Citation
An econometric model of international trade of frozen concentrated orange juice

Material Information

Title:
An econometric model of international trade of frozen concentrated orange juice
Creator:
Irias, Luiz José Maria, 1944-
Publication Date:
Language:
English
Physical Description:
xiv, 245 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Orange juice -- Marketing ( lcsh )
Frozen concentrated orange juice -- Marketing ( lcsh )
Food and Resource Economics thesis Ph. D
Dissertations, Academic -- Food and Resource Economics -- UF
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1981.
Bibliography:
Bibliography: leaves 238-244.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Luiz José Maria Irias.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
028133608 ( ALEPH )
07910432 ( OCLC )

Downloads

This item has the following downloads:


Full Text


















AN ECONOMETRIC MODEL OF INTERNATIONAL

TRADE OF FROZEN CONCENTRATED ORANGE JUICE










BY


LUIZ JOSE MARIA IRIAS
























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

UNIVERSITY OF FLORIDA 1981

































Copyright 1981

by

Luiz Jos6 Maria Irias
































To my wife, Raulina, my daughters, Ana Claudia and Ana Cristina, and my parents.















ACKNOWLEDGEMENTS

The author wishes to express his gratitude and

appreciation to Dr. Daniel S. Tilley, chairman of the supervisory committee, for his guidance, patience, encouragement and support in the preparation of this dissertation. He is quite grateful to Dr. Jonq-Ying Lee, cochairman of the supervisory committee, for his valuable assistance, especially with the computer programs. Warm appreciation and thanks are also extended to Dr. Max R. Langham and Dr. Frederick 0. Goddard, members of the committee, for their time, assistance and guidance. He was very fortunate in having this committee. Working with them was a unique experience full of intellectual stimulation. They taught him a great deal.

Special recognition is given to Dr. W. W. McPherson, graduate coordinator in the Food and Resource Economics Department, for his patience, assistance and guidance during the author's study program at the University of Florida. The author is also very proud of having Dr. McPherson attending his final examination.

The author is especially grateful to the Empresa

Brasileira de Pesquisa Agropecuiria, the Ford Foundation and the Florida Department of Citrus for providing financial support for the author's graduate study.



iv










April G. Burk merits special recognition for her

dedication in typing extensive early drafts and the final version. Likewise, Susan Howard deserves thanks for her excellent art work.

The author also appreciates the wonderful experiences and good times shared with colleagues, friends, staff, and faculty of the University of Florida's Food and Resource Economics Department and the Economic Research Department, Florida Department of Citrus.

Finally, and most importantly, the author is especially grateful to his wife, Raulina, for her continued support and understanding during these years of study.


































v
















TABLE OF CONTENTS


Page

ACKNOWLEDGEMENTS iv LIST OF TABLES. ix LIST OF FIGURES. xii ABSTRACT. xiii CHAPTER

I PROBLEM SETTING. 1

Statement of the Problem. 1 Objectives. 12

II AN ECONOMIC PROFILE. 14

Trade Flows. 14 Structural and Institutional Factors. 26

Import and Export Incentives. 27 Market Promotion Programs. 30 Exchange Rate Policies. 33 Trade Restrictions and Concessions. 35 III THE ECONOMIC MODEL. 38

Theoretical Background. 38 The Economic Model. 44

FCOJ and Other Orange Juices Market Structure
in the USA. 45 Retail FCOJ. 55 Wholesale FCOJ. 56 FCOJ Imports. 58
Retail and Wholesale Other Orange Juices. 60

FCOJ Market Structure in Canada. 61

Brazilian FCOJ Exports to Canada. 67 USA FCOJ Exports to Canada. 68



vi










TABLE OF CONTENTS--Continued


CHAPTER Page

FCOJ Market Structure in Europe. 69

Brazilian FCOJ Exports to the EEC7. 73 USA FCOJ Exports to the EEC7. 74
Brazilian and USA FCOJ Exports to the
European Non-EEC3. 75

Estimation. 75

IV EMPIRICAL RESULTS AND IMPLICATIONS. 82

Estimates of the Structural-Form Equations. 82

The USA Market. 83

Retail FCOJ. 83 Wholesale FCOJ. 88 FCOJ Imports from Brazil. 93 Retail and Wholesale Other Orange Juices. 98

The Canadian Market. 105

Brazilian FCOJ Exports to Canada. 106 USA FCOJ Exports to Canada. 111

The European Markets. 117

Brazilian FCOJ Exports to the EEC7. 118 USA FCOJ Exports to the EEC7. 122
Brazilian FCOJ Export to the European
Non-EEC3. 128
USA FCOJ Exports to the European Non-EEC3. 132

Estimates of the Derived Reduced-Form Equations. 137

Impact of Exchange Rate. 149 Impact of Orange Production. 151 Impact of Freeze. 153 Tests on Performance of the Estimated Model. 155

V SUMMARY AND CONCLUSIONS

Summary. 159 Major Conclusions. 162 Suggestions for Future Studies. 168






vii










TABLE OF CONTENTS--Continued


APPENDICES Page

APPENDIX A EXPORTS OF ORANGE JUICE. 172 APPENDIX B IMPORTS OF ORANGE JUICE. 180 APPENDIX C FLORIDA's ORANGE JUICE. 188 APPENDIX D REAL EXCHANGE RATE VALUES AND INDICES. 192 APPENDIX E VARIABLE DEFINITIONS AND DATA SOURCES. 203 APPENDIX F MEANS, STANDARD DEVIATIONS AND
COEFFICIENTS OF VARIATION FOR VARIABLES
USED IN THE MODEL. 213 APPENDIX G STRUCTURAL-FORM ESTIMATES OF
ELASTICITIES AND FLEXIBILITIES AT THE
VARIABLE MEANS. 218 APPENDIX H ACTUAL AND REDUCED-FORM ESTIMATED VALUES
OF ENDOGENOUS VARIABLES. 224 LIST OF REFERENCES. 238 BIOGRAPHICAL SKETCH. 245






























viii
















LIST OF TABLES

Table Page

1 Shares of total exports of orange juice for
major exporting countries, 1969-1978. 3

2 Orange production in the United States of
America, Brazil and other major orange-producing
countries, 1972-1979. 4

3 Shares of total imports of orange juice for
major importing countries, 1969-1978. 7

4 Market shares for major suppliers of the United
States of America imports of orange juice,
1969-1979. 18

5 The United States of America exports of FCOJ and
of other orange juices, 1972-1979. 19

6 Brazilian FCOJ weighted average export price (FOB
Santos) in major markets, 1972-1979 21

7 Weighted average wholesale and retail prices of
FCOJ and of other orange juices in the United
States of America, 1972-1979. 22

8 Brazilian and the United States of America FCOJ
weighted average export prices (FOB) in three
major markets, 1972-1979. 24

9 Orange juice imports per capita and major FCOJ
suppliers market share in EEC7, European nonEEC3 and Canada, 1972-1979. 25

10 Promotional expenditures of the three-party program in Europe and of advertising programs in the USA and Canada for orange juices from
Florida, 1972/73 1978/79. 32

11 Description of variables in the orange juice market structure in the USA. 49

12 Description of variables in the FCOJ market structure in Canada. 65



ix









LIST OF TABLES--Continued


Table Page

13 Description of variables in the FCOJ market
structures in the EEC7 and the European
non-EEC3 countries. 71

14 Structural-form estimates of the FCOJ retail
demand equation in the USA market (RQFSt,
Equation 1). 84

15 Structural-form estimates of the FCOJ retail
price equation in the USA market (RPFSt,
Equation 2) 85

16 Structural-form estimates of the FCOJ wholesale
demand equation in Florida (WQFFlt, Equation 3). 89

17 Structural-form estimates of the FCOJ wholesale
price equation in Florida (WPFFlt, Equation
4) . 90

18 Structural-form estimates of the USA FCOJ import
demand equation from Brazil (MQFSBt, Equation 7). 94

19 Structural-form estimates of the USA FCOJ import
price equation from Brazil (MPFSBt, Equation 9). 95

20 Structural-form estimates of the other orange
juices retail demand equation in the USA market
(RQOSt, Equation 10). 99

21 Structural-form estimates of the other orange
juices retail price equation in the USA market
(RPOSt, Equation 11) 100

22 Structural-form estimates of the other orange
juices wholesale demand equation in Florida
(WQOFlt, Equation 13). 101

23 Structural-form estimates of other orange juices
wholesale price equation in Florida (WPOF1t,
Equation 14) .102

24 Structural-form estimates of Brazil FCOJ export
demand equation to Canada (XQFCBt, Equation 17). 107

25 Structural-form estimates of Brazil FCOJ export
price equation to Canada (XPFCBt, Equation 18) 108





x










LIST OF TABLES--Continued


Table Page


26 Structural-form estimates of the USA FCOJ export
demand equation to Canada (XQFCSt, Equation 19). 112

27 Structural-form estimates of the USA FCOJ export
price equation to Canada (XPFCSt, Equation 20) 113

28 Structural-form estimates of Brazil FCOJ export
demand equation to the EEC7 countries (XQFEBt,
Equation 21) 119

29 Structural-form estimates of Brazil FCOJ export
price equation to the EEC7 countries (XPFEBt,
Equation 22) 121

30 Structural-form estimates of the USA FCOJ export
demand equation to the EEC7 countries (XQFESt,
Equation 23) 123

31 Structural-form estimates of the USA FCOJ export
price equation to the EEC7 countries (XPFESt,
Equation 24) 124

32 Structural-form estimates of Brazil FCOJ export
demand equation to the European non-EEC3
countries (XQFNBt, Equation 25) .129

33 Structural-form estimates of Brazil FCOJ export
price equation to the European non-EEC3 countries
(XPFNBt, Equation 26). 130

34 Structural-form estimates of the USA FCOJ export
demand equation to the European non-EEC3
countries (XQFNSt, Equation 27). 133

35 Structural-form estimates of the USA FCOJ export
price equation to the European non-EEC3
countries (XPFNSt, Equation 28). 134

36 Estimates of the derived reduced-form equations. 140

37 Derived reduced-form elasticity and flexibility
estimates for selected predetermined variables
evaluated at the means of the respective
variables. 143

38 Ratios of the root mean square error to observed
mean value for comparing actual and derived reduced-form estimated values of endogenous
variables for specific periods within the data
range used in the study. 156

xi
















LIST OF FIGURES


Figure Page


1 Orange juice trade flows for major
exporting and importing countries. 15

2 Effects of currency devaluation by exporting
country on equilibrium price and quantity for
a given commodity. 42

3 Orange juice market structure in the USA. 46

4 Orange juice market structure in Canada, the
EEC7 and the European non-EEC3 countries. 62



































xii
















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


AN ECONOMETRIC MODEL OF INTERNATIONAL
TRADE OF FROZEN CONCENTRATED ORANGE JUICE By


LUIZ JOSE MARIA IRIAS

August, 1981


Chairman: Dr. Daniel S. Tilley Cochairman: Dr. Jonq-Ying Lee Major Department: Food and Resource Economics


A simultaneous equation econometric model of international trade of frozen concentrated orange juice was developed for two major producing and exporting countries (the USA and Brazil) and four major importing countries or groups of countries, i.e., Canada, the USA, the EEC7 countries (West Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark and France) and the European non-EEC3 countries (Sweden, Norway and Finland). The model consists of demand and price equations for FCOJ and other orange juices for the USA retail, wholesale (Florida) and import (FCOJ from Brazil) markets, and for Brazil and the USA FCOJ export markets (Canada and Europe).





xiii










The system of structural-form equations was estimated by two-stage least squares.

Florida's wholesale prices, orange production forecasts, USA import and export prices, exchange rates and occurrence of freeze in Florida were found to play major explanatory roles in the various price equations. Florida's wholesale FCOJ price, together with the USA import price of FCOJ from Brazil, determines the price spread. This spread is the key variable in the simultaneous determination of imports by the USA, and the USA export price to Canada and Europe. USA export prices are the major variables in the simulatenous determination of Brazilian export prices to Canadian and European markets.

USA imports of FCOJ from Brazil were found to have an elastic response to the price spread (Florida's wholesale price minus import price). Brazilian exports to Canada and USA exports to Europe were found to have an elastic response to respective own-prices. Brazilian FCOJ was found to be a substitute for USA FCOJ in the Canadian market, while in the European markets the two products behaved quite independently of each other.















xiv















CHAPTER I

PROBLEM SETTING


Statement of the Problem

Volume of world trade of processed citrus products has increased dramatically within the last fifteen years. Among processed citrus, orange juice is by far the most important traded product as compared with fresh grapefruit and lemon juice, pulp, oil and other by-products. The majority of orange juice trade has been in the form of frozen concentrated juice with some hot pack and single strength juice also traded.

From 1969 to 1978 total orange juice exports increased from 197.6 to 646.1 million gallons of single strength juice equivalent, an increase of nearly 230 percent (Table A.l. Appendix A). In the USA and Canada, orange juice consumption

has been increasing. Myers (1978b) reports a consumption increase of over 85 percent in Canada from 1968 to 1976. From 1972 to 1979 Florida's domestic movement of frozen concentrated orange juice (FCOJ) increased more than 50 percent. Canned and chilled orange juice domestic movement increased nearly 70 percent during the same period (Florida Canners Association, 1971-1980). Similar trends in consumption took place in Europe beginning in the early sixties



1







2


when Brazil offered FCOJ at low prices (Lingens,1978). European countries increased imports of orange juice by 60 percent from 1969 to 1978 (Table B.l. Appendix B). Brazil's exports to this region, in the same period of time, increased by more than 750 percent (Table A.2. Appendix A).

Brazil, the United States of America (USA), Israel,

Spain, Italy, Morocco and Mexico are the leading orange juice producing and exporting countries. These countries have accounted for almost 90 percent of total exports (Table 1). This volume increased considerably from 1975 to 1978. Other exporting countries include Greece, South Africa, Belize and Argentina.

The USA is the world's largest producer of oranges and orange juice, and the state of Florida is the primary producing region (Table 2). According to the Florida Crop and Livestock Reporting Service (1979), practically 80 percent of the USA orange production is processed and Florida alone accounts for almost 90 percent of the entire portion which is processed. From 1969 to 1978, the USA average share of total orange juice exports was around 15 percent. Brazil's share of the export volume has increased steadily.

