demand variables were included in the sensitivity analysis.
Sensitivity Analysis
This section of the chapter will present and evaluate the results of
the sensitivity analysis. Each type of equation will be addressed
separately. The analysis and discussion will focus on major trading
regions. To ease the presentation and discussion, a graphical analysis
constructed using the sensitivity analysis indices was developed. The
figures will be used to evaluate individual market behavior and to compare
them among regions. The indices generated in the sensitivity analysis are
included in Appendix J.
The figures show, on the bottom axis, an index that represents the
right-hand side variable modified. The variable varies, depending on the

236
equation analyzed. By construction, this index always goes from .7 to 1.3
independently of the variable considered. That is, the right-hand side
variable has been modified over that range. On the left axis, the index
represents the endogenous or response variable. The specific variable on
the left axis could be total market demand, export supply, or product
demand depending on the equation studied. The index varies, depending on
the type of response of the endogenous variable in each case. The
response depends on the percentage change in the right-hand side variable
and the magnitude and sign of the estimated parameter.
The right-hand side variable index and the endogenous or response
variable index were used to construct the figures. Each figure shows the
regions in order of importance. The first region presented corresponds to
the most important region in the figure, the second to the next most
important, etc. The most important region corresponds to the largest
consuming region or importer for total market demands and to the largest
exporting region for export supplies. The criterion for product demands
was based on trade-flow volumes between partner regions and the final
market.
The first section will center on total market demands, the second on
export supplies, and the third on product demands. In each case, the
discussion will address consumers, importers, exporters, and trading
partners. A summary regarding this section of the chapter will be
presented at the end.

237
Total Market Demands
As shown in Table 6.1, six regions consumed 90.7% of total world
fresh orange consumption from 1966 to 1986. The regions were Latin
America, Far East, Middle East/North Africa, Mediterranean-EC, EC, and the
United States. A different group of six regions accounted for 99.5% of
total world fresh orange imports. In this case, the regions were EC,
Communist Bloc, rest of Western Europe, Canada, Far East, and Middle
East/North Africa. Given that some regions were important as consumers
and others as importers, the following analysis will cover both groups.
Two of the four variables included in the total market demand
equations were considered in the sensitivity analysis, the average market
price and income (GDP). These variables were selected based on the
implications of economic theory and the empirical results from Chapter 5.
Figures 6.1 to 6.4 present the sensitivity analyses for total market
demand equations. Figures 6.1 and 6.2 show total market demands while
changing the average market price for major world consumers and importers,
respectively. Figures 6.3 and 6.4 present total market demand responses
to changes in income (GDP) for major world consumers and importers,
respectively.
Average market price
Figure 6.1 presents the total market demands while changing the
average market price for major world consumers. The bottom axis shows the
average market price index with the index extending from .7 to 1.3. The
response or total market demand index is shown on the left axis. In

238
Table 6.1 World Demand, Imports and Exports Share Per Region
(Cummulative 21 Year Period 1966-1986)
Region
% of Total
Demand
% of Total
Imports
% of Total
Exports
United States
7.12
1.20
8.90
Canada
0.75
4.81
0.00
Latin America
28.59
0.06
3.92
Mediterranean-EC
10.59
0.07
44.00
EC
9.92
63.43
0.25
Rest of Western Europe
1.64
10.57
0.05
Middle East/North Africa
11.51
1.65
34.95
Rest of Africa
2.46
0.16
6.24
Far East
23.02
4.29
1.09
Oceania
0.81
0.20
0.48
Communist Bloc
3.62
13.55
0.12
100.00
100.00
100.00

239
TOTAL MARKET DEMAND INDEX (1986=1)
AVERAGE MARKET PRICE INDEX (1986=1)
Figure 6.1. Total Market Demand Changing Average Market Price (Major
World Consumers).

240
TOTAL MARKET DEMAND INDEX (1986=1)
INCOME INDEX (1986=1)
Figure 6.2. Total Market Demand Changing Average Market Price (Major
World Importers).

241
TOTAL MARKET DEMAND INDEX (1986=1)
AVERAGE MARKET PRICE INDEX (1986=1)
Figure 6.3. Total Market Demand Changing Income (GDP) (Major World
Consumers).

242
TOTAL MARKET DEMAND INDEX (1986=1)
1.6
1.4
0.6
0.4
EC
~\ COMMB RWE
-0- CAN -X- FE -£>- ME/NA 1
0.7 0.8 0.9 1 1.1 1.2 1.3
INCOME INDEX (1986=1)
Figure 6.4. Total Market Demand Changing Income (GDP) (Major World
Importers).

243
this case, the response index goes from below one up to approximately 2.2.
In other words, it ranges from below the 1986 level up to approximately
2.2 times that level. The figure indicates that major world consumers
have different responses to changes in the average market price. Since
the response index depends on the magnitude and sign of the original
parameter, a negative relationship is expected. As the average market
price index increases, the total market demand index should decrease and
vice versa.
The figure shows that responses are negative, with the exception of
the Far East, and the magnitude of the responses differs from region to
region. For example, a 10% increase in the average market price is
represented on the bottom axis as 1.1. The values corresponding to the
total market demand index, starting from the lowest, are .812 for United
States; .893 for EC; .897 for Middle East/North Africa; .928 for
Mediterranean-EC; .995 for Latin America; and 1.017 for the Far East (see
Appendix J). The index numbers are less than one, except for the Far
East. This result indicates that total market demands will be lower than
the 1986 level in all regions after a 10% increase in the average market
price. The only exception is the Far East. Recall from Chapter 5 that
unique problems with the Far East equations were discussed.
The rest of the analysis will be based on the graphical results.
The specific index numbers utilized to construct the figures are shown in
Appendix J. The relative importance for each region is captured by the
order in which the figures present the different regions. For example,
Latin America is the largest consumer of fresh oranges in the world,
therefore, Figure 6.1 lists Latin America in the first position.

244
Figure 6.1 indicates that, given changes in the average market price
index, the total market demand index stays practically fixed. This
implies that the average market price is not a major factor for local
consumption in Latin America. The empirical results indicate that the
average market price parameter, or elasticity, for Latin America is close
to zero and not significant (see Table 5.1). The Far East is the second
largest consumer in the world (based on reported data). The figure shows
that, as the average market price index increases, the total market demand
index also increases. The empirical results indicate that its average
market price parameter is significant and inelastic. A possible
explanation for the positive direction of the relationship was developed
in Chapter 5. The argument included the characteristics of the market and
consumers in terms of growth and quality requirements. Market growth is
probably the most important factor to justify the wrong sign.
The figure shows that the total market demand indices for the EC and
Middle East/North Africa are very similar. The empirical results indicate
that both have elastic and significant parameters, implying that these
regions react in a similar way to changes in their respective average
market prices. The total market demand index for Mediterranean-EC is
smaller than those for the EC and Middle East/North Africa. This
indicates that Mediterranean-EC consumers are less sensitive to changes in
their average market price. The empirical results indicate that it has an
inelastic and significant parameter. The United States has the most
elastic average market price parameter in the group. That is, fresh
orange consumption in the United States is highly sensitive to changes in
the average market price.

245
Figure 6.2 presents the same indices as Figure 6.1, but for major
world importers. The bottom axis shows the average market price index.
The left axis shows the total market demand index. In this case, the
response index goes from below one up to approximately 1.5 times the 1986
level. This figure indicates that major world importers have different
responses to changes in the average market price, but they are closer to
each other than to the responses of major world consumers. The responses
are negative as expected, with the exception of the Far East. The EC is
Che major world importer. It has an elastic price parameter or elasticity
which is very similar to the ones from the Communist Bloc, Middle
East/North Africa, and the rest of Western Europe. The empirical results
indicate that the four elasticities are significant and lie between 1 and
1.22 (see Table 5.1). Canada's response is inelastic, implying that it is
less sensitive to changes in the average market price than former regions.
Figures 6.1 and 6.2 indicate that regions with high import levels,
such as the EC, Communist Bloc, rest of Western Europe, and Middle
East/North Africa, have similar total market demand indices; and their
parameters or elasticities are elastic. Regions with high consumption
levels and low imports relative to consumption will tend to have lower
average market price indices. Therefore, they are not very sensitive to
changes in average market prices. This is the case for Latin America and
Mediterranean-EC. These conclusions show that if world prices increase,
major importers will tend to consume proportionally less than regions with
low import levels. If world prices decrease, importers will tend to
consume relatively more than regions with low import levels.

246
The figures also show that the United States has the largest
response. Its parameter is the most elastic in the group, indicating that
consumers in this market are highly sensitive to changes in the average
market price. If world prices increase, consumers in the United States
will consume proportionally less fresh oranges than other regions in the
world. If world prices decrease, United States consumers will tend to
consume more relative to the rest of the regions considered.
Income (GDP)
Figure 6.3 presents total market demand responses to changes in the
income (GDP) level for major world consumers. The income index is shown
on the bottom axis. The total market demand index is presented on the
left axis. The expected relationship between income and demand is
positive. As the income index increases, it is expected that the demand
index also increases. The figure indicates that Latin America, Far East,
EC, and Middle East/North Africa have positive relationships. The
empirical results indicate that only the parameters for the EC and Middle
East/North Africa are significant. The results also indicate that these
parameters are inelastic (see Table 5.1).
The figure shows that the United States and Mediterranean-EC have
negative responses. The empirical results indicate that the parameter for
Mediterranean-EC is not significant. A negative total market demand index
response produces lower consumption as income increases. This is the case
for the United States, and it is an unexpected result. The negative
relationship shows that fresh oranges are considered an inferior good in
the United States. Consumers are expected to consume more of all normal

247
goods and shift to different good bundles as their income increases. A
good is considered inferior when consumers reduce its consumption level as
their income increases. In other words, the good is excluded from the new
bundle selected. In Chapter 5, the "t" statistic for the parameter was
compared with the rest of the statistics in the equation. The results
were weak, when compared to the rest of the variables in the equation.
This condition indicates that the income parameter may not be significant
in this particular case.
Figure 6.4 presents total market demand responses to changes in
income (GDP) level for major world importers. The income index is shown
on the bottom axis. The total market demand index is presented on the
left axis. The figure indicates that major world importers have similar
income responses and most of them are positive as expected. The figure
shows that only one negative relationship exists. This is the case for
the rest of Western Europe; however, the empirical results indicate that
it is not significant. The Far East has a positive response, but the
empirical result was also not significant. The EC, Canada, and Middle
East/North Africa relationships are positive and inelastic. Total market
demand indices for the EC and Canada are practically the same. These
responses show that as income increases (decreases), consumption will
increase (decrease) in a lower proportion than the income change. The
curve for the Communist Bloc indicates that the relationship is elastic.
A 30% increase in income generates more than a 30% increase in
consumption. Consumers in the Communist Bloc are more sensitive to
changes in income levels than consumers in other regions.