Brazil is the largest exporter of orange juice (Table 1) and the second largest orange producer (Table 2). The state of Sao Paulo accounts for more than 62 percent of Brazil's commercial orange production and for practically all of the








Table 1. Shares of total exports of orange juice for major exporting countries, 1969-1978



Year Brazil USA Israel Spain Italy Morocco Mexico Other 1 Total
Countries

----------------------------Percent----------------------------------------1969 14.94 14.58 12.03 7.20 15.24 11.77 .19 24.05 100.00

1970 18.38 18.28 10.01 10.00 14.11 12.58 .44 16.20 100.00

1971 32.30 15.19 13.96 7.85 9.56 4.89 .75 15.50 100.00

1972 32.31 14.65 12.60 7.86 8.90 5.97 2.45 15.26 100.00

1973 38.04 15.47 10.74 6.07 6.16 5.66 3.27 14.59 100.00

1974 35.88 16.70 12.78 6.02 5.49 4.10 3.64 15.39 100.00

1975 52.84 16.13 9.83 4.79 3.83 2.45 1.02 9.11 100.00

1976 51.89 16.63 8.12 8.80 1.16 2.62 2.10 8.68 100.00

1977 53.52 17.08 7.55 4.26 1.73 2.66 5.99 7.21 100.00

1978 65.61 8.98 6.63 3.79 1.83 2.08 4.62 6.46 100.00

Average 44.49 14.99 9.85 6.25 5.33 4.47 2.89 11.73 100.00

1Includes Algeria, Argentina, Belize, Cyprus, Ghana, Greece, Jamaica, South Africa and Trinidad-Tobago.

Source: Computed from Table A.l. Appendix A.






Table 2. Orange production in the United States of America, Brazil and other major
orange-producing countries, 1972-1979



Brazil USAc Other
Year State of Other Ttab Flrd Other Ttl countriesd3 Total
Sao Paulo states states

-----------------------------million boxes of 90 pounds---------------------------1972 60.7 20.1 80.8 137.0 54.4 191.4 206.0 478.2

1973 62.4 58.3 120.7 169.7 55.0 224.7 209.9 555.3

1974 82.0 70.6 152.6 165.8 50.4 216.2 219.5 588.3

1975 84.7 70.4 155.1 173.3 64.5 237.8 216.8 609.7

1976 99.6 80.8 180.4 181.2 61.6 242.8 205.3 628.5

1977 92.0 83.5 175.5 186.8 56.1 242.9 208.2 626.6

1978 150.0 42.4 192.4 167.8 52.3 220.1 187.8 600.3

1979 151.5 45.3 196.8 164.0 46.5 210.5 180.9 588.2

Total 782.9 471.4 1,254.3 1,345.6 440.8 1,786.4 1,634.4 4,675.1

Percent 16.75 10.08 26.83 28.78 9.43 38.21 34.96 100.00


1Includes the states of Bahia, Minas Gerais, Parana, Rio Grande do Sul, Rio de Janeiro, Santa Catarina and Sergipe. 2Includes the states of Arizona, California and Texas. 3Includes 4Argentina, Belize, Israel, Italy, Jamaica, Mexico, Morocco, Spain and South Africa. Average percentage of total orange production. Sources: aU ited States Department of Agriculture Foreign Agricultural Service,1980a and 1980b. Food and Agriculture Organization of the United Nations,1973-1979;and United States Department of Agriculture, Foreign Aaricultural Service,1980c. cFlorida Crop and Livestock Reporting Service,1976 and 1979. Food and Agriculture Organization of the United Nations, 1973-1979.






5


citrus industry processing capacity. More than 80 percent of Sao Paulo's orange production is estimated to be processed into FCOJ (United States Department of Agriculture, Foreign Agricultural Service,1980a), and almost all of the FCOJ is exported.2 The Brazilian share of total orange juice exports increased from 14.94 percent in 1969 to 65.61 percent in 1978 with an average share for the period of 44.49 percent (Table 1).

Israel ranks third in orange juice exports followed by Spain, Italy, Morocco and Mexico (Table 1). Export markets for the fresh fruit are the primary outlets for Isreali citrus fruit, followed by the domestic fresh market. The citrus processing industry is the third most important market alternative for the Israeli product under the Citrus Marketing Board policies (Melamed,1977). Spain, Italy and Greece are orange juice exporting countries that are entirely dependent on the market in the European Economic Community (EEC). Among the remaining countries that export orange juice, Mexico is the most important. From an insignificant position in the late sixties, Mexico moved to


1Except for small plants located in the northeastern state of Sergipe and in the southern states of Santa Catarina and Rio Grande do Sul, Brazil's citrus processing industry is located entirely in the state of Sao Paulo. 2Domestic consumption of processed oranges in Brazil is estimated to be 5 to 7 percent (Myers, 1978a; United States Department of Agriculture, Foreign Agricultural Service, 1980a).







6


be the fourth most important exporting country in 1977 and 1978, its shares of total exports being 5.99 and 4.62 percent, respectively, in those years (Table 1).

From 1969 to 1978, Brazilian USA and Israeli orange

juice exports accounted for nearly 70 percent of the total trade. From 1975 to 1978, these countries exported almost 80 percent of the total. Brazil and the USA accounted for an average of 60 percent of the total orange juice exports and from 1975 to 1978 these two countries exported around 70 percent of the total (Table 1). They are the major competing suppliers in world orange juice markets.

The major orange juice importing countries are located

in Europe and North America. From 1969 to 1978, 12 countries in these two regions imported nearly 90 percent of total orange juice traded (Table 3). Most European imports go to EEC member countries and specifically to West Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark and France (the EEC7). These countries imported an average of 63.07 percent of the total world imports.

The European non-EEC countries are the next largest

orange juice importing region, accounting for 18 percent of the total volume traded. Within this region Sweden, Norway and Finland (the European non-EEC3) import 9 percent of the world total (Table 3) and have the highest per capita orange juice consumption indices in Europe (Lingens,1978).

North America, as represented by Canada and the USA, is ranked as the third most important orange juice importing








Table 3. Shares of total imports of orange juice for major importing countries, 1969-1978


EEC European non-EEC
EEr Eun- ean othErs Canada USA Other Total
Year EEC7 Others 2 non-EEC3 Others countries5
--------------------------------------Percent------------------------------------1969 67.20 .74 9.53 10.84 8.89 1.99 .81 100.00

1970 66.70 1.91 11.68 10.44 7.15 .65 1.47 100.00

1971 67.90 3.33 8.91 8.04 6.91 3.59 1.32 100.00

1972 66.65 2.63 7.75 8.05 6.51 7.32 1.09 100.00

1973 68.20 1.32 8.58 8.10 7.72 4.52 1.56 100.00

1974 64.22 1.12 9.24 10.52 8.06 3.42 3.42 100.00

1975 62.20 .43 8.14 9.23 9.49 4.80 5.71 100.00

1976 60.79 .19 9.80 8.73 8.95 4.80 6.74 100.00

1977 61.60 .17 9.11 8.89 9.35 6.75 4.13 100.00

1978 51.98 .29 7.53 7.64 11.37 18.92 2.27 100.00

Average 63.07 1.11 8.89 8.90 8.59 6.31 3.13 100.00

1EEC7 as defined in this study includes West Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark and France. 20ther EEC includes Italy and iceland. 3European non-EEC3 as defined in this study includes Sweden, Norway and Finland. Other European non-EEC3 includes Austria, East Germany, Iceland, Spain and Yugoslavia. 50ther countries include Antigua, Australia, Barbados, Jamaica, South Africa and Trinidad-Tobago. Source: Computed from Table B.1. Appendix B.







8


region accounting for an average of 15 percent of the total. While Canada's share of total imports has been relatively constant at about 8 percent, the USA import share has fluctuated from .65 percent in 1970 to 18.92 percent in 1978.

Most previous studies on production, consumption and trade of orange juices with an international perspective concentrate on one sector or on one country or world region. In the USA a number of studies focused on orange juice import and export trade issues with domestic and/or foreign perspectives (Polopolous, 1968; Ward & Niles, 1975; Ward, 1976a; Myers et al.,1979). Most of these studies viewed the import-export issues from the viewpoint of the Florida citrus industry. Ward's (1976a) study was the first major FCOJ trade model addressing import and export issues in a simplified world context. It provided an understanding of the major economic relationships of Florida's imports and exports. The model consists of a system of simultaneous equations explaining imports to Florida, export price from Florida to Europe, European demand for Florida FCOJ and a definitional equation for the price spread.3 European demand was found to be highly sensitive to Florida concentrate price changes.

3.
Price spread was defined as the difference between the quarterly average Florida domestic FOB price and the quarterly average price of imports to Florida from Brazil, excluding tariffs. The import price was adjusted by the cost of converting bulk concentrate to equivalent gallons of retail pack concentrate (Ward, 1976a, pp. 105-106).






9


The desirability of a two-pricing system between the domestic and European markets was clear. Price spread changes were found to be the major factor influencing the level of Florida imports of FCOJ. Myers et al. gave an overview of the FCOJ import-export policies and programs by evaluating how different governmental policies, industry programs and economic factors affect the volume and directions of product flows throughout the system. An important contribution of this study was to show the interdependency between price spreads, tariff levels, and duty drawback provisions and their effects on the quantity of Florida imports. It was shown that under the current set of import-export policies a trade-off exists between the percent of imports offset with exports and the magnitude of the price spread. Even at a zero price spread it is possible to import and export using the drawback provisions without losing money. The necessary condition is a relatively higher export price as compared with the domestic price. Studies by Prato (1969), Rausser (1971) and Buckley (1977) addressed the USA FCOJ domestic economic issues. They focused on the estimation and analysis of consumer demand structures within the USA domestic markets with little reference to the role of imports and exports.


4The duty drawback provisions of the U. S. Tariff Act of 1930 permit processors to import foreign FCOJ in bulk form and to recover 99 percent of the duties by exporting like amounts of FCOJ in bulk or consumer pack within a 5-year period from date of importation (Myers et al.,1979).






10


Brazilian participation in world trade of orange juice has been studied by Morais and Medeiros (1978), and by Moretti (1978). Morais and Medeiros describe the Brazilian citrus industry in a general fashion. Moretti's work uses an analytical framework to describe and to analyze important market factors affecting Brazilian trade of FCOJ. He used a single-equation model to estimate export demand relationships. Major study results indicate that concentrate exports were highly responsive to changes in prices and exchange rates.

European markets are quite important in orange juice

consumption. Priscott (1969), Ward (1976b), and Nguyen (1977) are the major studies which addressed trade, consumption patterns, and structural consumption relations in these markets. These studies focused on Florida's FCOJ exports to Western European countries. Their results showed highly price-elastic markets in most of the countries with changes in exchange rates affecting export demands. Ward's (1976b) study also addressed the benefits to the Florida citrus industry of discounting its export price to Europe relative to the domestic price by using the relatively cheaper imported product.

Chern (1973), and Tilley and Lee (1981) studied

consumer demand for orange juice in Canada. Chern used a single-equation model to estimate Canadian demand for FCOJ, canned single strength orange juice (CSSOJ) and chilled orange juice (COJ). An aggregation of the three products was






11


found to have a unitary price elasticity, an elastic response to income and a positive but a statistically insignificant effect of advertising. Individual orange juice product estimates indicated that FCOJ was price-inelastic and had a strong consumer response to advertising. CSSOJ and COJ demand was found to be price-elastic with no immediate response to advertising. All products were positively related to income changes. Tilley and Lee (1981) used a system of simultaneous equations to analyze Canadian orange juice consumer demand and import demand from Brazil and from the USA. Canada's imports from Brazil were found to be highly price-elastic, while imports from the USA were quite price-inelastic. Cross-price effects on imports indicate that an absolute change in the import price from the USA would have a greater impact on Canadian imports from Brazil than the impact on Canadian imports from the USA, given the same change in the import price from Brazil.

Despite significant research efforts a more complete world trade model for orange juice has not been developed. Such a model should explain levels of consumption, trade flows, inventories and prices for major producing and consuming countries. Also, the model should have the capability of evaluating the impact of the introduction of, and changes in, protective trade barriers, such as tariffs and quotas. These are major trade issues affecting economic, social and political decisions in the citrus industry around the world. A model of world orange juice trade would give a






12


better understanding of the economic relations and interrelations of these factors under domestic and international conditions.


Objectives

The overall objective of this study is to provide an economic analysis of domestic and foreign trade of orange juice products (frozen concentrate and other forms of orange juice traded) for major producing and consuming countries. This analysis will address the basic economic relations and interrelations in the determination of levels of consumption, trade flows and prices. An attempt will be made to:

1. identify the most important factors affecting trade of orange juice products;

2. evaluate to what extent changes in domestic and

foreign demand for the USA orange juice products affect the USA FCOJ imports;

3. evaluate whether foreign and domestic promotion programs for the USA orange juice products provide an "umbrella" for the market expansion for the same products from other producing-exporting countries;

4. determine the impact of pricing strategies and price changes on exports and imports of orange juice products among producing-exporting and importing countries.

In order to accomplish these objectives an economic

model for international trade in orange juice products will be developed. This model will focus on two basic orange juice products: frozen concentrated and an aggregate of






13


other product forms of orange juice. Unless otherwise specified, these products are expressed in gallons of single strength juice equivalent. For modeling purposes the world trade of these orange juice products consists of a simplified geographical scenario of two major producing and exporting countries (the United States of America and Brazil) and four major importing countries or groups of countries (Canada, the United States of America, the EEC7 and the European non-EEC3 countries).















CHAPTER II


AN ECONOMIC PROFILE


The purpose of this chapter is to describe and discuss the orange juice industry. The first section analyzed basic economic relations concerning production, consumption and trade of orange juices, while the second section discusses major structural and institutional factors affecting the industry.


Trade Flows

Figure 1 summarizes orange juice flows for major

exporting and importing countries. Solid arrows indicate export flows and broken arrows indicate import flows. Numbers shown are estimated average percentages of exports or imports from 1969 to 1979, except in the case of Florida where the percentages are for the 1972 to 1979 period.

The USA, as the largest producer and consumer and the second largest exporter of orange juice, is the primary economic indicator of this industry around the world. The state of Florida is by far the largest USA supplier of orange juices, followed by the states of Arizona, California and Texas. While Florida's orange production is practically all utilized for processed products, the production from


14








Stocks imports
Stocks 18 17 -j stic
Domestic
supply 79 78 Domestic

LORI demand
Foreign 4 FLORIDA
exports
U. S. A.


N, ~rg~84 9 IU

25 48 1

54 Other
S, European
Eoen EEC Canada 9 Importing
nonEECCountries


13 6 37

95


4? ~S~oPaulo F
Domestic 5
Other demand
exporting Brazil
countries II_ I




Figure 1. Orange juice trade flows for major exporting and importing countries.
Note: Solid arrows indicate export flow and broken arrows indicate import flows. Numbers are approximated average
percentage of exports or imports from 1969 to 1979 (Appendices A, B and C).







16


other states is mainly for fresh products (Myers, 1978b). Detailed data on processed orange products are not available for the states of Arizona, California and Texas. Therefore most of the discussion on the USA processed orange industry will be related to state of Florida data.

In the last eight years (1972 to 1979) an average of

approximately 79 percent of Florida's total supply of orange juices was produced from the current crop of domestic oranges. The remaining supply came from stocks from previous years (17 percent) and from imports (4 percent). In the past, Florida has been a net exporter of orange juice. Only in the recent two years (1978 and 1979) have Florida's processors imported more than they have exported. This occurred following a freeze in 1977 that substantially reduced domestic production. Florida accounted for more than 70 percent of USA total imports and 80 percent of total USA exports from 1972 to 1979 (Appendix B and Appendix C). Florida's canned and chilled orange juices exports represent about 50 percent of the USA other orange juices (single strength and hot pack) exports (Table A.4. Appendix A and Appendix C).