248
Major importers have significant and more consistent results for
changes in the income index than major consumers. Consumption of fresh
oranges for major consumers with local production is less sensitive to
changes in income. On the other hand, importers are more sensitive to
changes in income levels, and their consumption and imports will increase
or decrease as income increases or decreases.
If Figures 6.1 to 6.4 are compared, major importers (Figures 6.2 and
6.4) behaved in a similar way most of the time. The responses to the
average market price index and income index were close among the regions
and, in most cases, correct. Price and income parameters or elasticities
for regions with high import percentages relative to their total
consumption conform better to theoretical expectations and differ from
those obtained in regions with large local production.
The results and the analysis developed could be used by any region
to make policy decisions. The decisions could be related to domestic or
trade policy. Knowledge about the different reactions that major
consumers or major importers have to changes in the average market price
and income is valuable market information. The results could be used to
evaluate price policies, the competition level, the potential market,
market development and growth, the impact of trade barriers, and other
important factors.

249
Export Supplies
As shown in Table 6.1, three regions exported 87.8% of total world
fresh orange exports in the period considered. The regions were the
Mediterranean-EC, Middle East/North Africa, and the United States. The
sensitivity analysis will focus on these three regions.
The variables analyzed were the FOB average export price and fresh
orange production. Figures 6.5 and 6.6 present the sensitivity analyses
for export supply equations. Figure 6.5 shows export supplies while
changing the FOB average export price, and Figure 6.6 shows export
supplies while changing fresh production.
FOB average export price
Figure 6.5 has the FOB average export price index on the bottom axis
and the export supply index in the left axis. The expected economic
relationship between export supply and the FOB average export price is
positive. The figure shows that responses differ dramatically among
regions. Mediterranean-EC curve indicates that its response to changes in
the FOB average export price is close to zero. The response for the
Middle East/North Africa is positive as expected, but the one for the
United States is negative. The empirical results show that the "t"
statistic for the United States is one, indicating that it is significant
(see Table 5.2). The insignificance of the parameter obtained for
Mediterranean-EC and the sign of the parameter for the United States show
that the FOB average export price is not a major factor for export

250
EXPORT SUPPLY INDEX (1986=1)
Figure 6.5. Export Supply Changing FOB Average Export Price (Major World
Exporters).

251
EXPORT SUPPLY INDEX (1986=1)
FOB AVERAGE EXPORT PRICE INDEX (1986=1)
Figure 6.6. Export Supply Changing Fresh Production (Major World
Exporters).

252
decisions in these regions. Middle East/North Africa has the correct
positive response and it is also significant. The Mediterranean-EC and
United States major exports are directed to captive markets, while the
Middle East/North Africa has a larger clientele. These conditions
partially explain why the Middle East/North Africa has the correct
relationship.
Fresh production
Figure 6.6 presents export supplies changing fresh production. The
fresh production index is on the bottom axis and the export supply index
is on the left axis. The figure shows that, for the three regions, the
relationship is positive and strong as expected. Fresh production is an
important driving factor for exports. Mediterranean-EC and United States
have inelastic relationships indicating that their exports change less
than proportional to changes in fresh production (see Table 5.2). Middle
East/North Africa has an elastic relationship; therefore, its exports are
more sensitive to changes in fresh production. The differences in the
export supply indices indicate that Mediterranean-EC and United States
have larger local and/or captive markets than Middle East/North Africa.
However, the relationships are close between each other and close to one
in the three cases.
The sensitivity analyses indicate that export supply are driven
mainly by fresh production. Production was considered exogenous, there
may very well be a dynamic linkage between export growth and subsequent
supply responses, the model does not consider these linkages. The export
supply index while changing the FOB average price index is not conclusive,

253
with the exception of the Middle East/North Africa case. On the other
hand, the behavior of the three major exporters is consistent and
significant for the export supply index while changing fresh production.
The Middle East/North Africa is the region that has the most flexible
relationship regarding production and exports. This indicates that this
region has better opportunities to take advantage of new or growing
markets in the future.
Product Demands
A basic objective of the present study is to determine the major
factors affecting consumers' decisions regarding fresh orange imports from
alternative sources. Even though some of the 11 regions considered are
not major importers, they all have potential significance in terms of the
forecast and should be included in the sensitivity analysis.
Product demand equations include the relative price and total market
demand variables. The relative price variable refers to the import price
of a product coming from a given region relative to the average market
price in the final market. If the import price from a certain region
increases (decreases), it is expected that imports from that region will
decrease (increase) relative to other regions in the final market. For
simplicity in the following discussion, relative price will be called
import price.
The total market demand variable refers to total market consumption.
It represents the size of the final market. The change in the product
demands relative to this variable could be positive or negative. Product

254
demand will depend on consumers' preferences with respect to product
sources, as their total demand changes. For example, if the total market
demand increases in a certain region, the consumers' next step will be to
decide from which region to buy the extra product. The consumers'
decision could be in favor or against any potential source. The analysis
will consider market size increases and decreases. Market size decreases
below the 1986 level are less likely to occur, given the behavior of fresh
orange consumption in the last two decades.
The product demands were selected taking into consideration trade -
flow volumes among the regions. For every region, a group of partners
that accounted for over 90% of total imports were considered.
United States
Table 6.2 shows that Latin America, Middle East/North Africa, and
Mediterranean-EC accounted for 98.1% of total United States imports in the
period considered. Figures 6.7 and 6.8 present United States imports
while changing import prices and total market demand. The sensitivity
analysis presented in Figure 6.7 indicates that United States product
demand behavior differs dramatically, depending on the product source.
Latin America is the major exporter to the United States (based on
reported data). The relationship between the import price index and
product demand is positive in this event. This is an unexpected result,
probably related to the fact that the United States is self-sufficient and
imports exist only when local production is insufficient to fulfill
consumers' demand. If this is the case, then a positive relationship is
possible. As shown in the figure, the demand for Middle East/North Africa

Table 6.2
Region'
s Relative Imports Per
Partner
Region (%
Cummulative 21
Year Period
1966
-1986)
Region
us
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
COtMB
TOTAL
IMPORTS
us
0.04
83.62
2.79
0.04
0.00
11.71
0.04
1.67
0.10
0.00
100.00
CAN
79.67
1.23
1.08
0.01
0.00
5.48
3.44
8.39
0.69
0.00
100.00
LA
88.14
0.34
0.80
5.71
0.04
2.18
0.09
0.26
0.09
2.37
100.00
MED-EC
0.09
0.00
22.02
47.78
1.75
17.67
9.97
0.31
0.04
0.37
100.00
EC
1.82
0.00
3.28
54.19
0.03
32.69
7.69
0.01
0.11
0.18
100.00
RWE
0.94
0.00
1.02
43.41
1.50
47.55
5.33
0.04
0.16
0.05
100.00
ME/NA
1.49
0.00
14.93
8.37
1.12
1.85
28.48
39.47
4.23
0.05
100.00
RAF
0.18
0.00
5.89
7.81
13.86
0.05
68.18
0.15
3.83
0.05
100.00
FE
81.38
0.04
0.31
0.54
0.02
0.00
7.58
3.53
6.60
0.00
100.00
OCE
85.92
0.00
4.50
1.87
0.03
0.00
4.88
2.63
0.18
0.00
100.00
COMB
0.49
0.00
2.76
35.19
0.07
0.01
61.44
0.00
0.03
0.00
100.00
255

256
PRODUCT DEMAND INDEX (1986=1)
IMPORT PRICE INDEX (1986 = 1)
Figure 6.7. United States Imports Changing Import Prices.

257
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.8. United States Imports Changing Total Market Demand.

258
product has the correct relationship and is highly elastic. A small
change in the import price index produces a large change in the demand for
this product in the United States. The relationship for the
Mediterranean-EC is also negative, but the empirical results indicate that
it is not significant (see Table 5.3). Consumption of the Mediterranean-
EC product in the United States is not affected by changes in the import
price. A similar change in the import price index for the Middle
East/North Africa and Mediterranean-EC shows that the demand for the
latter region's product is more stable. This opens an interesting
opportunity for the Middle East/North Africa to increase its market
participation in the United States. A small decrease in their import
price in the United States would causes consumption of relatively more of
their product than that from the Mediterranean-EC and other sources.
Figure 6.8 shows that United States product demands are sensitive to
changes in the size of the market. Product demand indices for the three
regions are positive and highly elastic to changes in total market demand
indices. The figure and the empirical results show that Latin America has
the smallest parameter or elasticity. Parameters for the Middle
East/North Africa and Mediterranean-EC are over four and nine times the
parameter for Latin America, respectively. Under these conditions, if the
United States total market demand increases, consumers will prefer to buy
the extra fruit first from the Mediterranean-EC, second from the Middle
East/North Africa, and finally from Latin America. In all cases, these
volumes of imports are still very small relative to total U.S. consumption
of oranges.

259
Notice that the import price is not a relevant factor for demand of
Mediterranean-EC product in the United States, however, it has the largest
total market demand response. These characteristics imply that United
States consumers rate Mediterranean-EC product in a premium position with
respect to the rest of the fruit in the world market.
Canada
Table 6.2 indicates that 97.0% of Canada's imports came from four
regions, the United States, Far East, Middle East/North Africa, and rest
of Africa. Figures 6.9 and 6.10 show Canada's imports while changing
import price and total market demand, respectively. Figure 6.9 presents
the relationship between the import price index and the product demand
index. The results of the sensitivity analysis indicate that only the
demand for rest of Africa product has the incorrect relationship. The
empirical results show that this parameter is not significant (see Table
5.4). The rest of the regions have negative and inelastic relationships,
indicating that for similar percentage changes in the import price,
consumers in Canada will react differently depending on the region of
origin. The figure shows that, for a similar increase in the import price
index, consumers will consume proportionally less from the Far East and
United States than from the Middle East/North Africa. However, the
empirical results indicate that the parameters for the United States and
Middle East/North Africa are not significant.

260
PRODUCT DEMAND INDEX (1986=1)
IMPORT PRICE INDEX (1986 = 1)
Figure 6.9. Canada Imports Changing Import Prices.

261
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.10. Canada Imports Changing Total Market Demand.

262
Figure 6.10 presents Canada's imports while changing total market
demand. The figure indicates that, for the United States and Far East,
the relationship between the total market demand index and the product
demand index is positive. The rest of the regions have a negative
association. The expected direction in this case is either negative or
positive. Recall that, if the association between these indices is
negative, consumers tend to consume less from a given region when the
market size increases. The results imply that, as the Canadian market
increases, consumers shift from Middle East/North Africa and rest of
Africa to United States and Far East products. If the size of the market
decreases, then the relative consumption of Middle East/North Africa and
rest of Africa products with respect to the other regions will be larger.
The demand for United States and Far East products is more stable, and
both can take advantage of market size increases. The figure also shows
that product demands are more sensitive to reductions below the base than
to increases above the base market size. However, market size decreases
are less likely to occur.
While the import price may not be a relevant factor for United
States imports in Canada, it does have a positive and strong total market
demand parameter. This characteristic suggests that Canadian consumers
rate United States product in a premium position with respect to the rest
of the fruit in the world market. Also, the closeness of the two
countries and the ease of trade have likely impacted these results.