Florida imports and export data by country are not

available. Because of Florida's market dominance, United States Department of Commerce data can be used to identify sources and country of destination. Brazil has been the primary source of the USA imports of FCOJ. Brazilian imports averaged around 84 percent of total USA FCOJ imports from







17


1969 to 1979. Mexico and Belize accounted for 10 and 1 percent respectively (Figure 1, Table 4). USA FCOJ imports varied from 2.6 million gallons in 1970 to a record of 156.4 million single strength equivalent gallons in 1979 (Table B.2. Appendix B). FCOJ is the leading citrus export product in the USA accounting for approximately 77 percent of total exports. Single strength and hot pack accounted for the remaining 23 percent over the eight-year period shown in Table 5. In the six years preceeding the 1977 Florida freeze, the USA FCOJ exports were increasing rapidly. In 1978, exports decreased by almost 40 percent from those of 1977. Practically 48 percent of all the USA orange exports go to Canada (Figure 1 and Table A.3. Appendix A). The EEC countries, led by The Netherlands and West Germany, accounted for an average of more than 25 percent, and the European non-EEC countries accounted for almost 17 percent of USA exports (Table A.4. Appendix A). In 1974 the Brazilian orange industry faced its first crisis when exports declined by more than 10 percent. The economic recession in Europe, high stock levels in most of the importing countries, and quality control problems of the reconstituted Brazilian juice in those countries are factors contributing to the decline.

Prices for the Brazilian product (FOB Santos) declined

below the 1974 minimum export price of 560 dollars per metric CAE1
ton established by CACEX (Morais and Medeiros, 1978). 1CACEX Carteira de Com6rcio Exterior, Export Trade Office of the Bank of Brazil, Inc., associated with the Ministry of Finance. CACEX's major role is to implement and enforce official government policy regarding exports and imports.








Table 4. Market shares for major suppliers of the United States of America imports of
orange juice, 1969-1979


Year Brazil Mexico Belize Other 1 Total
countries
--------------------------------Percent-----------------------------------1969 53.30 9.77 23.99 12.94 100.00

1970 49.68 21.91 28.33 .08 100.00

1971 81.37 6.49 7.47 4.67 100.00

1972 77.61 12.87 2.69 6.83 100.00

1973 46.56 20.43 4.71 28.30 100.00 m

1974 67.20 24.57 1.73 6.50 100.00

1975 85.38 10.03 00 4.59 100.00

1976 89.14 4.22 00 6.64 100.00

1977 67.89 27.74 .36 4.01 100.00

1978 92.00 6.99 00 1.01 100.00

1979 95.07 4.72 00 .21 100.00

Total 84.41 10.23 1.29 4.07 100.00


1Includes Argentina, Israel, Italy, Japan, Spain, South Africa and eleven other countries. Source: Computed from Table B.2. Appendix B.







Table 5. The United States of America exports of FCOJ and of other orange juices,l
1972-1979



FCOJ Other orange juices Total
Year 2 22
Volume Percent Volume Percent Volume Percent


1972 34,585 68.89 15,620 31.11 50,205 100.00

1973 47,155 75.42 15,367 24.58 62,522 100.00

1974 45,155 76.68 14,951 23.32 64,106 100.00

1975 56,150 80.05 13,990 19.95 70,140 100.00

1976 69,096 80.89 16,322 19.11 85,418 100.00

1977 69,486 80.29 17,053 19.71 86,539 100.00

1978 42,428 73.11 15,609 26.89 58,037 100.00

1979 48,539 73.57 17,439 26.43 65,978 100.00

Total 416,594 76.73 126,351 23.27 542,945 100.00

1Includes orange juices exported as single strength and as concentrated but not frozen (hot pack).

2In 1,000 gallons of single strength juice equivalent. Source: Computed from Table A.4. Appendix A.







20


Nominal price changes for Brazilian FCOJ exports to major markets (Table 6). Average prices in 1975 were the lowest prices shown in all markets. Price levels recovered following the 1977 Florida freeze. Western Europe countries, and especially the EEC countries and the European non-EEC countries, have been the primary customers for Brazilian FCOJ exports accounting for almost 61 percent of Brazil's total concentrate exports from 1969 to 1979 (Figure 1 and Table A.2. Appendix A). West Germany, The Netherlands and Sweden accounted for more than 40 percent of Brazilian FCOJ exports. The USA bought an average of 23 percent of the Brazilian product. The USA recorded 187.4 million gallons of FCOJ in single strength juice equivalent imports from Brazil in 1978 when it emerged as Brazil's most important buyer.

As a result of the 1977 Florida freeze and the subsequent reduced orange production, domestic prices of orange juices increased in the USA. Both wholesale and retail prices increased substantially as compared with 1976 levels (Table 7). In 1978 FCOJ prices increased as much as 71 percent at wholesale and 59 percent at retail from respective price levels of the year before the freeze. Non-frozen orange juice prices increased approximately 30 percent in both markets. The increased price differential between wholesale domestic and imported juice made it profitable for the USA importers and processors to buy foreign orange juice







Table 6. Brazilian FCOJ weighted average export price (FOB Santos) in major markets,
1972-1979



Year EEC 2 Canada USA Other 3 Total
non-EEC countries
--------current cents per gallon of single strength juice equivalent--------1972 36.93 38.57 37.96 38.34 36.56 37.48

1973 41.51 40.54 38.43 44.47 39.41 41.39

1974 43.04 44.61 43.64 40.93 45.41 42.94

1975 35.38 36.85 35.35 35.57 37.00 35.77

1976 38.05 37.66 36.90 37.10 38.25 37.84

1977 62.91 59.82 59.02 75.04 73.54 65.26

1978 77.34 74.20 81.44 77.62 82.95 78.01

1979 75.61 73.90 76.50 75.29 79.35 75.78

1Includes West Germany, the United Kingdom, The Netherlands, Belgium, Luxembourg, Denmark and France. Includes Sweden, Norway, Finland and 7 other European countries. 3Includes Israel, Japan, Venezuela and eleven other countries. Source: Computed from Brazil, Servigo de Estatistica Econdmica e Financeira, Minist6rio da Fazenda, 1970-1979.








Table 7. Weighted average wholesale and retail prices of FCOJ and of other orange juices1
in the United States of America, 1972-1979



Year FCOJ Other orange juicesa
Wholesale Retailb Wholesalea Retailb
----- current dollar per gallon of single strength juice equivalent------1972 .81139 1.25078 1.09515 1.77025

1973 .75906 1.23319 .99571 1.71801

1974 .79321 1.27437 1.11686 1.80698

1975 .86490 1.37840 1.20589 1.88345

1976 .83684 1.41075 1.33023 1.94496

1977 1.14948 1.70821 1.54129 2.19211

1978 1.43292 2.23701 1.84367 2.69041

1979 1.43489 2.40002 1.97372 2.96256


1Includes canned single strength and chilled orange juices at Florida's wholesale level,
and single strength orange juices in cans, glasses, and other containers at the USA retail level.

Sources: a Florida Department of Citrus, Economic Research Department,1972-1980;and bFlorida Department of Citrus and A. C. Nielsen Company, 1972-1980.






23


concentrate, pay the USA import duty and sell the imported product in the domestic market. FCOJ export prices from the USA have been consistently quoted higher than prices for the Brazilian product (Table 8). Brazilian FCOJ exports are in 650 brix concentrate in bulk drums, while most of the USA exports are in consumer retail pack at 420 brix to Europe and Canada. Brazilian FCOJ exports to Canada have expanded substantially during the seventies (Table A.2. Appendix A). From under 6 million single strength equivalent gallons in 1969 and 1970, exports of the Brazilian orange juice to Canada increased by more than 600 percent, reaching a record of 43.0 million gallons in 1978. Total Brazilian FCOJ exports in 1979 declined to 371.4 million single strength juice equivalent gallons or approximately 12 percent less than the record level of 1978.

As previously stated, Europe and Canada are the most important importers of orange juice. Per capita import indices and FCOJ market shares to major suppliers were constructed for these regions for the last eight years (Table 9). Scandinavian countries (the European non-EEC3) have the highest per capita imports in Europe. According to Lingens (1978), per capita consumption of orange juice in Sweden is more than five gallons of single strength


2
Import duty on FCOJ imports has been 34 cents per pound of solids or 34.986 cents per gallon of single strength juice equivalent or 487.22 USA dollars per metric ton of 650 brix concentrate.








Table 8. Brazilian and the United States of America FCOJ weighted average export prices
(FOB) in three major markets, 1972-1979



Year EEC71 European non-EEC32 Canada
Brazil USA Brazil USA Brazil USA
------current USA dollars per gallon of single strength juice equivalent----1972 .36908 .62114 .38503 .82022 .37957 .90133

1973 .41549 .51342 .40967 .74340 .38433 .90879

1974 .43042 .59091 .44630 .81385 .43641 .92286

1975 .35365 .63207 .37995 .86196 .35354 1.00308 I

1976 .37413 .56843 .37662 .83142 .36899 1.00201

1977 .62909 .70637 .61148 .95143 .59024 1.22939

1978 .77343 1.49492 .73833 1.29239 .81436 1.71403

1979 .75615 1.47822 .73134 1.26381 .76502 1.75170


1Includes West Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark and France. 2Includes Sweden, Norway and Finland. Sources: Computed from Brazil, Servigo de Estatistica Econ8mica e Financeira, Ministerio da Fazenda, 1970-1979; and from United States Department of Commerce, Bureau of the Census, 1969-1979a.








Table 9. Orange juice imports per capita and major FCOJ suppliers market share in EEC7
European non-EEC32 and Canada, 1972-1979



Imports per capita3 FCOJ market shares4
Year EEC71 European2 Canada EEC7 European non-EEC32 Canada
non-EEC3 5
Brazil USA Brazil USA Brazil USA Others


1972 .38 1.00 1.81 88.71 11.29 57.23 42.77 36.63 51.05 12.32

1973 .63 1.53 2.22 89.29 10.71 59.56 40.44 39.12 47.00 13.88

1974 .43 1.89 2.21 90.36 9.64 68.48 31.52 19.18 60.36 20.46
U,
1975 .68 2.13 2.87 90.98 9.02 72.22 27.78 47.94 46.82 5.24

1976 .93 2.50 2.83 87.99 12.01 77.40 22.60 39.21 54.86 5.93

1977 .75 2.54 2.96 88.07 11.93 76.64 23.36 39.28 49.31 11.41

1978 .66 2.42 3.61 94.38 5.62 89.92 10.08 61.07 34.13 4.80

1979 .85 2.54 3.71 94.56 5.44 91.06 8.94 56.37 40.10 3.53


1Includes West Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark and France. 2Includes Sweden, Norway and Finland. 3Gallons of single strength juice equivalent per capita. In the EEC7 and the E ropean non-EEC3 countries includes only imports from Brazil and the USA. 4Percent. Includes Argentina, Belize, Israel, Jamaica, Mexico and nineteen other countries.

Sources: Computed from Tables B.3. to B.5., Appendix B; Table A.4., Appendix A and International Monetary Fund (population), 1970-1980.







26


juice equivalent per year. Considerable attention is given to juice quality and packing procedures in Sweden. In EEC7 total orange juice imports in 1979 from Brazil and the USA, were around 169.2 million single strength juice equivalent gallons, an increase of more than 590 percent from the 1969 import level (Table B.4. Appendix B). Orange juice imports per capita more than doubled in that region in the last eight years (Table 9). In Sweden, Norway and Finland, Brazil substantially increased its FCOJ market share from 57.23 percent in 1972 to 91.06 percent in 1979. Even though the USA's participation in this market increased in absolute terms, the USA share of total FCOJ exports declined from 42.77 to 8.94 percent from 1972 to 1979 (Table 9). All orange juice used in Canada is imported. Brazil and the USA are again the major suppliers with a large group of other countries far behind. In 1978 and 1979, Brazil's market share exceeded 50 percent. The USA share dropped below 40 percent in 1978 and was just over 40 percent in 1979.


Structural and Institutional Factors

The purpose of this section is to describe and discuss major structural and institutional factors affecting production, consumption and trade in the orange juice industry.

In Brazil the major government impact on trade of FCOJ is exercised through CACEX and the Citrus Juice Export Committee integrated by government officials (CACEX,







27


Ministry of Agriculture,and state of Sao Paulo) and by citrus grower and processor representatives. The most important activities of CACEX are to license FCOJ exports, to establish and enforce reference prices for oranges to be processed, to set minimum FCOJ export prices, and to control stocks. The remainder of this section is organized to cover import and export incentives, market promotion programs, exchange rate policies, protective restrictions, and preferential concessions.


Import and Export Incentives

In the USA the duty drawback provisions of the Tariff Act of 1930 are a very important mechanism influencing FCOJ trade. Under such provisions processors can import foreign FCOJ in bulk form and recover 99 percent of the import tariff (34 cents per pound of solids) by exporting a like amount of product in bulk or consumer pack within a fiveyear period from the date of importation (Myers et al., 1979). The rationale for such trade activity was discussed by Ward (1976a) and expanded by Myers et al. (1979). Import tariff, the price spread (the differential between domestic and import FCOJ prices), duty drawback provisions and export prices are key determinants of USA FCOJ imports and exports. If the price spread is higher than the tariff, then it is profitable to sell imported FCOJ in the domestic markets. However, if the price spread is equal to or less than the tariff on imports, then processors will not be able to







28


realize profits on the imported FCOJ unless they have an option to export part of the reprocessed product and then recover most of the tariff paid by taking advantage of duty drawback. As long as the price spread is less than the tariff level, more juice needs to be exported under duty drawback for a given export price. As the spread approaches the tariff level, fewer exports are needed to compensate for imports. Under the present set of USA FCOJ import-export policies, a trade-off exists between the percent of imports offset with exports and the magnitude of price spread. Myers et al. (1979) showed that, even at a zero price spread, it is possible to import and export FCOJ and use the duty drawback provisions without losing money. The necessary condition is a relatively higher export price as compared with the USA domestic price.

Brazil3 does not have specific incentives for citrus growers, but does have federal and state support programs for the agricultural sector. Among others, citrus producers are entitled to subsidized credit, research and extension programs. The government also gives some assistance in establishing grower prices for oranges to be processed through the Citrus Juice Export Committee. However, oranges are not included in the minimum price program carried out by the federal government for major agricultural products. The subsidized credit program of zero interest rates on loans


3Most of the discussion of export incentives in Brazil was organized from United States Department of Agriculture, Foreign Agricultural Service, 1980a and 1980b.