263
Latin America
Table 6.2 indicates that four regions accounted for 98.4% of total
imports in Latin America. The regions are the United States, EC,
Communist Bloc, and Middle East/North Africa. The sensitivity analysis
presented in Figure 6.11 shows Latin America imports given changes in
import prices. The EC and Middle East/North Africa have the correct
negative association. The United States and Communist Bloc have positive
relationships. Only the demands for the United States and Middle
East/North Africa products are significant (see Table 5.5). The United
States product demand has an unexpected result. It is probably related to
the fact that Latin America is self-sufficient, and imports exist only
when the market is experiencing a substantial shortage. If this is true,
the direction of the relationship could be positive. On the other hand,
it is expected that demand for Middle East/North Africa product will tend
to decline relative to other products as import prices increase. The
reverse could be expected if prices decrease.
Figure 6.12 presents Latin America imports while changing total
market demand. The figure indicates that the United States and Middle
East/North Africa have negative relationships, while the EC and Communist
Bloc have positive associations. As total market demand increases,
consumption for United States and Middle East/North Africa products will
decline relative to other regions. This is especially true in this case,
since the Communist Bloc and EC import price variables were not
significant, indicating that import prices will probably not affect their
demands. The Communist Bloc has the strongest relationship between

264
IMPORT PRICE INDEX (1986 = 1)
Figure 6.11. Latin America Imports Changing Import Prices.
1.3

265
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.12. Latin America Imports Changing Total Market Demand.

266
product demand and market size. The empirical results indicate that the
parameter for the Middle East/North Africa is not significant.
Mediterranean-EC
Table 6.2 indicates that four regions accounted for 89.2% of total
imports in the Mediterranean-EC. The number of observations for this
region was insufficient to estimate some of the equations. Lack of
sufficient import information was expected, given that Mediterranean-EC
was a net exporter of fresh oranges. The product demands estimated are
for the EC, Latin America, Middle East/North Africa, and rest of Western
Europe. The sensitivity analyses presented in Figure 6.13 indicate that
three of the four regions have the correct negative relationship. The
only positive association corresponds to Latin America, but the empirical
result indicates that it is not significant (see Table 5.6). Middle
East/North Africa and rest of Western Europe have similar responses.
Their response is smaller than the one from the EC. If import prices
increase, consumers will shift their relative consumption from the EC to
other regions. The figure also indicates that demand for EC product is
more sensitive to price decreases from the base year than it is to price
increases from the base year. If import prices decrease, then the EC will
capture most of Mediterranean-EC imports.
Figure 6.14 shows Mediterranean-EC imports changing total market
demand. The figure indicates that all relationships are positive. The
parameters for the Middle East/North Africa are not significant. With
increases in the size of the Mediterranean-EC market, imports will come
first from the rest of Western Europe, second from the EC and finally from

267
IMPORT PRICE INDEX (1986=1)
Figure 6.13. Mediterranean-EC Imports Changing Import Prices.
1.3

268
PRODUCT DEMAND INDEX (1986=1)
30.0
25.0
20.0
15.0
10.0
5.0
o.offl
0.7
a

EC
|
LA
*
ME/NA
-B-
RWE
0.8 0.9 1 1.1
TOTAL MARKET DEMAND INDEX (1986=1)
1.2
1.3
Figure 6.14. Mediterranean-EC Imports Changing Total Market Demand.

269
Latin America. Since rest of Western Europe is not a producer and EC
production is small, then it is apparent that consumers' first choice will
be Latin America and then the Middle East/North Africa. Given that Latin
America product demand is not affected by import price changes, its
position in the market is even stronger. As explained in Chapter 5,
United Nations trade data tapes included exports from rest of Western
Europe which are probably related to reexports.
EC
The EC is the world largest importer of fresh oranges. Table 6.2
indicates that five regions accounted for 99.6% of total imports during
the period considered. The regions are the Mediterranean-EC, Middle
East/North Africa, rest of Africa, Latin America, and the United States.
As shown in Figure 6.15, two regions have the correct negative association
between import price and product demand indices. The product demands for
the other three regions have positive relationships, but two of them are
not significant (see Table 5.7). Demand relationships for rest of Africa
and United States products are negative and significant. However, the
United States parameter is more elastic. As the import price index
increases, consumers in the EC will shift their consumption from United
States product to rest of Africa product. The demand for Middle
East/North Africa product is positive and significant. This result might
be related to the fact that the EC will buy product from the Middle
East/North Africa only when market prices are increasing, due to

270
PRODUCT DEMAND INDEX (1986=1)
0.7 0.8 0.9 1 1.1 1.2 1.3
IMPORT PRICE INDEX (1986 = 1)
Figure 6.15. EC Imports Changing Import Prices.

271
insufficient fruit being provided by its major supplier, the
Mediterranean-EC. Demands for Mediterranean-EC and Latin America products
are not significant.
The sensitivity analysis presented in Figure 6.16 shows EC imports
while changing total market demand. Mediterranean-EC, Middle East/North
Africa, and Latin America show a positive association between total market
size and product demands. The parameter for the Middle East/North Africa
is not significant. Demand for this product is not affected by changes in
the market size. United States and rest of Africa product demands show
negative relationships. Only the one for the United States product is
significant. As market size increases in the EC, consumers will shift
their relative consumption from United States product to Mediterranean-EC
and Latin America products. Mediterranean-EC and Latin America product
have the best position for changes in the market size in this region.
Product demands while changing the import price for Mediterranean-EC
and Latin America are not significant. Import prices may not play a
relevant role for consumer decisions about imports from these regions.
Still, both have positive and highly significant parameters for changes in
the market size. If market size increases, consumers' first and second
choices will definitely be Mediterranean-EC and Latin America products,
respectively. This is an important result, mainly related to trade
agreements between the EC and Mediterranean-EC.

272
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.16. EC Imports Changing Total Market Demand.

273
Rest of Western Europe
Table 6.2 indicate that four regions accounted for 98.7% of total
imports in the rest of Western Europe. The sensitivity analysis is
presented in Figures 6.17 and 6.18. Figure 6.17 shows the relationships
for the import price and product demand indices. The figure indicates
that three of the four relationships have the correct negative direction.
Only the demand for the rest of Africa product has a positive and
significant association (see Table 5.8). The responses of the other three
product demands with negative relationships are very similar, indicating
that consumers in the rest of Western Europe will not shift their
consumption among sources if all prices change proportionally.
Figure 6.18 indicates that the relationships between the total
market demand indices and the product demand indices are positive in all
cases. The parameter for the Middle East/North Africa is not significant,
indicating that demand for this product is not affected by change in
market size. The figure shows that responses are different among regions.
It shows that, if market size increases, consumers will increase their
consumption first from the EC product, second from the Mediterranean-EC,
and third from the rest of Africa.
The results indicate that the rest of Africa import price/product
demand relationship is not significant. The positive and relatively
strong association between product demand and market size gives this
region an opportunity to take advantage of potential market growth.

274
PRODUCT DEMAND INDEX (1986=1)
IMPORT PRICE INDEX (1986 = 1)
Figure 6.17. Rest of Western Europe Imports Changing Import Prices.

275
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.18. Rest of Western Europe Imports Changing Total Market Demand.

276
Middle East/North Africa
As indicated in Table 6.2, 96.9% of total Middle East/North Africa
imports came from the Far East, rest of Africa, Latin America,
Mediterranean-EC, and Oceania. The sensitivity analysis shown in Figure
6.19 presents the Middle East/North Africa product demand index while
changing the import price index. The figure indicates that three of the
five major product demands have negative relationships. Far East and
Oceania parameters are negative and significant (see Table 5.9). Demand
for the rest of Africa product has also a negative association, but it is
not significant. Latin America and Mediterranean-EC have positive but
insignificant relationships. The figure indicates that Far East and
Oceania responses are different to changes in the import price index.
Oceania has a stronger response than the Far East. Demand for Oceania
product in the Middle East/North Africa is more sensitive to changes in
import prices. If Oceania and Far East import prices increase in the same
proportion, consumers in the Middle East will consume more product from
the Far East relative to Oceania product. The reverse happens when import
prices decrease.
Figure 6.20 presents Middle East/North Africa imports while changing
total market demand. The sensitivity analysis indicates that all
responses are positive and strong. The demand for Latin America product
has the strongest response. The second strongest corresponds to the rest
of Africa product. The third, fourth, and fifth places correspond to
Oceania, Mediterranean-EC, and Far East, respectively. The results imply
that, if market size increases, Latin America product becomes the
consumers' first choice. The rest of the regions will also have a

277
PRODUCT DEMAND INDEX (1986=1)
IMPORT PRICE INDEX (1986 = 1)
Figure 6.19. Middle East/North Africa Imports Changing Import Prices.

278
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.20. Middle East/North Africa Imports Changing Total Market
Demand.

279
positive output for market size increases. Latin America and rest of
Africa will probably have the major relative gains. It is worth
mentioning that this market has been growing rapidly in the last two
decades. This establishes an excellent opportunity for exports,
especially from Latin America and rest of Africa.
Rest of Africa
Table 6.2 indicates that 99.6% of rest of Africa total imports came
from five regions: the Middle East/North Africa, EC, Mediterranean-EC,
Latin America, and Oceania. The sensitivity analysis shown in Figure 6.21
presents rest of Africa imports while changing import prices. The figure
indicates that three out of the five regions have negative relationships.
The correct negative associations correspond to the Mediterranean-EC,
Latin America, and Oceania. The demand elasticity for Mediterranean-EC
product is not significant. The results indicate that, if a similar
increase in the import price of Latin America and Oceania products occur,
consumers will tend to consume more product from Latin America relative to
Oceania. Demand elasticities for Middle East/North Africa and EC products
are positive, but only the former is significant. The wrong direction of
the relationship for the Middle East/North Africa product could be related
to the fact that the rest of Africa is self-sufficient and a net exporter
of fresh oranges. It is possible to argue that imports from its major
supplier occur only when domestic supply is insufficient and prices are
rising.
Figure 6.22 shows rest of Africa imports while changing total market
demand. As shown in the figure, only the demand for Oceania product has

280
PRODUCT DEMAND INDEX (1986=1)
IMPORT PRICE INDEX (1986 = 1)
Figure 6.21. Rest of Africa Imports Changing Import Prices.

281
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.22. Rest of Africa Imports Changing Total Market Demand.

282
a negative response to changes in the market size. If the rest of Africa
market size increases, consumers will move away from the Oceania product
to consume more product from the other regions.
Even though the Mediterranean-EC and EC have insignificant import
price/product demand relationships, they have the only significant
parameters for the market size index. Import prices may not affect demand
for these products in the rest of Africa; however, if the market size
increases, consumers will purchase the extra product from these regions
first. This is an important result that is probably related with some
type of trade agreement among the regions.
Far East
As indicated in Table 6.2, 99.IX of Far East imports came from four
regions, the United States, Middle East/North Africa, Oceania, and rest of
Africa. The United States is the major exporter and represented 81.4% of
total imports in the period considered. Figure 6.23 presents Far East
imports while changing the import price index. As shown in the
sensitivity analysis and the empirical results, the demand relationship
for United States product is positive and significant (see Table 5.11).
Given the characteristics of the Far East market and consumers in terms of
fast growth and high-quality products, the positive direction could be
justified. An interpretation of the positive association is not an easy
task. However, considering the high percentage of United States fruit in
the Far East markets, it is possible that import prices do not play a
relevant role for United States-product import decisions in the Far East.
The rest of the regions have negative relationships between import price

283
PRODUCT DEMAND INDEX (1986=1)
IMPORT PRICE INDEX (1986 = 1)
Figure 6.23. Far East Imports Changing Import Prices.

284
and product demand indices. Demand elasticity for the Oceania product is
not significant. If Middle East/North Africa and rest of Africa import
prices go up in similar proportions, consumers will consume relatively
more of the Middle East/North Africa product.
Figure 6.24 presents Far East imports while changing total market
demand. As expected, one of the strongest relationships corresponds to
United States. This implies that consumers are willing to import more
from the United States than from any other region in the world as market
size increases. This fact, and the possibility that the United States
import price may not be a major concern to consumers, give the United
States an interesting position to penetrate the Far East market with fresh
oranges. The rest of Africa also has a strong response to changes in
market size. However, the empirical results indicate that it is not
significant. Oceania has a positive and significant response, but it is
not as strong as the one for the United States. The Middle East/North
Africa response is negative but insignificant. The results show that, if
the Far East market size increases, consumers will increase their
consumption, mainly from the United States and Oceania. This is an
important opportunity for these regions, given that the Far East is one of
the regions with the fastest growth rates in the last two decades.
Oceania
Table 6.2 indicates that three regions accounted for 97.1% of
Oceania's total imports in the period considered. The regions are the
United States, Middle East/North Africa, and Latin America. The United
States is the major exporter and represented 85.9% of total imports. The

285
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.24. Far East Imports Changing Total Market Demand.