29


for fertilizer and of 15 to 18 percent a year for other production activities, can be a great incentive in view of Brazil's high rate of inflation in the past years (more than 40 percent). Processors and exporters have various incentives in Brazil. Under a national policy, given the non-existence of similar domestic products,processors are allowed to import plant parts and equipment duty-free. Also available are subsidized credit for investments in new factory equipment (in 1979 the interest rate was 22 percent a year for five- or six-year loans), and for financing production for exports (in 1979 the interest rate was 8 percent a year discounted at the time of the loan). Under the latter program, an exporter is allowed to borrow up to 30 percent of the value of his exports in the previous year and repay it in no more than a year. Oranges processed for export are not subject to state and federal value-added taxes (ICM and IPI) and profits from FCOJ export sales are exempt from corporate income taxes.

A tax credit of 20 percent on value of exports, which could be used to offset taxes on domestic sales, has been

reduced since 1979 by one percent per quarter so that it will be completely eliminated by the end of 1983. Initiated late


4ICM (Imposto de Circulagao de Mercadorias) is a statelevied value-added tax, and IPI (Imposto de Produtos Industrializados) is a federal-levied value-added tax on industrial products.







30


in 1979, when the Brazilian cruzeiro5 was substantially devaluated against the USA dollar, an export tax on FCOJ was introduced as a measure intended to support export prices and to guarantee an adequate domestic supply of oranges and processed products. Market Promotion Programs

Brazilian citrus industry market promotion activities have been minimal. Domestically, the only information available (Myers, 1978a) was related to industry participation in expanding internal use through the development of a school lunch market for FCOJ in Sao Paulo and in other central and southern Brazil states. Growers, in general, are not directly involved in any demand-creation efforts. In external markets the situation is similar. Only one Brazilian processor (Cutrale), jointly with Coca Cola, was participating in an effort to package and market concentrate in a form which requires no refrigeration. A successful market test was completed in Sweden, and production and distribution began in 1979 (Myers, 1978c). Processors appear to be quite unenthusiastic about market promotion programs, given the unusually strong current demand for their product. Practically all processed products are exported.


5Cruzeiro is the Brazilian monetary unit. Brazil's exchange rate policy is discussed with more detail in the subsection on that topic.







31


In the USA, and in particular the Florida citrus

industry, a great effort has been under way for several years to promote citrus products from Florida in domestic and in foreign markets. Domestically, Florida Department of Citrus expenditures for promotions on orange juice products are reported to be more than 12.6 million dollars as an average for the last seven fiscal years (Table 10). To this effort one should add specific brand and distributor's advertising programs around the country. Florida's shipments of orange juice to domestic markets has increased by more than 50 percent in the last eight years (Table C.3. Appendix C). In foreign markets, the Florida citrus industry has been promoting orange juices in Western Europe and in Canada. Since 1966-67 their major promotional effort in Europe has been through the Three-Party Market Development Program.6 The main target of the program, as measured by distribution of total expenditure, appears to be the Scandinavian countries (European non-EEC3). About half of the promotional expenditure has been used in Sweden, Norway and Finland (Table 10). The other half of the promotional expenditures has been used in the seven EEC countries. Effectiveness of the three-party program has been reported by the studies of Lee (1977, 1978) and


6This program has been supported by Florida Department of Citrus, United States Department of Agriculture, Foreign Agricultural Service, and European distributors. See Lee et al. (1978) for a more complete description.







Table 10. Promotional expenditures of the three-party program in Europe and of advertising
programs in the USA and Canada for orange juices from Florida, 1972/73 1978/79



1 Three party program in Europe
Fiscal year 2 European3 Others 4Total USA Canada
non-EEC3
--------------------------------1,000 dollars---------------------------1972-73 636.4 788.7 190.4 1,615.5 9,790.0 311.4

1973-74 723.3 756.7 137.0 1,617.0 6,743.4 434.1

1974-75 397.2 650.9 121.2 1,169.3 12,131.4 483.6

1975-76 518.0 733.8 105.9 1,357.7 12,938.5 557.2

1976-77 507.6 680.5 157.1 1.345.2 16,360.4 462.6

1977-78 507.3 444.0 113.9 1,065.2 12,792.9 659.4

1978-79 470.8 404.9 157.5 1,033.2 17,791.9 738.8

Total 3,760.6 4,459.5 983.0 9,203.1 88,548.5 3,647.1


1Includes expenditure from July to June (twelve months), except for the three-party program in 1975-76 (July 1975 to September 1976 or fifteen months) and from 1976-77 to 1978-79 (October to September or twelve months).
2
Includes Wes Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark and France. Includes Sweden, Norway and Finland. 4Includes Austria, Iceland, Italy and Switzerland.

Source: Florida Department of Citrus,1980.







33


Lee et al. (1979). Overall conclusions from those studies are favorable to the program. Additional dollar sales generated by the program are far in excess of its costs. These studies did not identify any direct link of the three-party program expenditures to Florida's imports of FCOJ. In Canada, the Florida Department of Citrus advertising program for orange juices has also increased substantially. From 1972/73 to 1978/79 promotional expenditures in Canada increased by practically 140 percent, achieving in that period an average fiscal year expenditure of more than one-half million dollars (Table 10). A multi-level study by Tilley and Lee (1981) on Canadian demand for orange juice found a weak positive sales response as advertising expenditure increased.


Exchange Rate Policies

Exchange rates express the monetary value of

a national currency per unit of another country's currency. Changes in this relation reflect changes in the domestic currency's purchasing power as compared with the foreign currency. In terms of commodity trade, an exporting country's devaluation is equivalent to a commodity price cut viewed by the foreign importer. A devaluation of the Brazilian cruzeiro relative to the USA dollar means that fewer dollars are needed to buy a unit of a given commodity from Brazil for any given price expressed in cruzeiros. The final effect of exchange rate changes on the amount of trade







34


depends on other factors such as commodity elasticities, government import-export policy and adjustments of the international monetary system. These factors will be incorporated in the theoretical discussion of exchange rates in the next chapter. The remaining part of this subsection will discuss exchange rate changes and exchange rate policies in Brazil and in the USA.

In Brazil the official exchange rate has been

historically controlled by the Government to a greater or lesser degree, depending on the time period analyzed (Veiga, 1974). Since 1968, Brazil has been using a system of miniadjustments of the cruzeiro against the USA dollar. These adjustments, most of them devaluations,have been made at short and random intervals (one or two months) to minimize the speculative demand for dollars. The cruzeiro is devalued at a rate approximately equal to the difference between Brazil's internal inflation rate and the average inflation rate of Brazil's major trading partners, basically the USA, Western Europe and Japan (Alves & Pastore, 1978, p. 866). A departure from the mini-devaluation policy by the Brazilian Government took place in December 1979 when the cruzeiro was devalued by 30 percent (Fundagao Gettlio Vargas, 1972-1980). This was part of an overall change of the country's economic programs in an expectation of better economic results that would help it pay its increasing international debt (especially oil bills and loans), as well as to fight the high and increasing levels of internal






35


inflation. The annual rate of inflation in 1979, estimated by Fundagao Getilio Vargas (Conjuntura Econ6mica 34, February,1980 p. 14), was 77.2 percent or almost twice as much the 40.8 percent level in 1978.

Foreign exchange market forces are the basic

determinants of exchange rate values for the dollar in the USA. The change from a fixed to a floating system of exchange rates by the international monetary community in early 1973 (Abrams, 1980) was another factor that can be used to explain increasing variability of the USA dollar against major foreign currencies. Schuh (1974) argued that the exchange rate has had an important role in explaining depressed prices of agricultural products associated with an overvaluation of the dollar prior to 1971, as well as the increases in the same prices in 1973-74 associated with a devaluation of the dollar. Studies by VellianitisFidas 1976, and by Johnson et al., 1977 that attempt to empirically evaluate the effects of exchange rate variability on the USA trade in agricultural commodities fail to support the hypothesis that devaluation had a substantial effect on prices. Both studies found that the exchange rate was found to be less important than other factors in explaining high USA agricultural products prices. Trade Restrictions and Concessions

Trade restrictions include a variety of protective

import tariffs, quotas and non-tariff barriers designed to






36


shield domestic industry and/or to enforce quality control measures as a way to protect consumers. Moretti (1978) gave a complete description of duties on citrus and citrus products entering the major importing countries. Great variation in those duties is apparent. For example, FCOJ is free of tariffs in Norway, however it is subject to 200 percent ad valorem duty in Venezuela. Orange juice entering the EEC is subject to a common external ad volorem tariff of 12 to 19 percent according to the specific gravity of the juice. However, the EEC allows for a partial or total suspension of imports and the establishment of a minimum import price, if the market for certain processed fruits, including orange juices, experiences or is threatened with serious distrubances (Taylor, 1978). Canada has an overall five percent ad valorem duty on orange juice. However, the product is free of duty to commonwealth suppliers and to all other supply sources of unsweetened orange concentrate if not less than 580 brix. Finland has a 30 percent duty on orange juices that can qualify for a substantial reduction if it is imported for industrial manufacture. In Sweden there is a duty of 5.00 to 7.50 krona per 100 kilograms for orange juice without added sugar, according to the container size, and 30.00 krona per 100 kilograms for orange juices with added sugar. In the USA FCOJ imports are subject to a tariff of 34 cents per pound of solids and, as previously discussed in this chapter, FCOJ importers can use the duty drawback provision to recover 99 percent of the duty paid






37


if a similar quantity is exported. Japan uses quotas as a way to protect its domestic industry.

Preferential concessions are forms of favorable

treatment (tariff reductions or other forms) given to one country or to a group of countries. The EEC allows a 70 percent preferential duty reduction on orange juice imports from Algeria, Israel, Morocco and Tunisia, while other suppliers must pay the full amount. Italy and Greece, as producers of citrus and citrus products within the community, have benefited by the condition of free trade between members.

This treatment will soon be extended to Spain if it becomes a member of the EEC in the near future.

Other regional organizations and trading blocs7 with a large variety of agreements and commitments may also affect trade of orange juices. Cuba, for example, has been reported as a potential producer of orange juice in the future (Wolff, 1978) and, as a member of the COMECON will have preferential treatment in that market.


7European Free Trade Association (EFTA) includes Austria, Norway, Portugal, Sweden, Finland, Iceland and Switzerland. Council for Mutual Economic Assistance (COMECON) includes Eastern European nations, the Soviet Union, Mongolia and Cuba. Yugoslavia does not participate fully and Finland, Iraq and Mexico are "cooperant" members of COMECON. Other groups are the Latin America Free Trade Area (LAFTA), the Caribbean Free Trade Area (CARIFTA) and the Central American Common Market (CACM; Ryan & Tontz, 1978).















CHAPTER III


THE ECONOMIC MODEL

In this chapter efforts are devoted to setting up the conceptual development of the economic model of the orange juice industry. The goal is to develop the reasoning for a set of equations that summarize the most important market segments of the industry. The first section gives an overview of the underlying economic theory. A brief discussion of theory of consumer behavior is followed by a review of recent theoretical developments of exchange rate effects on commodity prices and trade. In the second section the economic model is formally specified with flow charts and sets of equations. This specification results from putting together important economic considerations from Chapter II, and theoretical guidelines, logical reasoning, experience and basic knowledge of the orange juice industry. Finally, the third section summarizes the estimation of the empirical model.


Theoretical Background

The analytical framework of this study can be described as a system of demand and price equations for orange juice. The set of demand equations (retail, wholesale, import or export) are derived, in the sense that they reflect consumer


38






39


preferences.1 In some markets the derived concept will also include a chain effect, for example, retail variables may be

in awholesale equation, or retail and wholesale variables may be in an import or export equation. In such cases, the derived demand relations are also partial reduced-form equations derived from a more complete structure. The price equations stem from the simultaneous nature of the system. This formulation attempts to capture the price transmission effects within different levels of a given market and across markets. Bredahl et al. (1979) emphasize the role of elasticity of price transmission in estimating the elasticity of export demand for a given commodity. Their major concern relates to the implications of government policies that insulate domestic consumers and producers from external price fluctuations.

The basic economic foundations for the price equations for product sources other than the USA is price leadership.2 The USA sets the price and prices from other sources are derived from it. The USA price equations are basically derived from Florida's wholesale prices which have as major

determinant measures of orange juice stocks and crop sizes. Exchange rates are included as price equation shifters in


1The underlying theories of consumer behavior (utility function and revealed preference) will not be discussed in detail in this study. Most Microeconomics textbooks discuss these theories. For example, see Henderson and Quandt, 1980; Russell and Wilkinson, 1979; Phlips, 1974; and Silberberg, 1978. For modern treatments of revealed preference theory see Uzawa, 1960; and Richter, 1966. 2For more detail on price leadership, see for example Russell and Wilkinson, 1979; Gould and Ferguson, 1980. For application of price leadership in the citrus industry see Buckley, 1977.






40


most of the structure. The role of exchange rate should be reviewed, not only for its importance as discussed in Chapter II of this study, but also because of the recent attempts to develop a theoretical framework explaining its effects on commodity prices and trade.

Most of the work on exchange rates, as economic policy variables, has been in the context of their impact on the balance of payments of a country. In recent years attempts have been made to formulate a theoretical framework that incorporates the role of exchange rate policy on commodity analysis in a less aggregate basis (Schuh, 1974; Kost, 1976). Both studies use the same economic framework (supply and demand analysis in an industry with export potential). Schuh concentrated relatively more on the role o the exchange rate within the USA economy with particular interest in the distribution of benefits of economic progress and adoption of new technologies. Kost, on the other hand, was more general in the sense that his discussion attempted to cover the effects of changes in exchange rates on commodity production, consumption, trade levels, and prices for any two trading partners.3 Only the effects of devaluation by exporting countries will be discussed here under Kost's theoretical framework.4

3
Empirical works have, in general, provided support for exchange rate effects on commodity analysis with their major differences on its magnitude. See for example Greenshields, 1974; Vellianitis-Fidas, 1976; Johnson et al., 1977; and Collins et al., 1980.

4For more detailed analysis and effects of other sources of exchange rate changes see Kost's paper.







41


Figure 2 summarizes these major effects under a

simplified set of assumptions for a given commodity. This figure has three basic segments describing the economic conditions for this special commodity in the exporting country (exporter), in the importing country (importer), and at the international level (trade sector). Its horizontal axes measure quantity in any unit and its vertical axes measure prices. Solid lines in Figure 2 show the classical equilibrium in these two economies after trade, under the assumption that currency ratios are one to one. PE and QT are equilibrium price and traded quantity, respectively. Broken lines show a situation of equilibrium after trade associated with a devaluation of the exporting country currency (say the Brazilian cruzeiro) relative to the importing country currency (say the USA dollar). These exchange rate changes (from a ratio of one to one to a ratio of two cruzeiros to one dollar) are shown in the importer segment on Figure 2. Its vertical axis was stretched out with the devaluation of the exporter's currency. Under a new price scale the structure of demand


5These assumptions are: (1) The model consists of a twocountry world or one country-rest of the world (potential exporter and importer), (2) competitive economic systems exist in both countries, (3) a single homogenous commodity is traded, (4) there are no transport costs and no trade barriers, and (5) the market for the single homogenous commodity can be specified by a single downward-sloping demand function and a single upward-sloping supply function for each country (Kost, 1976).









Cr $ Cr $ $


8- 8 8 14.0


7 7 7 3.5


6- 6 -6 3.0
SE

5 -Sd 5 2


4. 4 4 2.0
PE PE1 72
3- 3 11.
PEO PEO I

2P 2 2 1


1- 101
Ddd
$\D$
DdI 0, 0 .44.
'0 Q 0 QT0 QT1 Q 0 Q

Exporter Trade Sector Importer


Figure 2. Effects of currency devaluation by exporting country on equilibrium price and quantity for a given commodity.







43


and supply functions for the importing country will pivot up from their respective intersection with the horizontal axis such that internal equilibrium price (dollars) and quantity will remain as before. Demand for imports (DI) function in the trade sector also moves in the same fashion, given that it is derived from importer demand and supply structure. Since this structure did not change in the exporting country, then the derived supply for exports (SE) in the trade sector segment does not change either. The new equilibrium implies a relatively large traded quantity of of the commodity (QT1), a higher export price in the exporting country currency (PE1) and a lower import price expressed in the importing country currency (PI1). The impact of exchange rate effects on prices and quantities is determined in large degree by the elasticity of supply for exports and by the elasticity of demand for imports which in turn depend on the respective domestic demand and supply elasticities in the exporting and importing countries. In this particular case of devaluation by the exporting country, the role of the export supply elasticity is quite important. One can see that the effects of currency devaluation by an exporter will be relatively more on prices (or traded quantity) if the supply export elasticity is relatively more inelastic (or elastic). Since agricultural products are


6Another important determinant is the magnitude of the exchange rate changes.






44


traditionally quite inelastic, then it is very likely that exchange rate effects will be larger on prices than on quantity.

In this study it is assumed that total orange juice

supply is relatively fixed for both sources (Brazil and the USA). Given previous discussion and the fact that domestic demand for orange juice in Brazil is very small, as compared with large demand in the USA, it is reasonable to expect that a devaluation of the Brazilian cruzeiro will have a relatively larger effect on export price than on export quantity. The same change in the USA dollar is expected to result in a relatively larger effect on export quantity than on export price if the orange juice export supply elasticities are relatively more inelastic in Brazil and relatively less inelastic in the USA.


The Economic Model

The economic model is essentially a model that explains the trade flows and prices. Supply of fruit is assumed to be exogenous.7 Variables expected to influence the demand structure, but not incorporated in previous discussions, will also be included in the proper segment of the economic model. Two flow charts, together with four sets of 7This assumption stems from the nature of orange production. New orange groves may require four to five years before reaching commercial production capacity. This is certainly a major constraint on the USA and Brazil supply responses from changes in orange juice market conditions in a given year. Thus, a simplified assumption of fixed or practically fixed supply is quite reasonable, especially in this study that uses bimonthly data.






45


equations, describe the economic model for orange juices

in the USA, Canada, the EEC7 (seven countries of the

European Economic Community), and the European non-EEC3

(three European countries non-EEC members) markets.8


FCOJ and Other Orange Juices Market Structure in the USA

Relations and interrelations of the major variables

of the model describing the orange juice market structure

in the USA are shown in Figure 3 and Equations 1 through

16. The set of equations formally specifies the derivation

of 16 endogenous variables0 in behavioral relations

(Equations 1 through 4, Equation 7, Equations 9 through 11,

and Equations 13 and 14) or in identity relations

8
The Brazilian domestic market is not included in the model because of its small size (less than 5 percent of the orange juice production) and lack of systematic data.

9A variable superscript sign plus (+) or minus (-) indicates the expected directional relationship between variables of the equation. Variables without a sign do not have a basis for hypothesizing the relationship. The subscript t stands for time period (December-January 1971/72 to February-March of 1979). Unless otherwise specified, most of the nonmonetary variables are expressed in gallons of single strength or reconstituted juice equivalent per thousand persons. Monetary variables are expressed in USA dollars in terms of 1975 prices. For more details see Appendix E.

10Endogenous, or current endogenous or jointly dependent variables,are those variables whose values are determined within a set of equations. Predetermined variables, which can be subdivided into exogenous, lagged exogenous, or lagged endogenous, are those variables used as explanatory variables in the set, but which are completely determined outside the system (exogenous) or are past values of current endogenous variables













RPShSt AS t1 MPFic



MQFSB

RQOS + + RQFSt WQFFI WQOFi
+ + tD +

I I t --6


I t


RPOS t RPFS t WP)FF1 tt-1 WPOFit MPFSB







e.mmummmm. mmmm mmm atRxB

\ I% I WmkOt-RPOF
XQFCS+ + t t

\ / m .S~
WQsFFi WSkFSt -- X
tIIWkF W kSt

XQFESt XQFNSt+ XQFFtcOt M SI W~s Ft + WWkOt





LEGEND

Endogenous Predetermined Endogenous variables which '% Solid arrows indicate Broken arrows indicate variables variables derivation is not shown functional relations identity relations



Figure 3. Orange juice market structure in the USA.






47


(Equations 5, 6, 8, 12, 15 and 16).


RQFSt = fl (RPFS, RPShS+, RPOS+, iSi+, AS+, AS+


Zlt, Z2t, Z3t, Z4t, Z5t) (01)


RPFS = f (WPFF1 WPFF1 +) (02)
t 2 t t-1

WQFF1t = f3 (WPFF1t, WPOF1t, RQFS RQOSt,


Zlt, Z2t, Z3t, Z4t, Z5t (03)


WPFF1t = f4 (WSkFSt, WSkFSt-i, PdSt, PdSt-1'


Dt, Tt) (04)


WSkFSt = WQkFSt/WQFF1t (05)


WQkFSt = [(WQkFSt-1 + WQsFF1t + MQFSBt + MQFSOct


WQFF1 t) Pist (XQFCS t) PlCt (XQFESt)


P1Et (XQFNSt) P1Nt (XQFOcSt) Ploct]/

PiSt (06)


MQFSBt = f5 (SpBt, WQOFlt, MPFSOct, WSkFSt,


WSkFSt-1, WSkOSt, WSkOSt-1, Zlt, Z2t,


Z3t, Z4t, Z5t) (07)


SpBt = WPFF1t MPFSBt (08)






48


MPFSB = f (WPFF1 WPFF1+ WPOF1 +, WPOF1 ,
t 6 t t1 t t-1

PdSpt-6, RxBSt) (09)


RQOSt = f7 (RPOS, RPShS, RPFSt, isi, AS, ASt1


Zlt, Z2t, Z3t, Z4t, Z5 t) (10)


RPOSt = f (WPOFl, WPOF1t-1, RpOSt) (11)


RpOSt = RQObSt/RQOSt (12)


WQOFlt = f9 (WPOFl-, WPFF1 RQOS+, Zlt, Z2t,


Z3t, Z4t, Z5t) (13)


WPOF1t = f10 (WSkOSt, WSkOSt-1, PdSt, PdSt-1'


Dt, Tt) (14)


WSkOSt = WQkOS/WQOF1t (15)


WQkOSt = [(WQkOSt-1 + WQsOFlt WQOFlt) PiS t


XQOCSt) P1C t ]/P1St (16)


A description of all variables in the specification of the USA market relationships is presented in Table 11.

The demand structure for orange juices in this market is formulated separately for FCOJ (Equations 1 through 9) and for other orange juices (Equations 10 through 16). The underlying reasons stem from the importance of these two forms of orange juice marketed in the USA, as discussed






49


Table 11. Description of variables in the orange juice
market structure in the USA



Variables1 Description


Endogenous variables

RQFSt Retail quantity of FCOJ in the USA

RPFSt Retail price of FCOJ in the USA

WQFF1t Wholesale quantity of FCOJ in Florida

WPFF1t Wholesale price of FCOJ in Florida

WSkFS t Wholesale stock of FCOJ in the USA, expressed
in bimonths of Florida supplies

WQkFSt Wholesale quantity in stock of FCOJ in the USA

MQFSBt Import quantity of FCOJ by the USA from Brazil

SpB t Price spread between Florida's FCOJ wholesale
price (WPFF1 ) and the USA FCOJ import price
from Brazil (MPFSBt)

MPFSBt Import price of FCOJ by the USA from Brazil

RQOS t Retail quantity of other orange juices in the
USA

RPOSt Retail price of other orange juices in the USA

RpOSt Retail proportion of other orange juices
marketed in cardboard containers in the USA

WQOF1t Wholesale quantity of other orange juices in
Florida

WPOFlt Wholesale price of other orange juices in
Florida

WSkOSt Wholesale stock of other orange juices in the
USA, expressed in bimonths of Florida supplies

WQkOS t Wholesale quantity in stock of other orange
juices in the USA

(continued)






50


Table 11 (continued)



Variables1 Description


XQFCSt Export quantity of FCOJ to Canada from the USA

XQFES t Export quantity of FCOJ to the EEC7 countries
from the USA

XQFNS t Export quantity of FCOJ to the European nonEEC3 countries from the USA Predetermined variables

MQFSOc t Import quantity of FCOJ by the USA from other
countries

MPFSOc t Import price of FCOJ by the USA from other
countries

XQFOcSt Export quantity of FCOJ to other countries
from the USA

XQOCS t Export quantity of other orange juices to
Canada from the USA

RPShS t Retail price of frozen orange-flavored
synthetics and drinks in the USA

RQObS t Retail quantity of other orange juices in
cardboard containers in the USA

WQkFS t-1 Wholesale quantity in stock of FCOJ in the
USA in the previous period

WQsFFlt Wholesale quantity supplied of FCOJ in Florida

WQkOSt-1 Wholesale quantity in stock of other orange
juices in the USA in the previous period

WQsOF1 t Wholesale quantity supplied of other orange
juices in Florida

WPFF1t-1 WPFF1t lag by one period

WQOFlt-1 WPOF1t lag by one period

WSkFFlt-1 WSkFFlt lag by one period


(continued)






51


Table 11 (continued)



Variables1 Description

WSkOFlt-i WSkOFlt lag by one period

RxBSt Brazil exchange rate expressed in cruzeiros
per USA dollar

iSt Income per capita in the USA, seasonally
adjusted at annual rates, expressed in
thousand dollars

ASt, ASt-i Current and lagged advertising expenditures
(generic and brand) by the Florida Department of Citrus in the USA, expressed in dollar per
thousand persons

PdSt, PdSt-l Current and lagged estimates of orange
production in the USA, expressed in billion
boxes of 90 pounds

PdSpt-6 Estimates of orange production in the state
of Sao Paulo (Brazil) lag of six periods,
expressed in billion boxes of 90 pounds

PlSt, P1Ct, PlEt, PlNt and PlOct

Estimates of population in the USA, Canada,
the EEC7, the European non-EEC3 and in other
countries, respectively, expressed in million
persons

D t Dummy variables to account fot the effects of
the USA freeze in 1977. Dt is equal to one
for observations from February-March bimonth
of 1977 on, and zero otherwise

T tLinear time trend to account for time-related
changes in prices

(continued)






52


Table 11 (continued).



Variables Description


Zi t Five (i=1,.,5) dummy variables to account
for bimonth seasonal effects on demand. Zi
is equal to one for December-January bimont s,
and zero otherwise. Similarly, Z2 through
Z5t is equal to one for February-MArch, AprilMay, June-July and August-September bimonths,
respectively, and zero otherwise


1Unless otherwise specified non-monetary variables are expressed in gallons of single strength or reconstituted juice equivalent per thousand persons. Monetary variables are expressed in the USA dollar in terms of 1975 prices. Price variables are expressed in the USA dollars per gallon of single strength or reconstituted juice equivalent. A "t" subscript on a variable stands for "time period," which is from December-January bimonth of 1971/1972 to February-March bimonth of 1979. For more detail see Appendix E.







53


previously (Chapter II), and the existence of systematic data for both of them. There is no doubt about the significance of estimating demand relations for FCOJ. However, the quantity of other forms of orange juice marketed in the USA has increased considerably. For example, Florida's wholesale domestic sales of canned and chilled orange juices increased by almost 70 percent from 1972 to 1979 (Table C.2. Appendix C). Thus, the inclusion of a set of demand and price equations for other orange juice is required to complete the model in this market. The retail and wholesale sectors for FCOJ and other orange juices are described by Equations 1 through 6 and by Equations 10 through 16. The wholesale sector is specified only for Florida since this state accounts for almost 90 percent of the total oranges processed in the country (Florida Crop and Livestock Reporting Service, 1979) and detailed data are not available for the other states. USA imports11 are described by Equations 7 through 9. In the import segment only the demand structure for FCOJ imports is specified, since the USA does not import significant quantities of other forms of orange juices.

As shown by the flow chart in Figure 3 all sectors of the USA market are linked together so that changes in any variable are likely to have some effects on all variables determined by the system. It should be pointed out that


USA export components will be described in the formulation of the demand structure of the other markets as defined in this study.






54


wholesale prices, especially for FCOJ, and the FCOJ import price from Brazil, are variables that play major roles in the model. Wholesale prices are important determinants of retail prices in both forms of orange juices and wholesale prices of FCOJ together with the FCOJ import prices from Brazil determine the price spread, a key variable in this model. Another role of the domestic-import price relation would be in the derivation of the USA export prices which will be discussed in the next sub-section, where the FCOJ demand structures for other markets are conceptually formulated.

The USA market structure for orange juices is

conceptually quite similar to the model developed by Ward (1976a).12 However, it is different in the sense that it also includes the retail sector of FCOJ and other orange juices. In addition, the wholesale sector also addresses the demand structure for other orange juices, and domestic and import prices are endogenous. Another important difference is the simultaneous linkage of major supply sources of orange juices in the most important markets (the USA, Canada, the EEC7 and the European non-EEC3).

The remaining part of this sub-section will discuss the formulation of each equation in the system, as well as the expected directional effects of each variable.


12Ward's FCOJ model endogenously determines Florida's domestic demand and demand for imports (domestic and import prices are exogenous), and Florida's export quantities and prices to Canada and Europe.






55


Retail FCOJ

Retail quantity of FCOJ in the USA (RQFS t) is formulated as a function of its own price (RPFSt) prices of orangeflavored synthetics and drinks and other orange juices as substitute products (RPShSt and RPOSt), income (ISt), advertising (ASt and ASt-1),and seasonal variation in consumption (Zlt through Z5t) (Equation 1). The hypothesized signs are consistent with traditional neoclassical demand theory and previous FCOJ demand research by Tilley (1979) and Ward and Tilley (1980). The impact of advertising on demand is assumed to have a lag structure of one time period. The underlying assumption is that some of the effects of advertising on consumption are expected to be reflected at later periods. Lee (1980) reviewed the Florida Department of Citrus, Economic Research Department research program in the area of measuring advertising effectiveness. A common agreement is that the effects of advertising are not immediately perceived. That is, consumer responses to advertising have a decay or carryover effect over time. Since, in this model a time period aggregates two months (bimonthly data base), the specified structure of the advertising variables (ASt and AS t-l) will include four months, so that most of the effects of advertising on FCOJ retail demand are expected to be captured.