286
associations between import price and product demand indices presented in
the sensitivity analysis are negative and significant for United States
and Middle East/North Africa products (see Figure 6.25). Latin America
has an incorrect positive and significant relationship (see Table 5.12).
The direction of this association could be the result of a relatively
small trade between the regions. The strongest negative relationship
corresponds to the Middle East/North Africa. That is, if the United
States and Middle East/North Africa import prices increase proportionally,
consumers will consume relatively more product from the United States. If
import prices decrease, consumers will tend to consume relatively more
from the Middle East/North Africa.
Figure 6.26 and the empirical results indicate that the United
States has an important advantage in Oceania. It is the only region with
a positive and significant relationship between total market demand and
product demand indices. The other regions have significant and highly
negative relationships. If Oceania market size increases, consumers will
import most product from the United States.
Communist Bloc
As indicated by Table 6.2, three regions accounted for 99.4% of the
Communist Bloc imports in the period considered. The regions are the
Middle East/North Africa, Mediterranean-EC, and Latin America. Figure
6.27 presents Communist Bloc imports while changing import prices. The
figure shows that two of the three regions have the correct negative
relationship. The Mediterranean-EC has a positive but insignificant
relationship (see Table 5.13). That is, import prices from this region

287
PRODUCT DEMAND INDEX (1986=1)
IMPORT PRICE INDEX (1986=1)
Figure 6.25. Oceania Imports Changing Import Prices.

288
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.26. Oceania Imports Changing Total Market Demand.

289
PRODUCT DEMAND INDEX (1986=1)
IMPORT PRICE INDEX (1986 = 1)
Figure 6.27. Communist Bloc Imports Changing Import Prices.

290
are not a major factor for import decisions in the Communist Bloc. The
figure also indicates that demand for Latin America product has the
strongest response. If import prices for Latin America and the Middle
East/North Africa increase proportionally, consumers will consume
relatively more product from the Middle East/North Africa than from Latin
America. If import prices decrease, then the reverse is true.
Figure 6.28 shows Communist Bloc imports changing total market
demand. The results indicate that the three product demands have a
positive and significant response to changes in total market demand. The
strongest response is for Latin America product, followed by Middle
East/North Africa product. The response for the Mediterranean-EC is small
but significant. The sensitivity analysis implies that, if the Communist
Bloc market size increases, Latin America product will have the strongest
position to penetrate the market. The Middle East/North Africa will have
the second position and Mediterranean-EC the last one.
Mediterranean-EC has an insignificant import price/product demand
relationship. This condition, combined with the analysis on the market
size, gives this region an advantageous position to penetrate the
Communist Bloc market.
Summary
This section of the chapter discussed the sensitivity analysis
results. Regions were analyzed and compared with other regions. Each
major equation of the model was discussed separately. The results in

291
PRODUCT DEMAND INDEX (1986=1)
TOTAL MARKET DEMAND INDEX (1986=1)
Figure 6.28. Communist Bloc Imports Changing Total Market Demand.

292
each section indicated that domestic and trade-policy decisions can be
enhanced using the information generated in the sensitivity analysis.
Total market demand analysis concluded that if world prices
increase, major importers will consume proportionally less than regions
with low import levels and high local production. If world prices
decrease, importers will consume relatively more than regions with low
import levels. The analysis also shows that the largest response
coincides with the United States, implying that consumers in this market
are highly sensitive to changes in the average market price. If world
prices increase, the percentage adjustment in demand by U.S. consumers
will be greater than the percentage downward adjustment seen in the other
regions. If world prices decrease, United States consumers will consume
more relative to the rest of the regions considered. If a region is
interested in increasing exports to another region, the information
provided in this section of the chapter could be used for price policy
decisions and price discrimination.
Export supply equations show weak FOB average export price
parameters versus very strong and significant fresh production parameters.
This indicates that major export decisions are driven mainly by fresh
production (see Sparks results for vegetables). Since fresh orange
production implies a long-term decision, the results are appealing.
Middle East/North Africa exports have the correct relationship. This is
probably related to the fact that its industry is completely open, while
the Mediterranean-EC has important trade agreements with major partners,
and the United States with Canada. The results indicate that the Middle
East/North Africa is the region with the most flexible relationship

293
regarding production and exports. This region has better opportunities to
take advantage of new or growing markets in the future. However, product
quality has to be improved in order to take advantage of any new
opportunity.
Product demand analysis provides important information for exporters
and importers. The following conclusions will be made from the point of
view of the exporting regions. The United States has good opportunities
to increase and/or maintain its market share in four regions: Canada,
Latin America, Far East, and Oceania. In most regions, the United States
holds a premium position. The most promising region is the Far East,
which is one of the faster growing regions in the world. Latin America
apparently has better opportunities in the EC, Mediterranean-EC, Middle
East/North Africa, and Communist Bloc. Its best market position is found
in the Middle East/North Africa, which is one of the faster growing
regions. Mediterranean-EC has better opportunities in the United States,
EC, rest of Africa, and the Communist Bloc. Its holds a premium position
in all regions, indicating that its fruit quality levels are high and
consumers are willing to pay the price. The Middle East/North Africa has
opportunities in the United States and the Communist Bloc. Its
competitive position in most markets is not good. In most regions, its
market size parameter is negative, probably indicating that quality
standards are poor. The rest of Africa has opportunities in the rest of
Western Europe and Middle East/North Africa. Its position is especially
strong in the rest of Western Europe. The Far East holds a strong second
position in the Canadian market. Oceania has an important opportunity in
the Far East market. Its geographical position is better than its major

294
competitor, the United States, and the Far East market is growing fast.
Finally, the Communist Bloc, in particular Cuba, has an opportunity to
export mostly to Latin America.
The conclusions presented above represent the point of view of the
exporting regions. Conclusions regarding importers will not be presented
here since most of them are included in each region's analysis. In
addition, if an exporting region A is said to have a good market position
in region B, it means that for region B its best choice is A.
The results and the analysis developed could be used by any region
to make policy decisions. The decisions could be related to domestic or
trade policy. Knowledge about the different reactions that consumers,
importers, and exporters have to changes in market price, income, the FOB
average export price, fresh production, the relative import price, and
market size is valuable market information. The results could be used to
evaluate policy decisions about price, price discrimination, the
competition level, the potential market, market development and growth,
the impact of trade barriers, and other important factors.
Conclusion
The objective of the present chapter was to develop a framework to
perform a comparative static analysis of the estimated parameters. A
sensitivity analysis procedure was developed in such a way that the
results could be evaluated and compared among regions.
The purpose of the analysis was to assess the impact in the
dependent variables given changes in the right-hand side variables. Major

295
consumers, importers, and traders were included in the analysis. The
results of the sensitivity analysis were used to develop a simple
graphical framework to study the different markets' behavior and compare
them among regions.
The results and the analyses turn out to provide relevant
information for every participating region in the fresh orange trade
model. The discussion also provided additional information to complement
the more technical analysis developed in Chapter 5.

CHAPTER 7
SUMMARY AND CONCLUSIONS
Introduction
The present study developed a fresh orange trade model to study the
major factors affecting exporter and consumer decisions in 11 regions of
the world. The specific objectives were: to specify a multiple-region
equilibrium world trade model for the fresh orange industry; to analyze
the implications contained in the estimated model; to use the estimated
parameters to study analytically the reasons for changes in market shares;
and to develop a sensitivity analysis under different economic scenarios
to make contributions to specific policy issues.
Even though the fresh orange market has experienced important
growth, several countries, including the United States, have experienced
pronounced changes in their trade patterns. The fresh orange industry is
of enormous importance for some regions, especially for the United States,
South America, Mediterranean-EEC, Middle East/North Africa, and Far East,
as producers, consumers, and exporters. Producers and exporters need to
understand the major driving factors for fresh consumption and their
competitive position in foreign markets. Such information will allow them
to compete with its benefits, possibly achieve international success, and
help to develop new markets. This industry is also important for net
importers such as Canada, EEC, rest of Western Europe, and the Communist
296

297
Bloc. These regions are interested in knowing which are the major driving
factors for fresh consumption, and demand and price linkages between the
region and its major trading partners.
Studying the fresh orange trade flows and modeling these changes
provided information to help understand the reasons for changes in market
shares among major suppliers and facilitate longer term forecasts and
policy analyses. To accomplish the objectives of the present study,
international trade linkages among the major trading regions were
identified. It was also necessary to recognize the current and emerging
problems in the industry. This information was helpful in studying the
changes in trade patterns arising from changes in supply and demand
conditions, and from changes in policy variables such as tariff levels and
institutional constraints.
The dissertation includes six chapters. Chapter 1 presents a
discussion related to the importance of world trade and, in particular,
agricultural trade. It also presents a discussion about the orange
industry including fresh and processed oranges. The chapter concludes
that it is important to many countries and regions to study the trade
flows and market shares of the fresh orange industry from a world
perspective. Chapter 2 presents a discussion about world production and
trade flows for the 11 regions selected. Trade volumes by region and
partner region are discussed from 1966 to 1986. The chapter provides
important insights about trade-flow and market-share changes in the
different regions for the period considered. Chapter 3 presents a
literature review regarding trade models. Agricultural trade models and,
in particular, fresh and processed orange trade models are covered.