FCOJ retail price (Equation 2) is assumed to be a positive fucntion of current and lagged wholesale price (WPFF1t and WPFF1 t-). Within this time period (four months)






56


it is expected that market agents will have sufficient time to fully adjust retail prices to most changes in wholesale prices.


Wholesale FCOJ

The wholesale FCOJ demand structure is described by

Equations 3 through 5. Florida's wholesale quantity (WQFF1t) is hypothesized to be negatively related to its own wholesale price (WPFF1t ), and positively related to wholesale price of other orange juices (WPOF1t ), retail quantity of FCOJ (RQFSt), and retail quantity of other orange juices (RQOSt). Seasonal variation (Zlt through Z5t) in the wholesale demand for FCOJ is also expected. Underlying most of this formulation is the assumption that wholesale demand is derived from retail demand. Retail quantities (RQFSt and RQOSt) are expected to translate consumers' preferences to the wholesale sector. Retail quantity of other orange juices (RQOSt) is hypothesized to affect FCOJ wholesale demand (WQFF1 t) positively, since it can be reconstituted from bulk FCOJ.13

Wholesale price of FCOJ in Florida (Equation 4) is assumed to be determined by an estimated of wholesale stocks of FCOJ in the USA (WSkFSt), orange production in the USA (PdSt), a dummy variable to account for the effects


3It should be recalled that FCOJ, in this study, is defined so as to include all container forms of marketing the product. Florida's processors supply FCOJ in retail-size, bulk and institutional containers.
4An estimate because data on WSkFS determinant variables
t
are available only for Florida processors.






57


of the January of 1977 freeze (D t), and a linear time trend variable expected to capture the effects of time-related changes (Tt). WSkFSt and PdSt are assumed to be negatively related to wholesale prices under a lag structure. The FCOJ stock variable (WSkFSt) is defined by equations 5 and 6 (identities) as the ratio of FCOJ quantity in stocks in the USA at the end of the bimonthl5 to Florida's domestic movement (WQFFlt) for the same time period. Under this definition, WSkFSt is expressed in bimonths of domestic (Florida) supplies, such that the seasonal effect of stocks on wholesale prices can be accounted for. A specification that includes current and lagged FCOJ stock variables (WSkFSt and WSkFSt-1) is expected to capture most of its effects on FCOJ wholesale prices. Large stocks are expected to cause processors to reduce current prices in order to increase sales. Current and lagged orange production estimates (PdSt, 16
PdS t1) in the USA are also assumed to affect WPFF1 The
t-1 t
reasoning for it is that changes in expected production would lead processors to adjust current prices, given anticipated juice supplies. The freeze in 1977 (Dt) is expected to have a positive impact on wholesale prices, reflecting current and/or future reductions on supply.


15FCOJ quantity in stocks in the USA at the end of the bimonth (Equation 6) is determined by the quantity in stocks in the USA from previous bimonth plus current Florida domestic supply and the USA imports from Brazil and from other countries minus current Florida domestic demand and the USA exports to Canada, the EEC7, the European non-EEC3, and other countries.

16It is assumed that the wholesale price equation for FCOJ (Equation 4), as well as the wholesale price equation for






58


FCOJ Imports

USA import demand for FCOJ from Brazil is specified in Equation 7 through 9. In this formulation, Florida's wholesale prices (WPFF1t and WPOFlt) and the import price of FCOJ from Brazil (MPFSBt) are the major factors influencing the USA import demand equations.

The price spread (SpBt) between the FCOJ domestic

wholesale price in Florida (WPFF1t) and the import price from Brazil (MPFSBt) plays a major role, as suggested by Ward (1976a),in the determination of USA imports from Brazil (Equations 7 and 8). SpBt is expected to be positively related to imports, meaning that, as this price differential increases (decreases), quantity of FCOJ imported from Brazil is likely to increase (decrease). While the preceding

relationship determines the USA FCOJ import demand, the other variables in Equation 7 are assumed to be major import demand shifters. Wholesale quantity of other orange juices (WQOFlt) is expected to be positively reflected to FCOJ imports under the assumption that pressure from domestic wholesale sales of other forms of orange juices may lead to more FCOJ imports.



other orange juices (Equation 14), are both affected by changes on orange production in the USA as whole, even through the equations are for Florida's orange juices. The reasoning for such formulation is as follows. If Florida's orange production changes, the state supply of orange juices, and consequently their prices, is very likely to change, given that, in Florida, practically all orange production is processed. If orange production in other states changes, the most likely result would be changes in the market for fresh oranges.






59


Import price of FCOJ from other sources (MPFSOc t), as the price of an expected substitute product, is hypothesized to be positively related to imports from Brazil (MQFSBt). In this formulation stocks play a direct rolel7 on FCOJ imports. Imports are seen as the immediate source to raise stocks to some desired level, independent of future allocation. In this context stocks and imports move in opposite directions. Larger stocks should lessen the need for imports, while as stocks fall below some critical level, imports are viewed as immediately available supply. Even with stocks above the critical level, strong pressure from Canadian and European buyers of the USA product may lead to more imports in order to average down export prices. Bimonthly dummy variables (Zlt through Z5t) are included to account for seasonal influences on FCOJ imports.

It is assumed that current and lagged Florida FCOJ

wholesale price (WPFFlt, WPFFlt-1) are the major determinants of import prices from Brazil (MPFSBt) (Equation 9). In other words, to determine MPFSBt' Brazilian export agents and institutions first consider FCOJ market conditions in Florida, directly through market information and indirectly through actions of importers. WPOF1t and WPOF1t-1 are included in the import price equation because pressure on the domestic wholesale market of other orange juices is


17The indirect role of stock on imports comes from its effects on domestic prices (WPFF1t) and, consequently, on price spreads (SpBt) and on the level of imports.






60


expected to put upward pressure on prices in that market. Brazilian FCOJ exporters may react to this movement by arguing for higher prices, assuming that some FCOJ imports are reconstituted and sold in that market. Sao Paulo's orange production (PdSpt)18 enters the import price equation only under a lag of six periods (PdSpt-6), which is assumed to be the best formulation to capture the traditional negative effect of changing production on prices. This lag structure is directly associated with the fact that most Brazilian crop official estimates--as in the case of oranges-are done in a yearly basis. Production reports are more likely to influence prices in the next period of production and marketing. The last determinant of import prices, as in Equation 9, is cruzeiros per USA dollar (RxBSt). RxBSt is expected to be negatively related to MQFSBt


Retail and Wholesale Other Orange Juices

Most of the discussions with regard to FCOJ in the USA domestic markets can be fully extended to the formulation of demand structures for other forms of orange juices as described by Equations 10 through 16. In Equation 11, retail prices of other orange juices (RPOS t) are functions of WPOFlt, WPOFt-1 and of the proportion of orange juices sold in cardboard containers (RfOSt), as defined by Equation 12.


18In this case, only orange production in Sao Paulo is included as that state is the source of all FCOJ exports.






61


It is expected that as this proportion increases retail prices will decrease because cardboard containers are cheaper than glass and cans. It should also be pointed out that the USA does not import other orange juices. In the determination of the wholesale stocks of other orange juices in the USA (SWkOSt), as described by Equation 15, the supply side (see Equation 16) is comprised only of the USA quantity in stocks from the previous period and current domestic supply (WQkOS t-1 and WQsOFl t). The demand side includes Florida's domestic demand (WQOFl t) and exports to Canada (XQOCSt). USA exports of other orange juices to Europe are of relatively minor importance.


FCOJ Market Structure in Canada

The conceptual structure of the economic model showing the relations and interrelationships of variables for the Canadian and European markets is described by the flow chart in Figure 4. These segments of the model were simplified to include only the international trade sector and, specifically, only FCOJ exports to these markets from Brazil and from the USA. In most of these markets the domestic producing sector of orange juices is quite small or non-existent so that it is reasonable to assume that all domestic demand forces can be captured by the trade demand structures. Thus, the system of equations resulting from this formulation is a partially reduced form system. It is reduced in the sense that variables, characteristically defined as most likely to have a direct effect on wholesale (orange production) or on

















zK,
i 5








MPFC~cXPCS t
+

ACt + XOFESt A Et XOFEBt
XOFCBt A C XOFCS t o r A N o r o r
+ t-1 +XOFNSt t XOFNBt


Pdit J5 WPFFI
tt T


+ LE + 0
V M 1S XPFFESt XPFEBt
XPFCBt + I XPFCS t or I+ or
t /XPFNS XPFNBt
+ ++ t l


RxBCt PdSpt-6

Rx~tMPFSOct or or o RxSNtRxt


LEGEND
Endogenous Predetermined Endogenous variables which 1 Solid arrows indicate
variables 0 variables derivation is not shown j functional relations



Figure 4. Orange juice market structure in Canada, the EEC7 and the European non-EEC3 countries.






63


retail (income and advertising expenditures) sectors, are included in the international trade equations. It is partial because the set of equations still contains endogenous variables. The simplification that only Brazil and the USA export FCOJ to these markets is possible because other sources of orange juices have been supplying relatively small quantities.

This formulation of Canadian and European demand for FCOJ from Brazil and the USA is in some way similar to the model developed by Tilley and Lee (1981) for the Canadian retail and import level purchases and prices for orange juices. The similarities are translated in the common way of treating some variables as endogenous or as predetermined. For example Canadian FCOJ trade with the two suppliers (quantities and prices) is determined by the system. However, in the Tilley and Lee model Canadian import prices from Brazil and the USA are exogenous. Another difference is that their model includes retail demand structures (quantity and price) for all orange juices, as well as an aggregated demand for other sources of Canadian imports.

Major roles in the specification are again played by Florida's FCOJ wholesale price (WPFF1t) and by the import price of FCOJ from Brazil (MPFSBt). Together their important role is through the price spread (SpBt) in the determination of USA imports from Brazil and the impact of these imports






64


on the USA export prices. It was shown that widening price spreads associated with declining import prices from Brazil or higher domestic prices are likely to result in more imports into the USA. Higher price spreads may lead to a greater export price discounting because more imports are available to average down export prices, and Brazilian imports also have a lower relative price. The net effect of a wider spread on USA exports will thus depend on the degree of discounting and the substitution relationships in the import markets.

In the Canadian market four endogenous variables are determined within the structure (Equations 17 through 20) and there are 16 other variables in the system. Among these variables, WPFFlt and MPFSBt are endogenously determined in the USA market. The other variables are predetermined. The following equations represent the FCOJ market structure in Canada and Table 12 has a description of all variables of the structure.

XQFCBt = f11 (XPFCBt, XPFCSt, MPFCOct, XPOCSt,


IC AC ACt-i1, Zlt, Z2t, Z3t, Z4t, Z5t) (17)


XQFCBt = f (XPFCS PdSp RxBCt) (18)
t 12 t Pt-6 t

XQFCSt = fl3 (XPFCSt, XPFCBL, MPFCOc, XPOCS ICt,


AC AC +, Z i, Z2t, Z3t, Z4t, Z5 (19)
t t1 tt t t

+ + +
XPFCSt = f (WPFF1, MPFSBt, MPFSOc RxSCt) (20)
t 14 t t t t






65




Table 12. Description of variables in the FCOJ market
structure in Canada



Variables1 Description


Endogenous variables

XQFCBt Export quantity of FCOJ to Canada from Brazil

XPFCBt Export price of FCOJ to Canada from Brazil

XQFCS t Export quantity of FCOJ to Canada from the
USA

XPFCSt Export price of FCOJ to Canada from the USA

WPFF1t Wholesale price of FCOJ in Florida

MPFSBt Import price of FCOJ by the USA from Brazil

Predetermined variables

MPFSO t Import price of FCOJ by the USA from other
countries

MPFCO t Import price of FCOJ by Canada from other
Ct countries

XPOCSt Export price of other orange juices to Canada
from the USA

RxBCt Brazil exchange rate expressed in cruzeiros
per Canadian dollar

RxCS t The USA exchange rate expressed in cents of
dollar per Canadian dollar

IC t Income per capita in Canada, seasonally
adjusted at annual rates, expressed in
thousands of Canadian dollar

ACt, AC t- Current and lagged advertising expenditures of the Florida Department of Citrus in Canada,
expressed in dollars per thousand persons

(continued)






66

Table 12 (continued)



Variables1 Description


PdSp t-6 Estimates of orange production in the state
of Sao Paulo (Brazil) lag of six periods expressed in billion boxes of 90 pounds

Zit Five (i=l,.,5) dummy variables to account
for bimonth seasonal effects on demand. Zlt
is equal to one for December-January bimonths,
and zero otherwise. Similarly Z2t through
Z5t is equal to one for February-March,
April-May, June-July and August-September
bimonths, respectively, and zero otherwise.


1Unless otherwise specified non-monetary variables are expressed in gallons of single strength juice equivalent per thousand persons. Monetary variables are expressed in the USA dollar in terms of 1975 prices. Price variables are expressed in the USA dollars per gallon of single strength juice equivalent. A "t" subscript on a variable stands for "time period," which is from December-January bimonth of 1971/1972 to February-March bimonth of 1979. For more details see Appendix E.






67


Brazilian FCOJ Exports to Canada

In Equation 17, quantity of FCOJ exported to Canada from Brazil (XQFCBt) is hypothesized to be negatively related to export price (XPFCBt), and positively related to export price of FCOJ to Canada from the USA (XPFCSt), the import price of FCOJ by Canada from other countries (MPFCOct), the export price of other orange juices to Canada from the USA (XPOCSt), disposable income in Canada (ICt), and advertising expenditures of the Florida Department of Citrus in Canada (ACt). XPFCSt, MPFCOct and XPOCSt enter the equation as competing substitute products differentiated by the supply source and/or the product characteristics. Florida's advertising in Canada (AC t) is designed to promote orange juice from Florida. If, due to Florida's advertising, Canadian imports from the USA increase, the likely result would be a reduction in Canada's imports from other sources, assuming that other factors do not change. However, if the net effect of Florida's product advertising programs in Canada promotes orange juice in general, then it is possible that Florida advertising may have a positive effect on Canadian demand for FCOJ from other suppliers. Current and lagged advertising expenditures (ACt and ACt-1) are included to test these hypotheses. The remaining variables (Zlt through Z55 account for seasonal effects.

In the Brazilian export price equation to Canada

(Equation 18), the USA FCOJ export price to Canada (XPFCS t) is included because it is expected that the USA acts as price






68


leader. As the price changes, the Brazilian price (XPFCBt) changes in the same direction. Sao Paulo's orange production (PdSpt) and Brazil's exchange rate for the Canadian dollar (RxBC t) are expected to be negatively related to XPFCBt. Orange production in Sao Paulo enters the equation as a six-period lag (PdSpt-6) for the same reasons as in previous discussions.


USA FCOJ Exports to Canada

Export quantity of FCOJ to Canada from the USA (Equation 19) is expected to be negatively related to its own price (XPFCS t). It also has six export demand shifter variables (XPFCBt, MPFCOct, XPOCSt, ICt, ACt, ACt-1) that are expected to be positively related to XQFCSt. All the export demand shifter variables are expected to behave as in previous discussions. XPFCBt, MPFCOct and XPOCSt are prices of competing substitute products positively related to XQFCSt. ICt is disposable income in Canada, and ACt and ACt-1 are Florida's promotional expenditures in Canada, both positively shifting the export demand.