298
Chapter 4 develops the fresh orange trade model used in the present study.
The theoretical background and the empirical model to be estimated are
presented. Chapter 5 discusses the methods used for the estimation of the
model. It also develops graphical, statistical, and economic analyses to
study the performance of the model and the implications of the results.
Chapter 6 presents a sensitivity analysis to study the changes in the
total market demands, export supplies, and product demands, given changes
in the right-hand side variables.
Data Limitations
The data required for this model has serious limitations. It is
necessary to have all trade flows, import and export value, and quantity
for every country of the world, showing the partner country. The data are
then aggregated by region. If all countries of the world are included,
the data are not available except from the United Nations trade data
tapes. These data are gathered by each member country and sent to the
statistics office in New York. The price data used in this dissertation
are unit prices obtained by dividing value by quantity for each trade
flow. As expected in trade data, many errors were found. Most of them
were probably related with gathering problems and inconsistencies. Where
errors were detected, the data were corrected in what was believed to be
the most appropriate way.
Tariff barriers for fresh oranges were not available in a single
document for all countries. It was necessary to review many different
sources to obtain the final data presented in Appendix E. Tariffs of the

299
individual countries were averaged, using different methods to obtain the
regional tariff. Nontariff barriers were not considered in the study,
given that most of them are seasonal and the model uses annual data.
Fresh and processed orange utilization was not available for most
countries. It was necessary to do a detailed literature review, including
government reports, books, magazines, other publications and personal
contacts to obtain the necessary information for each country included in
the study. The information was then aggregated by region.
The regional CPIs based on Edwards and Ng (1985) theory were not
available for the regions considered. It was necessary to create the data
set for each country and region. Appendix H presents the detailed
procedure utilized to get the final numbers. The first step was to obtain
the domestic CPIs or inflation rates and the exchange rate index per
country. The domestic CPIs were divided by the exchange rate index to get
the CPIs per country. Finally, the CPIs per country were weighted using
the 1986 trade volume to obtain the regional CPIs.
Estimation and Sensitivity Analysis Difficulties
A nonlinear two stage least squares procedure was utilized for the
estimation of the model. While the model is simultaneous and large, it
was still possible to estimate the model by sections in a Personal
Computer, using TSP. Estimation capabilities have improved considerably
in the last two years, thus greatly facilitating the use of the personal
computer.

300
The size of the fresh orange trade model developed here, with 440
equations of which 242 were estimated, leaves little space to improve
individual equations by correcting the functional form, the variables
included, or any other alternative solution. Large trade econometric
models like the one developed here are used to provide information about
major trends and shifts of trade flows and market shares through the years
among the different regions. The model provided important information
about the behavior of the fresh orange industry. This information could
be used for policy decisions in the different regions and countries.
If a particular trade flow is of interest and more information is
needed, it is possible to review the particular functional form and obtain
better results. However, if a single-equation estimation procedure is
used, the results suffer from simultaneity bias.
An important limitation of the present study is that it was not
possible to obtain the reduced-form parameters. If the reduced-form
parameters were found, then the whole system of equations could have been
simulated, given changes in the exogenous variables. Given the
simultaneity embodied in the model, this is an important drawback. The
limitation implies that the sensitivity analysis has to be developed on an
equation by equation basis.
Performance of the Model and Results
The graphical and the statistical analyses provided sufficient
information to determine that the model has a good fit, is well specified
and predicted most turning points. The economic analysis shows that the

301
signs and magnitudes of the estimated parameter meet economic expectations
in a majority of cases.
The model consists of 11 regions, including all countries of the
world. It was found that trade is concentrated in a few regions. The
performance of the model is better where significant trade took place.
However, regions with small participation have important growing export or
import markets; reducing the size of the model will hide important
information and opportunities for some regions.
The analysis of the demand parameters showed the likely future
direction of trade. Price elasticities were used to predict responses in
the different markets to changes in prices. The role of prices as an
allocative tool was shown. Income and population elasticities gave an
indication of possible adjustments in consumption and trade patterns.
Fresh production was found to be the most important factor contributing to
world exports. Relative import price and market size were found to be
important product demand drivers for most trade flows.
The model made forecasts of trade patterns among importers and
exporters possible. The model was used to construct a sensitivity
analysis to predict and compare total market demands, export supplies, and
product demands responses among regions. Simulations were completed
giving shocks in the different variables including average market price,
income, relative prices, market size, FOB average export price and fresh
production.

302
Contributions to Agricultural Economics Research
The dissertation represents the first multiple-region world trade
model for the fresh-orange industry. The study provides a conceptual
framework and model which can be used for international trade research on
other individual agricultural products. The model is a modified spatial
equilibrium model that follows Armington's demand theory that products are
differentiated by place of origin. The model is a revised version of the
Armington model, which is more flexible and capable of predicting most
trade flows and market shares accurately. There has been only one other
study that used a similar model (Sparks, 1987); however, in that case the
model was used to study a highly aggregated commodity, fresh vegetables.
This is the first time that the model has been applied to an individual
good, which is more appealing, given that aggregated goods are difficult
to differentiate.
Exchange rates are explicitly included and uniquely introduced in
the present study for this type of model. The use of the United States
CPI, instead of the regional CPIs, implies the assumption of purchasing
power parity in all regions. The model utilized regional CPIs to obtain
real prices and income.
The model was estimated using a simultaneous system of equations.
There have been only two other studies which estimated this type of model
in a simultaneous system (Deardorff and Stern, 1986; Sparks, 1987).
Finally, the estimation procedure uniquely introduced different nonlinear
relationships for the first step of the principal components procedure.

303
Further Research
There are many areas to which future research could be directed.
The first would be to use the conceptual framework and model developed
here in a different agricultural product. The changes to be made are
minor and, obviously, related to the individual characteristics and trade
patterns of the product selected.
Another interesting area of research would be to work with the fresh
orange industry, modifying the model by reconsidering the number of
regions and the country composition. It is also important to investigate
and evaluate alternative functional forms for some of the equations. This
would represent a tremendous amount of work, but it would probably provide
a better model that could be used for many different products. It is
important to recognize the significant changes in Eastern Europe which may
affect some of the conclusions of the present study.
It was not possible to obtain the reduced form of the fresh orange
trade model. It is important that future research pursue the possibility
of obtaining the reduced-form parameters of the model. The procedure
developed will be useful in many ways, since the same model can be used
for other products.
The results of the present study suggest the presence of some
specification problems. Specification tests other than the Durwin Watson
where not conducted. An interesting area for future research will be to
apply specification tests for this particular model and evaluate and
measure the specification errors properly. Models such as used here with
large number of equations do not lend themselves to certain types of test.

304
Furthermore, it is almost impossible to make some of the corrections that
might be suggested by the specific test because of the interrelationship
among the equations.
During the estimation process, the model was first estimated
assuming purchasing power parity. The use of the work developed by
Edwards and Ng (1985) improved the results of the model. Given that trade
models usually utilize the United States CPI, trade research could benefit
from this finding. More research is necessary to assure that world
trade models will actually improve by using the CPIs per country, instead
of using the United States' CPI.
One must constantly be aware of the inheritant data limitations and
reporting problems using world trade data. Even so, this analysis shows
that such data can be successfully used in modeling while recognizing the
limitations.

APPENDIX A
COUNTRY COMPOSITION OF THE REGIONS
United States (US): United States.
Canada (CAN): Canada.
Latin America (LA): Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador,
Mexico, Paraguay, Peru, Uruguay, Venezuela, Costa Rica, El Salvador,
Guatemala, Honduras, Nicaragua, Antigua, Bahamas, Barbados, Dominica,
Dominican Republic, Grenada, Guadeloupe, Haiti, Jamaica, Netherlands-
Antilles, Saint Lucia, Saint Vincent, Trinidad Tobago, Belize, Guyana,
Panama, Panama Canal, Suriname.
Mediterranean-EC (MED-EO: Spain, Italy, Portugal and Greece.
EC: Belgium-Luxembourg, Denmark, France, West Germany, Ireland,
Netherlands, United Kingdom.
Rest of Western Europe (RWE): Austria, Finland, Iceland, Norway, Sweden,
Switzerland, Malta.
Middle East/North Africa (ME/NA): Algeria, Liby Arab JM, Morocco, Sudan,
Tunisia, Egypt, Israel, Bahrain, Cyprus, Iran, Iraq, Jordan, Kuwait,
Lebanon, Oman, Qatar, Saudi Arabia, Dem. Yemen, Syrian Arab RP, United
Arab EM, Turkey, Yemen AR, Afghanistan.
Rest of Africa (RAF): South Africa, Cameroon, Central Africa REP., Chad,
Congo, Gabon, Burundi, Cape Verde, Comoros, Zaire, Benin, Ethiopia,
Djibouti, Gambia, Ghana, Cote Divoire, Kenya, Liberia, Madagascar, Malawi,
Mali, Mauritania, Mauritius, Niger, Nigeria, Reunion, Rwanda, Sao Tome
305

306
Prn., Senegal, Seychelles, Sierra Leone, Somalia, Zimbabwe, Togo, Uganda,
United RP. Tanzania, Burkina Faso, Zambia.
Far East (FE~) : Japan, Bangladesh, Burma, Sri Lanka, Hong Kong, India,
Indonesia, Korea Republic, Malaysia, Maldives, Nepal, Pakistan,
Philippines, East Timor, Singapore, Thailand, Vietnam, China.
Oceania (OCE') : Australia, New Zealand, Solomon Islands, Fiji, New
Caledonia, Papua New Guinea, Samoa.
Communist Bloc (COMMB): Yugoslavia, Albania, Bulgaria, Czechoslovakia,
East Germany, Hungary, Poland, Romania, USSR, Cuba.

APPENDIX B
DERIVATION OF THE PRODUCT DEMAND EQUATIONS
If the market demand equations follow the CRES quantity index
function of the product demands, then:
(B.l) Xt. = [b^X,,*11 + bi2*Xi2ai2 + ... + bim*Ximaim](1/Ql-)
Defining the term in parenthesis as Q, the market demand equation can be
written as follows:
(1/a, )
(B. 2) X,. = Q' *'
Taking the partial derivative of the market demand (X, ) w.r.t. the product
demands (X,j) the following result is obtained:
(B. 3) (Xi.VaCXij) = [l/ai.]*[Q(1/ai-)1]*[a(Q)/3(XiJ)]
= [ 1/a,. ] [X,. *X,. 'ai ] [aij*bij*Xij (aiJ'1} ]
Equation (4) follows from the first order condition of utility
maximization:
(B.4) aXi.vaXy) P,. = Pij
which implies that
(b.5) p,. = Pij/tacxo/acXij)]
by substituting a(X, )/a(X,j) in (5), equation (6) holds:
(B. 6) P,. = Pij/f(l/ai.)*(Xi.'Qi-*Xi.)*(aij*bij*Xij(aiJ'1))]
Rearranging terms and solving for the product demands (X^), equations (7)
and (8) follow:
(B 7) X^'1 = [Oi./Co^by)] [Pij/P,.] tX1.(0iJ'1)]
307

308
(B.8) Xy [>)
*[xi.(i'1,/]
Equation 4.24 in the text is the same as B.l. Equations 4.25 and
4.31 in the text follow directly from equation B.8.