In Equation 20, the USA FCOJ export price to Canada (XPFCS) is assumed to be positively related to Florida's domestic wholesale price of FCOJ (WPFF1t ) and to the import prices of FCOJ by the USA from Brazil (MPFSB t) and from other countries (MPFSOc t). This specification recognizes, explicitly, the role of USA imports of FCOJ in export price determination and that USA prices may be affected by competitive prices of product from other sources.






69


FCOJ Market Structure in Europe

The European market structure is hypothesized to be similar to the Canadian market structure. There are only two suppliers (Brazil and the USA) and the structure is formulated only for FCOJ. The structure embraces the same partial reduced form concept and the key variables are the same (WPFFlt, MPFSBt and SpBt).

The economic background on the relevant markets for orange juices presented in Chapter II stressed important differences in major European FCOJ markets. On one side the European Economic Community (EEC) countries were said to be quite important due to their market size and specific economic policies with respect to orange juice imports into those countries. On the other side, European countries not EEC members were said to be also important because of their high rate of per capita consumption of orange juices, market sophistication as far as juice quality and packing procedures are concerned, and the fact that, as non-EEC members, they do not have the same structure of protective restrictions and preferential concessions as the EEC members. In order to deal with these different economic aspects, the FCOJ demand structures in Europe are formulated for two groups of countries. The EEC7 group (West Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark and France) represents EEC members. The European non-EEC3 group (Sweden, Norway and Finland) represents other European non-EEC members.






70


The EEC7 and European non-EEC3 market structures are described by Equations 21 through 28.

XQFEBt = f (XPFEB-, XPFES+, T+, AE+, Zit, Z2t,


Z3t, Z4t, Z5t) (21)


XPFEB = f (XPFES PdSp RxBE ) (22)
t 16 t t-61

XQFESt = f (XPFES-, XPFEB, T+, AE+, Zlt, Z2t,


Z3t, Z4t, Z5t) (23)


XPFESt = fl8 (WPFF1+, MPFSBt, MPFSOc RxSEt) (24)


XQFNBt = fl9 (XPFNBt, XPFNS, T+, AN +, Zlt, Z2t,


Z3t, Z4t, Z5 t (25)


XPFNBt f 0 (XPFNS PdSp RxBN ) (26)
t 20 t t-6

++ +
XQFNSt = f21 (XPFNSt, XPFNB TI+, ANt, Zit, Z2t,


Z3t, Z4t, Z5t (27)


XPFNSt = f22 (WPFF1, MPFSBt, MPFSOc RxSN) (28)


Eight endogenous variables are determined by the

structure. Wholesale prices in Florida (WPFFlt) and the USA import price from Brazil (MPFSBt) are endogenous variables in the system determined in the USA market. The other variables are predetermined. A description of all variables of the FCOJ market structure in the EEC7 and the European non-EEC3 countries is presented in Table 13.





71


Table 13. Description of variables in the FCOJ market
structures in the EEC7 and the European non-EEC3
countries



Variables1 Description


Endogenous variables

XQFEB t Export quantity of FCOJ to the EEC7 from
Brazil

XPFEBt Export price of FCOJ to the EEC7 from Brazil

XQFES t Export quantity of FCOJ to the EEC7 from the
USA

XPFESt Export price of FCOJ to the EEC7 from the USA

XQFNB t Export quantity of FCOJ to the European nonEEC3 from Brazil

XPFNB t Export price of FCOJ to the European non-EEC3
from Brazil

XQFNS t Export quantity of FCOJ to the European nonEEC3 from the USA

XPFNSt Export price of FCOJ to the European non-EEC3
from the USA

WPFF1t Wholesale price of FCOJ in Florida

MPFSBt Import price of FCOJ by the USA from Brazil

Predetermined variables

MPFSOct Import price of FCOJ by the USA from other
countries

RxBEt Brazil exchange rate index relative to the
EEC7 countries

RxBN t Brazil exchange rate index relative to the
European non-EEC3 countries

(continued)





72


Table 13 (continued).



Variables Description


RxSE t The USA exchange rate index relative to the
EEC7 countries

RxSN t The USA exchange rate index relative to the
European non-EEC3 countries, expressed in
dollars per thousand persons

AN t Advertising expenditures of the Three-Party
Program in the European non-EEC3 countries,
expressed in dollars per thousand persons

PdSp t-6 Estimates of orange production in the state
of Sdo Paulo (Brazil); lag of six periods
expressed in billion boxes of 90 pounds.

T t Time trend variables as a proxy variable for
income per capita in the EEC7 and the
European non-EEC countries

Zlt Five (i=l,.,5) dummy variables to account
for bimonth seasonal effects on demand. Zlt
is equal to one for December-January bimonths,
and zero otherwise. Similarly Z2t through
Z5t is equal to one for February-March,
April-May, June-July and August-September
bimonths, respectively, and zero otherwise.


1Unless otherwise specified non-monetary variables are expressed in gallons of single strength juice equivalent per thousand persons. Monetary variables are expressed in the USA dollar in terms of 1975 prices. Price variables are expressed in USA dollars per gallon of single strength juice equivalent. A "t" subscript on a variable stands for "time period" which is from December-January bimonth of 1971/1972 to February-March bimonth of 1979. For more details see Appendix E.





73


Brazilian FCOJ Exports to the EEC7

Export quantity of FCOJ to the EEC7 countries from Brazil (Equation 21) is hypothesized to be negatively related to its own price (XPFEBt) and positively related to the USA export price to the same market (XPFESt). Tt is a time trend variable used as a proxy for disposable income in the EEC7 market, which is also expected to be positively related to demand for FCOJ from Brazil. AEt stands for advertising expenditures in the EEC7 market and it includes only financial support of the Three-Party Program which is designed to promote Florida's citrus in Europe. Three-Party Program data are available only on a yearly basis and AE t accounts for advertising expenditures in the seven countries of the EEC7 market. The impact of the program expenditures may be positive or negative, depending on whether the promotion of FCOJ from Florida causes a substitution of the Florida product for the Brazilian product, or creates an umbrella effect that causes orange juice from both the USA and Brazil to be purchased. Zlt through Z5t are seasonal measures of demand with the same objectives and expected behavior as in previous market demand structure specifications.

Equation 22 describes the derivation of Brazil's FCOJ export price to EEC7 (XPFEB t). XPFEBt is expected to be positively related to the USA export price of FCOJ to the same market. Orange production in Sao Paulo enters the equation under the same specification as in the previous case (PdSpt-6) and with the same negative expected relationship.





74


Exchange rate in this case is summarized as an index value of the Brazilian cruzeiro against currencies of EEC7 countries.19 Its effect on the export price from Brazil to EEC7 (XPFEBt) is expected to be positive, since the index is an inverse function of the exchange rate values of individual countries. For example, if exchange rates associated with individual countries increase (devaluations), then the index is expected to decrease as would the corresponding export price.


USA FCOJ Exports to the EEC7

The USA export quantity of FCOJ to the EEC7 (XQFESt),

as specified by Equation 23, has the same logical formulation as discussed in the Brazilian case, with the only difference being that AEt now is expected to have a positive effect on the USA FCOJ export demand to the EEC7 market.

USA export price to the EEC7 (Equation 24) is hypothesized to be positively related to Florida's FCOJ wholesale prices (WPFF1t ), to the import prices of FCOJ by the USA from Brazil and from other countries (MPFSBt and MPFSOct), and to the USA exchange rate index for the EEC7 countries (RxSEt). Florida's FCOJ wholesale price (WPFF1t) has, as already discussed, a direct and an indirect role in the derivation of export prices. Directly, since an increase in the wholesale price is likely to lead to an increase in export


19For more details on the construction of the exchange rate index see Appendix D.





75


prices, and indirectly through price spreads in the derivation of FCOJ imports into the USA that may be used to average down export prices. This last rationale also explains the role of FCOJ import prices by the USA in the export price equation.


Brazilian and USA FCOJ Exports to the European non-EEC3

The formulation of the European non-EEC3 market

structure for FCOJ from Brazil and USA is conceptually the same as in the case of the EEC7 market. Equations 25 through 28 are the same as 21 through 24 except that N in these formulations means the European non-EEC3 countries.


Estimation

This section provides a brief discussion of a system of simultaneous linear equations and of its major elements, the procedure used to estimate the empirical model, and a test of model validation.

From previous discussions in this chapter with regard to the theoretical setting and model formulation the simultaneous nature of the economic model of this study is quite clear. Most equations in the system include more than one endogenous21 variable. This condition introduces a 20Most of this discussion is organized from Goldberger, 1964; Kennedy, 1979; Kmenta, 1971; and Maddala, 1977. 21In the economic model there are 28 endogenous variables as defined by equations 1 through 28 (left hand side variables) and, among the explanatory variables (right hand side variables), 41 are predetermined variables (the ones that are explanatory, but not current endogenous variables).







76


violation in the set of assumptions of the classical linear regression model. If endogenous variables are used as explanatory variables it can be proved (Goldberger, 1974, p. 305-306) that they are contemporaneously correlated with the error term in all other equations in the system. As a result, the ordinary least squares (OLS) estimator would be biased, even asymptotically, and, consequently, an alternative estimator should be used.

By adding an error term to each behavioral equation of the system formulated in the previous section (called economic model), the econometric model of this study is formally specified in the so-called structural form. In matrix notation, the whole structure can be written as:


SY(t) + FX(t) = U(t) (29)


where 8 is the M by M matrix of coefficients of the endogenous variables; Y(t) is the M by 1 row vector of the tth observation on the endogenous variables; r is the M by K matrix of coefficient of the predetermined variables; X(t) is the K by 1 row vector of the tth observation on the predetermined variables; and U(t) is the M by 1 row vector of the tth (unobserved) values of the error terms.

Each error term should satisfy the assumptions of the classical linear regression model, that is, in matrix notation,

U(t) = N(0,0), (30)

E[U(t.)U'(s)] = 0, t / s (31)





77


and, E[U(t)U'(s)] = P, t=s (32)

where D is the M by M variance-covariance matrix (nonnegative definite)22 of the structural error term. The underlying assumptions imply that each structural error vector (a) is normally distributed with mean zero and constant (homoskedastic) unknown variance and (b) is nonautoregressive, implying that error terms are contemporaneously uncorrelated.

If the structural-form of the system (Equation 29) is solved for the endogenous variables, that is, by expressing endogenous variables in terms of predetermined variables and error terms, the result is the so-called reduced-form of the system (Equation 33 or 34).23


Y(t) = -C 1lX(t) + $-U(t) (33)


Y(t) = HX(t) + V(t) (34)

Equations 33 and 34 are called, respectively, restricted and unrestricted versions of reduced-form equations of the system. If structural-form estimates are used to derive the reduced-form (Equation 33), restrictions imposed on the structural-form formulation are accounted for through the estimates of 6 and P. The same is not necessarily true if


22In the case of existence of identities, (D refers only to the equations that are not identities. 23In solving for reduced form, the matrix 6 in Equation 29 is implicitly assumed to be a nonsingular matrix.





78


the reduced-form parameters are estimated directly on the predetermined variables (Equation 34). From a direct comparison of these two equations, if follows that:


H = R -1 (35)


and


V(t) = i-1U(t) (36)


where H is the M by K matrix of reduced-form coefficients of predetermined variables, and V(t) is the M by 1 row vector of the tth (unobserved) values of the reduced-form error terms.

The variance-covariance matrix (M) of the reduced-form error terms can be proved to be given by


T = -1 -(1)' (37)

It should be pointed out that in the reduced-form system each endogenous variable is expressed in terms of only predetermined variables and an error term, after accounting for the interdependence among current endogenous variables. Thus, a given reduced-form coefficient indicates the total effect of a change in its respective predetermined variable on the corresponding endogenous variable, assuming that other variables are held constant. These reduced-form coefficients are called long-run multipliers associated with the model (Kennedy 1979, p. 107). In contrast, a structuralform coefficient indicates only its variable partial direct




79


effect on the respective dependent endogenous variable. However, estimates of the structural-form coefficients are quite important for identifying the underlying economic hypotheses used as major support in developing the model. Thus, depending on a study's objectives, different strategies can be used to estimate reduced-forms--directly from values of predetermined variables as in Equation 34, or deriving it, as in Equation 33, by first estimating structural-form parameters. Associated with these strategies is the identification problem in simultaneous equation systems. It is a mathematical problem of going from estimates of the reduced-form structure back to meaningful estimates of the structural-form parameters. An equation can be said to be overidentified, or exactly identified, or underidentified by zero restrictions on certain parameters, if there are more than enough, or exactly enough, or less than enough predetermined variables excluded from the given equation to act as instruments of endogenous explanatory variables used in that equation. By applying the usual tests for identification (rank and order conditions), the structuralform equations of this study were found to be overidentified.

Various procedures are suggested in the literature to estimate a system of simultaneous equations. Based on the objectives of this study, the problems associated with the


A summary of these procedures is presented by Kennedy, 1979, p. 112-126.





80


formulation of its model,25 the nature of the model identification problem, and major characteristics of available procedures, the two-stage least squares (2SLS) procedure was chosen to estimate the parameters of the over-identified structural-form equations of the system. This procedure not only gives consistent estimators of the structural parameters, but studies have also shown it to have small-sample properties superior on most criteria to all other estimators, in addition to being quite robust (Kennedy 1979, p. 115). However, 2SLS estimators are not invariant with respect to the endogenous variable which is normalized. One would like this property in a simultaneous equation estimator. Basically, the 2SLS procedure requires that, in the first stage, each endogenous variable acting as explanatory variable be regressed on all the predetermined variables of the system (estimation of the reduced-form version) and their estimated values are calculated. In the second stage these estimated values (called instrumental variables) of endogenous variables and the included exogenous variables are used in an OLS regression.


25There are some errors of specification in the formulation of the structural form version of the model, especially in the case of Canadian and European market structures, where lack of systematic data for retail and wholesale sectors forced the formulation of only import equations. Available retail and/or wholesale variables were used as explanatory variables of import equations. Single-equation estimation procedures are said to keep these misspecification problems within their source in the system, rather than to incorporate them in the estimates of all the structural parameters as in the case of systems methods procedures (Kennedy, p. 116).





81


For testing how well values predicted by the model

conform to observed data, a test based on the ratio of the root mean square error to the respective observed mean value of endogenous variables for different periods within the data range is used. For a given endogenous variable and the ith time period, the ratio (RN.) is given by
1

Ni 2 1
RM. = t t-A /M (38)
Kt=1 i
N.i


Where, Pt and At stand, respectively, for predicted (forecasted) and actual (observed) values; N. is the number of
1
observations in the ith time period; and M. is the corre1
sponding observed mean value of the endogenous variable. This ratio will give an idea of the magnitude of the forecast error relative to the respective actual mean value for the endogenous variable in consideration.