APPENDIX C
PROCEDURE TO OBTAIN REGIONAL CPIs
The procedure developed by Edwards and Ng (1985) to obtain the
regional CPIs (Consumer Price Indices) is the following:
1.- Get the percentage change of the CPIs per country (annual inflation)
2.- Get the exchange rates with respect to the U.S. dollar per country
3.- Get an index of the exchange rate for a base year
4.- Divide the CPIs by the exchange rate index to obtain the CPIs by
country
5.- The individual country's CPIs are weighed using trade levels to obtain
the regional CPIs aggregate values
309

1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
rce:
APPENDIX D
PROCESSED ORANGE UTILIZATION
us
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
COM
66.1
0.0
4.5
12.3
PERCENTAGE
0.0 0.0
6.7
10.5
5.5
18.7
0.0
71.5
0.0
4.4
13.4
0.0
0.0
8.2
10.0
5.8
21.5
0.0
71.3
0.0
8.2
14.8
0.0
0.0
10.0
12.0
8.0
24.7
0.0
74.0
0.0
8.0
14.8
0.0
0.0
10.5
10.7
8.8
28.4
0.0
74.4
0.0
10.6
14.8
0.0
0.0
16.0
16.5
7.5
32.7
0.0
75.0
0.0
15.1
16.4
0.0
0.0
17.3
14.0
7.5
23.3
0.0
75.5
0.0
23.1
15.6
0.0
0.0
13.8
16.2
10.4
27.9
0.0
79.6
0.0
27.1
12.0
0.0
0.0
16.2
15.6
11.9
32.3
0.0
77.4
0.0
28.2
12.3
0.0
0.0
13.4
14.8
6.3
36.7
0.0
75.4
0.0
27.3
12.1
0.0
0.0
10.3
13.9
12.9
41.2
0.0
77.5
0.0
31.2
12.2
0.0
0.0
8.4
13.6
13.8
45.5
1.7
78.8
0.0
30.8
12.9
0.0
0.0
9.3
12.1
14.7
51.6
2.5
78.1
0.0
37.9
12.7
0.0
0.0
7.8
13.4
12.4
50.6
2.5
79.1
0.0
38.2
12.6
0.0
0.0
8.9
15.1
14.9
50.7
1.8
78.6
0.0
40.8
13.4
0.0
0.0
8.8
13.3
14.4
57.4
3.1
78.1
0.0
43.5
16.8
0.0
0.0
12.0
16.5
8.5
45.8
1.4
72.4
0.0
46.3
13.1
0.0
0.0
9.5
9.4
10.4
52.9
1.5
72.2
0.0
45.4
13.1
0.0
0.0
9.8
6.4
8.0 ,
53.4
3.5
70.1
0.0
48.4
19.4
0.0
0.0
12.1
10.2
4.4
54.0
5.2
68.1
0.0
54.6
14.4
0.0
0.0
12.8
14.0
8.4
60.0
11.5
67.7
0.0
49.3
15.4
0.0
0.0
11.6
10.1
8.8
56.2
12.4
Prepared from different sources.
310

APPENDIX E
TARIFF DATA
Region
US
CAN
LA
MED-EC
EC
RWE
ME/NA
RAF
FE
OCE
COttdB

X OF FOB EXPORT
PRICE
-
USa
0
22.05
22.05
22.05
22.05
22.05
22.05
22.05
22.05
22.05
22.05
CAN
0
0
0
0
0
0
0
0
0
0
0
SA
25
25
25
25
25
25
25
25
25
25
25
MED-EECbc
66
12
12
12
10
2.8
12
6.7
12
12
12
12
67
12
12
12
10
2.8
12
6.7
12
12
12
12
68
12
12
12
10
2.8
12
6.7
12
12
12
12
69
12
12
12
10
2.8
12
6.7
12
12
12
12
70
12
12
12
10
2.8
12
6.7
12
12
12
12
71
12
12
12
10
2.8
12
6.7
12
12
12
12
72
12
12
12
10
2.8
12
6.7
12
12
12
12
73
12
12
12
8.2
2.8
12
6.39
12
12
12
12
74
12
12
12
8.2
2.8
12
6.39
12
12
12
12
75
12
12
12
8.2
2.8
12
6.39
12
12
12
12
76
12
12
12
8.2
2.8
12
6.39
12
12
12
12
77
12
12
12
8.2
2.8
12
6.39
12
12
12
12
78
12
12
12
8.2
2.8
12
6.39
12
12
12
12
79
12
12
12
8.2
2.8
12
6.19
12
12
12
12
80
12
12
12
8.2
2.8
12
6.19
12
12
12
12
81
12
12
12
8.2
2.8
12
6.19
12
12
12
12
82
12
12
12
8.2
2.8
12
6.19
12
12
12
12
83
12
12
12
8.2
2.8
12
6.19
12
12
12
12
84
12
12
12
8.2
2.8
12
6.19
12
12
12
12
85
12
12
12
8.2
2.8
12
6.19
12
12
12
12
86
12
12
12
8.2
2.8
12
6.19
12
12
12
12
EECbc
66
15
15
15
11.43
0
15
6.5
15
15
15
15
67
15
15
15
11.43
0
15
6.5
15
15
15
15
68
15
15
15
11.43
0
15
6.5
15
15
15
15
69
15
15
15
11.43
0
15
6.5
15
15
15
15
70
15
15
15
11.43
0
15
6.5
15
15
15
15
71
15
15
15
11.43
0
15
6.5
15
15
15
15
72
15
15
15
11.43
0
15
6.5
15
15
15
15
73
15
15
15
8.59
0
15
5.98
15
15
15
15
74
15
15
15
8.59
0
15
5.98
15
15
15
15
75
15
15
15
8.59
0
15
5.98
15
15
15
15
76
15
15
15
8.59
0
15
5.98
15
15
15
15
77
15
15
15
8.59
0
15
5.98
15
15
15
15
78
15
15
15
8.59
0
15
5.98
15
15
15
15
79
15
15
15
8.76
0
15
5.65
15
15
15
15
80
15
15
15
8.76
0
15
5.65
15
15
15
15
81
15
15
15
8.76
0
15
5.65
15
15
15
15
82
15
15
15
8.76
0
15
5.65
15
15
15
15
83
15
15
15
8.76
0
15
5.65
15
15
15
15
84
15
15
15
8.76
0
15
5.65
15
15
15
15
85
15
15
15
8.76
0
15
5.65
15
15
15
15
86
15
15
15
8.76
0
15
5.65
15
15
15
15
RWE
0
0
0
0
0
0
0
0
0
0
0
ME/NA
5
5
5
5
5
5
5
5
5
5
5
RAF
5
5
5
5
5
5
5
5
5
5
5
FE
40
40
20
40
40
40
40
20
40
20
40
OCE
0
0
0
0
0
0
0
0
0
0
0
COMMB
10
10
2.5
10
10
10
2.5
10
10
10
0
U.S. dollars per metric ton.
bZ of CIF import price and tariffs vary by year.
cTariffs differ by year.
311

APPENDIX F
PRINCIPAL COMPONENT PROCEDURE AND PROGRAM
FREQ A;
SMPL 66,86;
READ (FORMAT-LOTUS FILE-' C: \LOTUS\DATA. WK1') ;
? START OF PROGRAM;
LIST EXOG
POPI GDP1
PRD1 CPU YEAR
PEN BAVAL1
POP2 GDP2
PRD2 CPI2
BAVAL2
POP3 GDP3
PRD3 CPI3
BAVAL3
P0P4 GDP4
PRD4 CPI4
BAVAL4
P0P5 GDP5
PRD5 CPI5
BAVAL5
P0P6 GDP6
PRD6 CPI6
BAVAL6
P0P7 GDP7
PRD7 CPI7
BAVAL7
P0P8 GDP8
PRD8 CPI8
BAVAL8
P0P9 GDP9
PRD9 CPI9
BAVAL9
POPIO GDP10
PRD10 CPI10
BAVAL10
P0P11 GDP11
PRD11 CPI11
BAVAL11;
PRIN (NAME-PC.NCOM-6,FRAC-,98.NOPRINT) EXOG;
PRINT PCI PC2 PC3 PC4 PC5 PC6;
END;
The exogenous variables included in the principal component procedure are
Population (POP), Gross Domestic Product (GDP), Production (PRD), Consumer
Price Index (CPI), and Bananas and Apples Price Index per region (BAVAL).
The Year Trend (YEAR) and the Price Index for Energy (PEN) were also
included. PCI to PC6 refer to the principal components obtained with the
procedure.
312

APPENDIX G
ESTIMATION PROGRAM
The Program #1 below was used to estimate the market demand and
export supply equations. The Program #2 was used to estimate the product
demand equations for the United States. Since all regional programs are
similar to the one presented for the United States, they will not be
included here. The only differences among the regional programs are the
variables used and parameters names.
Program #1
FREQ A;
SMPL 66,86;
READ (FORMAT-LOTUS,FILE-'C:\LOTUS\DATA.WK1');
? START OF PROGRAM;
LIST
EXOG;
POPI
GDP1
PRD1
CPU YEAR
PEN BAVAL1
POP2
GDP2
PRD 2
CPI2
BAVAL2
POP3
GDP3
PRD 3
CPI3
BAVAL3
P0P4
GDP4
PRD4
CPI4
BAVAL4
P0P5
GDP5
PRD 5
CPI5
BAVAL5
P0P6
GDP6
PRD6
CPI6
BAVAL6
P0P7
GDP7
PRD 7
CPI7
BAVAL7
P0P8
GDP8
PRD 8
CPI8
BAVAL8
P0P9
GDP9
PRD 9
CPI9
BAVAL9
POPIO GDP10
PRD10 CPI10
BAVAL10
P0P11 GDP11
PRD11 CPI11
BAVAL11
PRIN (NAMEPC,NCOM-6,FRAC.98,NOPRINT) EXOG;
? Alternative #1
PCI-.96**PC1;
PC2-.96**PC2;
PC3-.96**PC3;
PC4-.96**PC4;
? Alternative #2
PCI-PCI;
PC2-PC2;
PC3-PC3;
PC4=PC1*PC1;
PC5=PC2*PC2;
PC6=PC3*PC3;
? Alternative #3
PC1-.96**PC1;
PC2-.96**PC2;
PC3-.96**PC3;
PC4-.96**PC4;
PC5-.96**PC5;
PC6-.96**PC6;
313

314
LEXQD1
LEXQD2
LEXQD3
LEXQD4
LEXQD5
LEXQD6
LEXQD7
LEXQD8
LEXQD9
LEXQD10
LEXQD11
LOG(EXPORT1)
LOG(EXPORT2)
LOG(EXPORT3)
L0G(EXP0RT4)
LOG(EXPORT5)
LOG(EXPORT)
LOG(EXPORT7)
LOG(EXPORT8)
LOG(EXPORT9)
LOG(EXPORTIO)
LOG(EXPORTll)
LIQ1D -
LIQ2D -
LIQ3D -
LIQ4D -
LIQ5D -
LIQ6D -
LIQ7D -
LIQ8D =
LIQ9D -
; LIQ10D
; LIQ11D
LOG(IQID)
LOG(IQ2D)
LOG(IQ3D)
L0G(IQ4D)
LOG(IQ5D)
LOG(IQ6D)
LOG(IQ7D)
LOG(IQ8D)
LOG(IQ9D)
LOG(IQIOD);
LOG(IQllD);
PARAM
RHOl
1
RH11
-1
RH21
1
RH31
1
DHOl
1
DH11
1
DH21
1
RH02
1
RH12
-1
RH22
1
RH32
1
DH02
1
DH12
1
DH22
1
RH03
1
RH13
-1
RH23
1
RH33
1
DH03
1
DH13
1
DH23
1
RH04
1
RH14
-1
RH24
1
RH34
1
DH04
1
DH14
1
DH24
1
RH05
1
RH15
-1
RH25
1
RH35
1
DH05
1
DH15
1
DH25
1
RH06
1
RH16
-1
RH26
1
RH36
1
DH06
1
DH16
1
DH26
1
RH07
1
RH17
-1
RH27
1
RH37
1
DH07
1
DH17
1
DH27
1
RH08
1
RH18
-1
RH28
1
RH38
1
DH08
1
DH18
1
DH28
1
RH09
1
RH19
-1
RH29
1
RH39
1
DH09
1
DH19
1
DH29
1
RH010 1 RH110 -1 RH210 1 RH310 1 DH010 1 DH110 1 DH210 1
RH011 1 RH111 -1 RH211 1 RH311 1 DH011 1 DH111 1 DH211 1;
PARAM
RH41 1 RH51 1 RH47 1 RH57 1
RH42 1 RH52 1 RH48 1 RH58 1
RH43 1 RH53 1 RH49 1 RH59 1
RH44 1 RH54 1 RH410 1 RH510 1
RH45 1 RH55 1 RH411 1 RH511 1
RH46 1 RH56 1;
REPD1 EPD1/CPI1;LREPD1
REPD2 EPD2/CPI2;LREPD2
REPD3 = EPD3/CPI3;LREPD3
REPD4 EPD4/CPI4;LREPD4
REPD5 = EPD5/CPI5;LREPD5
REPD6 = EPD6/CPI6;LREPD6
REPD7 = EPD7/CPI7;LREPD7
REPD8 EPD8/CPI8;LREPD8
REPD9 EPD9/CPI9;LREPD9
REPDIO EPDIO/CPIIO;LREPDIO
REPD11 = EPD11/CPI11;LREPD11
LOG(REPDl)
LOG(REPD2)
LOG(REPD3)
L0G(REPD4)
LOG(REPD5)
LOG(REPD)
LOG(REPD7)
LOG(REPD8)
LOG(REPD9)
LOG(REPDIO);
LOG(REPD11);
OLSQ LREPD1,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6
LREPD1H=@FIT
OLSQ
LREPD2,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6
LREPD2HH§FIT
OLSQ
LREPD3,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6
LREPD3HH§FIT
OLSQ
LREPD4,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6
LREPD4HH3FIT
OLSQ
LREPD5,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6
LREPD5H= OLSQ
LREPD6,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6
LREPD6H=@FIT
OLSQ
LREPD7,
C,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6
LREPD7H=@FIT

OLSQ LREPD8, C, PCI, PC2, PC3, PC4, PC5, PC6; LREPD8H@FIT;
OLSQ LREPD9, C, PCI, PC2, PC3, PC4, PC5, PC6; LREPD9H-@FIT;
OLSQ LREPD10, C, PCI, PC2, PC3, PC4, PC5, PC6; LREPD10HH§FIT;
OLSQ LREPD11, C, PCI, PC2 PC3 PC4, PC5, PC6 ; LREPDllH= RMP1D MP1D/CPI1;LRMP1D LOG(RMPID);
RMP2D MP2D/CPI2;LRMP2D LOG(RMP2D);
RMP3D MP3D/CPI3;LRMP3D LOG(RMP3D);
RMP4D = MP4D/CPI4;LRMP4D LOG(RMP4D);
RMP5D MP5D/CPI5;LRMP5D LOG(RMP5D);
RMP6D MP6D/CPI6;LRMP6D LOG(RMP6D);
RMP7D = MP7D/CPI7;LRMP7D = LOG(RMP7D);
RMP8D MP8D/CPI8;LRMP8D LOG(RMP8D);
RMP9D MP9D/CPI9;LRMP9D LOG(RMP9D);
RMPIOD MP10D/CPI10;LRMP10D LOG(RMPIOD);
RMP11D MP11D/CPI11;LRMP11D LOG(RMPllD);
OLSQ LRMP1D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP1DH=@FIT
OLSQ LRMP2D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP2DH=@FIT
OLSQ LRMP3D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP3DH=@FIT
OLSQ LRMP4D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP4DH=@FIT
OLSQ LRMP5D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP5DH=@FIT
OLSQ LRMP6D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP6DH= OLSQ LRMP7D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP7DH=@FIT
OLSQ LRMP8D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP8DHH2FIT
OLSQ LRMP9D,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LRMP9DH=@FIT
OLSQ LRMP10D,
c,
PCI,
PC2
PC3
PC4
PC5,
PC6
LRMP10DH=@FIT
OLSQ LRMP11D,
c,
PCI,
PC2
PC3
PC4
PC5,
PC6
LRMP11DH=@FIT
FRML EQiy/19 LEXQD1 (DH01 + DH11*(LREPD1H) + DH21*LOG(PRDl));
FRML EQ2#19 LEXQD2 = (DH02 + DH12*(LREPD2H) + DH22*LOG(PRD2));
FRML EQ3//19 LEXQD3 = (DH03 + DH13*(LREPD3H) + DH23*LOG(PRD3));
FRML EQ4#19 LEXQD4 = (DH04 + DH14*(LREPD4H) + DH24*LOG(PRD4));
FRML EQ5#19 LEXQD5 = (DH05 + DH15*(LREPD5H) + DH25*LOG(PRD5));
FRML EQ6#19 LEXQD6 (DH06 + DH16*(LREPD6H) + DH26*LOG(PRD6));
FRML EQ7#19 LEXQD7 (DH07 + DH17*(LREPD7H) + DH27*LOG(PRD7));
FRML EQ8//19 LEXQD8 (DH08 + DH18*(LREPD8H) + DH28*LOG(PRD8)) ;
FRML EQ9#19 LEXQD9 (DH09 + DH19*(LREPD9H) + DH29*LOG(PRD9));
FRML EQ10//19 LEXQD10 (DHOIO + DH110*(LREPD10H) + DH210*LOG(PRD10));
FRML EQliy/19 LEXQD11 (DH011 + DH111*(LREPD11H) + DH211*LOG(PRDll));
FRML EQ1#20 LIQ1D = (RHOl + RH11*(LRMP1DH) + RH21*L0G(GDP1/CPI1)
+ RH3l*LOG(POPI) + RH41*L0G(BAVAL1/CPI1));
FRML EQ2#20 LIQ2D (RH02 + RH12*(LRMP2DH) + RH22*LOG(GDP2/CPI2)
+ RH32*LOG(POP2) + RH42*LOG(BAVAL2/CPI2));
FRML EQ3#20 LIQ3D (RH03 + RH13*(LRMP3DH) + RH23*LOG(GDP3/CPI3)
+ RH33*LOG(POP3) + RH43*LOG(BAVAL3/CPI3));
FRML EQ4//20 LIQ4D = (RH04 + RH14*(LRMP4DH) + RH24*LOG(GDP4/CPI4)
+ RH34*LOG(POP4) + RH44*L0G(BAVAL4/CPI4));
FRML EQ5#20 LIQ5D (RH05 + RH15*(LRMP5DH) + RH25*LOG(GDP5/CPI5)
+ RH35*LOG(POP5) + RH45*LOG(BAVAL5/CPI5));

316
FRML EQ6#20 LIQ6D (RH06 + RH16*(LRMP6DH) + RH26*LOG(GDP6/CPI6)
+ RH36*LOG(POP6) + RH46*L0G(BAVAL6/CPI6));
FRML EQ7//20 LIQ7D (RH07 + RH17*(LRMP7DH) + RH27*LOG(GDP7/CPI7)
+ RH37*LOG(POP7) + RH47*L0G(BAVAL7/CPI7)) ;
FRML EQ8//20 LIQ8D (RH08 + RH18*(LRMP8DH) + RH28*LOG(GDP8/CPI8)
+ RH38*LOG(POP8) + RH48*LOG(BAVAL8/CPI8));
FRML EQ9#20 LIQ9D = (RH09 + RH19*(LRMP9DH) + RH29*LOG(GDP9/CPI9)
+ RH39*LOG(POP9) + RH49*LOG(BAVAL9/CPI9));
FRML EQ10#20 LIQ10D = (RHOIO + RH110*(LRMP10DH) + RH210*LOG(GDP10/CPI10)
+ RH310*LOG(POPIO) + RH410*LOG(BAVAL10/CPI10));
FRML EQ11#20 LIQ11D (RHOll + RH111*(LRMP11DH) + RH211*L0G(GDP11/CPI11)
+ RH311*LOG(POPll) + RH411*L0G(BAVAL11/CPI11));
(NOPRINT,
(NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
(NOPRINT,
(NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ
LSQ
LSQ
LSQ
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
EQ1#19;
EQ2#19;
EQ3#19;
EQ4//19;
EQ5//19;
EQ6#19;
EQ7//19;
EQ8#19;
EQ9#19;
EQ10y/19
EQliy/19
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
¡PRINT @RSQ
¡PRINT @RSQ
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW
@DW
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
, @FST;
, @FST;
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
LSQ (NOPRINT,
END;
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
SILENT)
EQiy/20;
EQ2#20;
EQ3#20;
EQ4#20;
EQ5#20;
EQ6#20;
EQ7//20;
EQ8#20;
EQ9y/20;
EQ10#20
EQliy/20
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT (3RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
PRINT @RSQ,
¡PRINT @RSQ,
¡PRINT @RSQ,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@DW,
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
@FST;
Program y/2
? REGION 1;
FREQ A;
SMPL 66,86;
READ (FORMAT=LOTUS,FILE='C:\LOTUS\DATA.WK1');
? START OF PROGRAM;
LIST EXOG;
POPI GDP1 PRD1 CPU YEAR PEN BAVAL1;
POP2 GDP2 PRD2 CPI2 BAVAL2;
POP3 GDP3 PRD3 CPI3 BAVAL3;
P0P4 GDP4 PRD4 CPI4 BAVAL4;

P0P5 GDP5 PRD5 CPI5 BAVAL5;
P0P6 GDP6 PRD6 CPI6 BAVAL6;
P0P7 GDP7 PRD7 CPI7 BAVAL7;
P0P8 GDP8 PRD8 CPI8 BAVAL8;
P0P9 GDP9 PRD9 CPI9 BAVAL9;
POPIO GDPIO PRD10 CPIIO BAVALIO;
POP11 GDP11 PRD11 CPI11 BAVAL11;
PRIN (NAME-PC,NCOM-6,FRAC-.98,NOPRINT) EXOG;
? Alternative #1
PC1=.96**PC1;
PC2-.96**PC2;
PC3-.96**PC3;
PC4-.96**PC4;
? Alternative #2
PCI-PCI;
PC2-PC2;
PC3-PC3;
PC4=PC1*PC1;
PC5-PC2*PC2;
PC6=PC3*PC3;
? Alternative #3
PCI-.96**PC1;
PC2-.96**PC2;
PC3-.96**PC3;
PC4-.96**PC4;
PC5-.96**PC5;
PC6-.96**PC6;
LIQ1D = LOG(IQ1D)
LIQ1_2 =
LIQ1_3
LIQ1_4 -
LIQ1_5 =
LIQ1_6 =
LIQl_7 -
LIQ1_8 -
LIQ1_9 -
LIQ1_10
LIQ1_11
LOG(IQl_:
LOG(IQl_
LOG(IQl_
LOG (IQ1_:
LOG (IQ1_J
LOG(IQl
LOG(IQl_
L0G(IQ1J
= LOG(IQ1
= LOG(IQ1
LIP1_2 = LOG(IPl_2);
2)
3)
A)
5)
6)
7)
8)
9)
LIP1_3
LIP1_4
LIP1_5
LIP1_6
LIP1_7
LIP1_8
LIP1_9
_10); LIP1
11); LIPl"
- LOG(IPl_3)
- L0G(IP1_4)
= LOG(IPl_5)
= LOG(IPl_6)
- LOG(IPl_7)
- LOG(IPl_8)
= LOG(IPl_9)
10 LOG(IP1_10);
11 L0G(IP1 11);
PARAM
TH012 1 TH112 -1 TH212 1 LH012 1 LH112 1 LH212 1 LH312 1
TH013 1 TH113 -1 TH213 1 LH013 1 LH113 1 LH213 1 LH313 1
TH014 1 TH114 -1 TH214 1 LH014 1 LH114 1 LH214 1 LH314 1
TH015 1 TH115 -1 TH215 1 LH015 1 LH115 1 LH215 1 LH315 1
TH016 .1 TH116 .1 TH216 .1 LH016 1 LH116 1 LH216 1 LH316 1
TH017 1 TH117 -1 TH217 1 LH017 1 LH117 1 LH217 1 LH317 1
TH018 1 TH118 -1 TH218 1 LH018 1 LH118 1 LH218 1 LH318 1
TH019 1 TH119 -1 TH219 1 LH019 1 LH119 1 LH219 1 LH319 1
TH0110 1 TH1110 -1 TH2110 1 LH0110 1 LH1110 1 LH2110 1
LH3110 1
TH0111 .1 TH1111 .1 TH2111 .1 LH0111 1 LH1111 1 LH2111 1
LH3111 1;
LEP2_1 LOG(EP2_l);LEP3_1 LOG(EP3_l);
LEP4_1 L0G(EP4_1);LEP5_1 = LOG(EP5_l);
LEP6_1 LOG(EP6_l);LEP7_1 = LOG(EP7_l);
LEP8_1 = LOG(EP8_l);LEP9_1 = LOG(EP9_l);
LEP10_1 = LOG(EP10_1);LEP11_1 = L0G(EP11_1);
OLSQ LEP2_1, C, PCI, PC2, PC3, PC4, PC5, PC6; LEP2_1H=@FIT;
OLSQ LEP3_1, C, PCI, PC2, PC3, PC4, PC5, PC6; LEP3_1H=@FIT;

OLSQ LEP4_1, C, PCI, PC2, PC3, PC4, PC5, PC6; LEP4_1H-@FIT;
OLSQ LEP5_1, C, PCI, PC2, PC3, PC4, PC5, PC6; LEP5_lH- OLSQ LEP6_1, C, PCI, PC2, PC3, PC4, PC5, PC6; LEP6_lH= OLSQ LEP7_1, C, PCI, PC2, PC3, PC4, PC5, PC6; LEP7_1H-@FIT;
OLSQ LEP8_1, C, PCI, PC2, PC3, PC4, PC5, PC6; LEP8_1HH§FIT;
OLSQ LEP9_1, C, PCI, PC2, PC3, PC4, PC5, PC6; LEP9_1H-@FIT;
OLSQ LEP10_1, C, PCI, PC2, PC3, PC4, PC5, PC6; LEP10_1H=@FIT;
OLSQ LEP11_1, C, PCI, PC2, PC3, PC4, PC5, PC6; LEP11_1H=@FIT;
PM1_2=MP1_2/MP1D;
PM1_3=MP1_3/MP1D;PM1_4-MP1_4/MP1D;
PM1_5=MP1_5/MP1D;PM1_6-MP1_6/MP1D;
PM1_7=MP1_7/MP1D;PM1_8=MP1_8/MP1D;
PM1_9=MP1_9/MP1D;PM1_10=MP1_10/MP1D;
PM1_11=MP1_11/MP1D;
LMP1_2 LOG(PMl_2);
LMP1_3 = LOG(PM1_3);LMP1_4 L0G(PM1_4);
LMP1_5 = LOG(PM1_5);LMP1_6 = LOG(PMl_6);
LMP1_7 LOG(PMl_7);LMP1_8 LOG(PMl_8);
LMP1_9 = LOG(PMl_9);LMP1_10 = LOG(PM1_10);
LMP1_11 LOG(PMl_ll);
OLSQ LMP1 2,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LMP1 2H=@FIT
OLSQ LMP1_3,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LMPl_3H= OLSQ LMP1_4,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LMP1 4H=@FIT
OLSQ LMP1_5,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LMP1 5H=@FIT
OLSQ LMP1_6,
c,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LMP1 6H(9FIT
OLSQ LMP1 7,
C,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LMP1 7H= OLSQ LMP1_8,
C,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LMP1 8H=@FIT
OLSQ LMP1 9,
C,
PCI,
PC2,
PC3,
PC4,
PC5,
PC6;
LMP1 9H-@FIT
OLSQ LMP1_10,
C,
PCI,
PC2, PC3
PC4,
PC5,
PC6;
LMP1 lOHHaFIT
OLSQ LMP1 11,
C,
PCI,
PC2
, PC3
PC4,
PC5,
PC6;
LMP1_11H=@FIT
OLSQ LIQ1D, C, PCI, PC2,
PC3, PC4, PC5, PC6; LIQ1DH=@FIT;
FRML EQ1#50 LIQl_2 = (TH012 + TH112*(LMP1_2H) + TH212*(LIQ1DH));
FRML EQiy/22 LIQ1_3 = (TH013 + TH113*(LMP1_3H) + TH213*(LIQ1DH));
FRML EQ1#24 LIQ1_4 (TH014 + TH114*(LMP1_4H) + TH214*(LIQ1DH));
FRML EQ1#26 LIQ1_5 = (TH015 + TH115*(LMP1_5H) + TH215*(LIQ1DH));
FRML EQ1#28 LIQ1_6 = (TH016 + TH116*(LMP1_6H) + TH216*(LIQ1DH));
FRML EQ1#30 LIQ1_7 = (TH017 + TH117*(LMP1_7H) + TH217*(LIQ1DH));
FRML EQ1#32 LIQ1_8 (TH018 + TH118*(LMP1_8H) + TH218*(LIQ1DH));
FRML EQiy/34 LIQ1_9 (TH019 + TH119*(LMP1_9H) + TH219*(LIQ1DH));
FRML EQ1#36 LIQ1_10 (TH0110 + TH1110*(LMP1_10H) + TH2110*(LIQ1DH));
FRML EQiy/38 LIQ1_11 (TH0111 + TH1111*(LMP1_11H) + TH2111*(LIQ1DH)) ;
FRML EQ1#21 LIP1_2 (LH012 + LH112*(LEP2_1H) + LH212*LOG(YEAR)
LH312*LOG(PEN));
FRML EQ1//23 LIP1_3 = (LH013 + LH113*(LEP3_1H) + LH213*LOG(YEAR) +
LH313*LOG(PEN));

319
(LH014 + LH114*(LEP4 1H) + LH214*L0G(YEAR) +
FRML EQiy/25 LIP1_4
LH314*LOG(PEN));
FRML EQ1//27 LIP1_5 = (LH015 + LH115*(LEP5_1H) + LH215*LOG(YEAR) +
LH315*LOG(PEN)) ;
FRML EQiy/29 LIP1_6 = (LH016 + LH116*(LEP6_1H) + LH216*LOG(YEAR) +
LH316*LOG(PEN)) ;
FRML EQiy/31 LIP1_7 (LH017 + LH117*(LEP7_1H) + LH217*LOG(YEAR) +
LH317*LOG(PEN));
FRML EQ1//33 LIP1_8 -
LH318*LOG(PEN));
FRML EQiy/35 LIP1_9 =
LH319*LOG(PEN));
FRML EQiy/37 LIP1_10
LH3110*LOG(PEN));
FRML EQiy/39 LIP1_11 = (LH0111 + LH1111*(LEP11_1H) + LH2111*LOG(YEAR) +
LH3111*LOG(PEN));
(LH018 + LH118*(LEP8_1H) + LH218*LOG(YEAR) +
(LH019 + LH119*(LEP9_1H) + LH219*LOG(YEAR) +
- (LHOllO + LH1110*(LEP10_1H) + LH2110*LOG(YEAR) +
SELECT LIQ1_2 >0; IF @NOB > 6; THEN; DO;
LSQ (NOPRINT,SILENT) EQ1#50; PRINT @RSQ, @DW, @FST;
ENDDO; ELSE; DO; PRINT @NOB;
ENDDO;
SELECT LIQ1_3 > 0; IF @NOB > 6; THEN; DO;
LSQ (NOPRINT, SILENT) EQiy/22; PRINT @RSQ, @DW, @FST;
ENDDO; ELSE; DO; PRINT @NOB;
ENDDO;
SELECT LIQ1_4 > 0; IF @NOB > 6; THEN; DO;
LSQ (NOPRINT,SILENT) EQ1#24; PRINT @RSQ, @DW, @FST;
ENDDO; ELSE; DO; PRINT @NOB;
ENDDO;
SELECT LIQ1_5 >0; IF @NOB > 6; THEN; DO;
LSQ (NOPRINT,SILENT) EQiy/26; PRINT @RSQ, @DW, @FST;
ENDDO; ELSE; DO; PRINT @NOB;
ENDDO;
SELECT LIQ1_6 >0; IF @NOB > 6; THEN; DO;
LSQ (NOPRINT,SILENT) EQ1#28; PRINT @RSQ, @DW, @FST;
ENDDO; ELSE; DO; PRINT @NOB;
ENDDO;
SELECT LIQ1_7 > 0; IF @NOB > 6; THEN; DO;
LSQ (NOPRINT,SILENT) EQ1#30; PRINT @RSQ, @DW, @FST;
ENDDO; ELSE; DO; PRINT @NOB;
ENDDO;
SELECT LIQ1_8 > 0; IF @NOB > 6; THEN; DO;
LSQ (NOPRINT,SILENT) EQ1#32; PRINT @RSQ, @DW, @FST;
ENDDO; ELSE; DO; PRINT @NOB;
ENDDO;
SELECT LIQ1_9 >0; IF @NOB > 6; THEN; DO;
LSQ (NOPRINT,SILENT) EQ1#34; PRINT @RSQ, @DW, @FST;
ENDDO; ELSE; DO; PRINT @NOB;
ENDDO;
SELECT LIQ1_10 > 0; IF @NOB > 6; THEN; DO;
LSQ (NOPRINT,SILENT) EQ1#36; PRINT @RSQ, @DW, @FST;
ENDDO; ELSE; DO; PRINT @NOB;

320
ENDDO;
SELECT LIQ1_11 > 0;
IF @NOB
> 6; THEN; DO;
LSQ
(NOPRINT.SILENT)
EQiy/38;
PRINT
@RSQ,
@DW,
@FST
ENDDO; ELSE; DO; PRINT @NOB;
ENDDO;
SMPL 66,86;
LSQ
(NOPRINT,SILENT)
EQiy/21;
PRINT
@RSQ,
@DW,
@FST
LSQ
(NOPRINT,SILENT)
EQiy/23;
PRINT
@RSQ,
@DW,
@FST
LSQ
(NOPRINT,SILENT)
EQiy/25;
PRINT
@RSQ,
@DW,
@FST
LSQ
(NOPRINT,SILENT)
EQ1#27;
PRINT