CHAPTER IV

EMPIRICAL RESULTS AND IMPLICATIONS

This chapter is organized in two major parts with the common goal of discussing the empirical results and implications of estimates of the statistical model. In the first section, 2SLS estimates of the structural-form equations are presented and discussed. In the second section, estimates of the derived reduced-form equations are presented and discussed. Validation tests of the model in its derived reduced-form version are also presented and discussed in the second section.


Estimates of the Structural-Form Equations

In this section, the structural-form estimates of the model are presented and discussed for each of the four markets in this study. Special emphasis is given to the implications of these results within a given market context and to the market supply sources. Economic analysis in this section is intended to evaluate the underlying theoretical framework of the model, as well as particular implications of the results. Comparisons with other findings are also made, whenever possible.

Explanatory variables are listed in the tables as the respective structure where theoretically specified in the 82





83


model. This, however, does not necessarily mean that all of those variables are included in the final estimates of the equation.


The USA Market

Most of the structural-form equations were estimated as stated in the theoretical formulation in Chapter III. However, some variables were deleted from specific equations due to multicolinearity, plausibility of the respective signs, and/or the magnitude of a given estimate relative to its standard error. In this particular market, only the equations concerning FCOJ imports from Brazil (Equations 8 and .9), and the equations for the wholesale level of other orange juices in Florida (Equations 13 and 14) deviate from the theoretical specification. Practically, all signs of the estimated coefficients were as theoretically hypothesized.2 Most of the estimated coefficients are considerably larger than their estimated standard errors.


Retail FCOJ

Structural-form estimates for retail FCOJ demand (RQFSt) and the price (RPFSt) equations are presented in Tables 14 and 15. Except for RPOSt in the FCOJ retail demand equation,

1t
1Multicolinearity was detected based on the estimates of the correlation matrices of the parameters. Only extremely high levels of multicolinearity were regarded as major problems, given that 2SLS has been shown to be quite insensitive to this problem (Kennedy 1979, p. 115). 2Means, standard deviations and coefficients of variation for variables used in the model are presented in Appendix F.





84


Table 14. Structural-form estimates of the FCOJ retail
demand equation in the USA market (RQFSt,
Equation 1)


Explanatory Standard
variables a oeicie error

Intercept 204.6033 238.2164

Endogenous
variables

RPFSt -98.6950 68.8457

RPOS -112.2447 59.4951
t
Predetermined
variables

RPShS 128.9164 73.3681
t
iS 62.1050 28.7421
t
AS 1.4025 1.1662
t
+
AS .8982 .8695
t-1
Zlt 17.5453 8.4340

Z2t 14.2254 8.2063

Z3t -10.2295 7.4022

Z4t -20.6681 8.3289

Z5t -10.9591 7.3398


1Variable superscript indicates the a priori expected sign.





85


Table 15. Structural-form estimates of the FCOJ retail price
equation in the USA market (RPFSt, Equation 2)


Explanatory Standard
vaiaes1 Coefficientero
variables error

Intercept .4532 .0801

Engodenous
variables

WPFF1 .8201 .1780
t
Predetermined
variables

WPFF1 .2638 .1759
t-l


1Variable superscript indicates the a priori expected sign.





86


all other coefficients were consistent with expectations. The negative effect of prices of other orange juices (RPOSt) on retail FCOJ demand (Table 14) does not necessarily imply a departure from the theoretical setting where a positive response was expected. Similar problems of consistency in orange juice demand estimates were reported by Tilley (1979), Ward and Tilley (1980), and Malick (1980). A common agreement of these studies is that no statistical substitution evidences were found between FCOJ and chilled orange juice (COJ). Tilley, and Ward and Tilley also found that canned single strength orange juice (CSSOJ) was negatively related to FCOJ demand.3 The net effect of a change in the price of any commodity (P.) on the quantity demand of any other good (Qi) can be decomposed into a substitution effect and an income effect (Phlips,1974, p. 40-47).4 Since the income effect of price changes is not necessarily symmetric and that it will be larger for a commodity which takes a larger proportion of total expenditures (Tomek & Robinson,1981, p. 56), then the negative effect of RPOSt on the demand for FCOJ does not necessarily invalidate the expected substitute characteristic between the two products. This result should also be


It should be recalled that in this study orange juices are grouped as frozen concentrated (FCOJ) and other orange juices. The latter includes all other forms of orange juices but FCOJ. At retail level, other orange juices group is basically chilled and canned single strength (COJ and CSSOJ). 4In terms of partial derivatives such decomposition can be written as Qi/DPj = Kij Qj (aQi/ Y), where Kij is the substitution effect and Qj (DQi/DY) is the income effect. Note that Qj ( Qi/ Y) is not necessarily equal to Qi (9Qj/DY).




Full Text

PAGE 1

AN ECONOMETRIC MODEL OF INTERNATIONAL TRADE OF FROZEN CONCENTRATED ORANGE JUICE BY LUIZ JOSE MARIA IRIAS A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1981

PAGE 2

Copyright 1981 by Luiz Jose Maria Irias

PAGE 3

To my wife, Raulina, my daughters, Ana Claudia and Ana Cristina, and my parents.

PAGE 4

ACKNOWLEDGEMENTS The author wishes to express his gratitude and appreciation to Dr. Daniel S. Tilley, chairman of the supervisory committee, for his guidance, patience, encouragement and support in the preparation of this dissertation. He is quite grateful to Dr. Jonq-Ying Lee, cochairman of the supervisory committee, for his valuable assistance, especially with the computer programs. Warm appreciation and thanks are also extended to Dr. Max R. Langham and Dr. Frederick O. Goddard, members of the committee, for their time, assistance and guidance. He was very fortunate in having this committee. Working with them was a unique experience full of intellectual stimulation. They taught him a great deal. Special recognition is given to Dr. w. w. McPherson, graduate coordinator in the Food and Resource Economics Department, for his patience, assistance and guidance during the author's study program at the University of Florida. The author is also very proud of having Dr. McPherson attending his final examination. The author is especially grateful to the Empresa Brasileira de Pesquisa Agropecuaria, the Ford Foundation and the Florida Department of Citrus for providing financial support for the author's graduate study. iv

PAGE 5

April G. Burk merits special recognition for her dedication in typing extensive early drafts and the final version. Likewise, Susan Howard deserves thanks for her excellent art work. The author also appreciates the wonderful experiences and good times shared with colleagues, friends, staff, and faculty of the University of Florida's Food and Resource Economics Department and the Economic Research Department, Florida Department of Citrus. Finally, and most importantly, the author is especially grateful to his wife, Raulina, for her continued support and understanding during these years of study. V

PAGE 6

TABLE OF CONTENTS ACKNOWLEDGEMENTS. . . . . . . . . iv LIST OF TABLES. . . . . . . . . . ix LIST OF FIGURES. . . . . . . . xii ABSTRACT ......... ...... .............. ...... . xiii CHAPTER I PROBLEM SETTING ............... . . . . 1 Statement of the Problem......................... 1 Objectives. . . . . . . . . . 12 II AN ECONOMIC PROFILE. . . . . . . 14 Trade Flows. . . . . . . . . . 14 Structural and Institutional Factors ........... 26 Import and Export Incentives ................... 27 Market Promotion Programs...................... 30 Exchange Rate Policies ....................... 33 Trade Restrictions and Concessions ............. 35 III THE ECONOMIC MODEL............................... 38 Theoretical Background ........................... 38 The Economic Model. . . . . . . . 44 FCOJ and Other Orange Juices Market Structure in the USA................................... 45 Retail FCOJ. . . . . . . . . 5 5 Wholesale FCOJ.. . . . . . . . 56 FCOJ Imports . . . . . . . . 5 8 Retail and Wholesale Other Orange Juices ..... 60 FCOJ Market Structure in Canada ................ 61 Brazilian FCOJ Exports to Canada ............. 67 USA FCOJ Exports to Canada ................... 68 vi

PAGE 7

TABLE OF CONTENTS--Continued CHAPTER Page FCOJ Market Structure in Europe ................ 69 Brazilian FCOJ Exports to the EEC? .......... 73 USA FCOJ Exports to the EEC? ................. 74 Brazilian and USA FCOJ Exports to the European Non-EEC3 .......................... 75 Estimation. . . . . . . . . . 7 5 IV EMPIRICAL RESULTS AND IMPLICATIONS ............... 82 Estimates of the Structural-Form Equations ....... 82 The USA Market. . . . . . . . . 8 3 Retai 1 FCOJ. . . . . . . . . 8 3 Whole sale FCOJ. . . . . . . . 8 8 FCOJ Imports from Brazil..................... 93 Retail and Wholesale Other Orange Juices .... 98 The Canadian Market ............................ 105 Brazilian FCOJ Exports to Canada ............. 106 USA FCOJ Exports to Canada .................. 111 The European Markets ....................... 117 Brazilian FCOJ Exports to the EEC? ........... 118 USA FCOJ Exports to the EEC7 ................. 122 Brazilian FCOJ Export to the European Non-EEC3 ................................... 128 USA FCOJ Exports to the European Non-EEC3 .... 132 Estimates of the Derived Reduced-Form Equations .. 137 Impact of Exchange Rate ........................ 149 Impact of Orange Production .................... 151 Impact of Freeze ............................... 153 Tests on Performance of the Estimated Model .... 155 V SUMMARY AND CONCLUSIONS Summary. . . . . . . . . . . 15 9 Major Conclusions ... ............................ 162 Suggestions for Future Studies ................... 168 vii

PAGE 8

APPENDICES APPENDIX A APPENDIX B APPENDIX C APPENDIX D APPENDIX E APPENDIX F APPENDIX G APPENDIX H TABLE OF CONTENTS--Continued EXPORTS OF ORANGE JUICE ................ 172 IMPORTS OF ORANGE JUICE ............... 180 FLORIDA'S ORANGE JUICE .............. 188 REAL EXCHANGE RATE VALUES AND INDICES 192 VARIABLE DEFINITIONS AND DATA SOURCES ..... 203 MEANS, STANDARD DEVIATIONS AND COEFFICIENTS OF VARIATION FOR VARIABLES USED IN THE MODEL. . . . . . 213 STRUCTURAL-FORM ESTIMATES OF ELASTICITIES AND FLEXIBILITIES AT THE VARIABLE ME.ANS. . . . . . . 218 ACTUAL AND REDUCED-FORM ESTIMATED VALUES OF ENDOGENOUS VARIABLES ................. 224 LIST OF REFERENCES.......... .......................... 238 BIOGRAPHICAL SKETCH .................................... 245 viii

PAGE 9

LIST OF TABLES Table 1 Shares of total exports of orange juice for major exporting countries, 1969-1978 ............ 3 2 Orange production in the United States of America, Brazil and other major orange-producing countries, 1972-1979............................. 4 3 Shares of total imports of orange juice for major importing countries, 1969-1978 ............ 4 Market shares for major suppliers of the United States of America imports of orange juice, 7 1969-1979........................................ 18 5 The United States of America exports of FCOJ and of other orange juices, 1972-1979 ................ 19 6 Brazilian FCOJ weighted average export price (FOB Santos) in major markets, 1972-1979 .............. 21 7 Weighted average wholesale and retail prices of FCOJ and of other orange juices in the United States of America, 1972-1979 .................... 22 8 Brazilian and the United States of America FCOJ weighted average export prices (FOB) in three major markets, 1972-1979 ......................... 24 9 Orange juice imports per capita and major FCOJ suppliers market share in EEC7, European nonEEC3 and Canada, 1972-1979 ....................... 25 10 Promotional expenditures of the three-party program in Europe and of advertising programs in the USA and Canada for orange juices from Florida, 1972/73 1978/79 ....................... 32 11 Description of variables in the orange juice market structure in the USA ...................... 49 12 Description of variables in the FCOJ market structure in Canada .............................. 65 ix

PAGE 10

LIST OF TABLES--Continued Table Page 13 Description of variables in the FCOJ market structures in the EEC7 and the European non-EEC3 countries ......................... 71 14 Structural-form estimates of the FCOJ retail demand equation in the USA market (RQFSt, Equation 1) . . . . . . . . . 84 15 Structural-form estimates of the FCOJ retail price equation in the USA market {RPFSt, Equation 2)...................................... 85 16 Structural-form estimates of the FCOJ wholesale demand equation in Florida {WQFFlt, Equation 3) .. 89 17 Structural-form estimates of the FCOJ wholesale price equation in Florida (WPFFlt, Equation 4) . . . . . . . . . . . 9 0 18 Structural-form estimates of the USA FCOJ import demand equation from Brazil {MQFSBt, Equation 7). 94 19 Structural-form estimates of the USA FCOJ import price equation from Brazil (MPFSBt, Equation 9) 95 20 Structural-form estimates of the other orange juices retail demand equation in the USA mark~t (RQOSt, Equation 10)............................. 99 21 Structural-form estimates of the other orange juices retail price equation in the USA market (RPOSt, Equation 11) .......................... 100 22 Structural-form estimates of the other orange juices wholesale demand equation in Florida (WQOFlt, Equation 13) ........................... 101 23 Structural-form estimates of other orange juices wholesale price equation in Florida (WPOFlt, Equation 14) . . . . . . . . . 102 24 Structural-form estimates of Brazil FCOJ export demand equation to Canada (XQFCBt, Equation 17) .. 107 25 Structural-form estimates of Brazil FCOJ export price equation to Canada (XPFCBt, Equation 18) ... 108 X

PAGE 11

LIST OF TABLES--Continued Table 26 Structural-form estimates of the USA FCOJ export demand equation to Canada (XQFCSt, Equation 19) 112 27 Structural-form estimates of the USA FCOJ export price equation to Canada (XPFCSt, Equation 20) ... 113 28 Structural-form estimates of Brazil FCOJ export demand equation to the EEC7 countries (XQFEBt, Equation 21) . . . . . . . . . 119 29 Structural-form estimates of Brazil FCOJ export price equation to the EEC7 countries (XPFEBt' Equation 22) . . . . . . . . . 121 30 Structural-form estimates of the USA FCOJ export demand equation to the EEC7 countries (XQFESt, Equation 23)..................................... 123 31 Structural-form estimates of the USA FCOJ export price equation to the EEC7 countries (XPFESt, Equation 24) . . . . . . . . . 124 32 Structural-form estimates of Brazil FCOJ export demand equation to the European non-EEC3 countries (XQFNBt, Equation 25) .................. 129 33 Structural-form estimates of Brazil FCOJ export price equation to the European non-EEC3 countries (XPFNBt, Equation 26)............................ 130 34 Structural-form estimates of the USA FCOJ export demand equation to the European non-EEC3 countries (XQFNSt' Equation 27) .................. 133 35 Structural-form estimates of the USA FCOJ export price equation to the European non-EEC3 countries (XPFNSt, Equation 28) ................. 134 36 Estimates of the derived reduced-form equations .. 140 37 Derived reduced-form elasticity and flexibility estimates for selected predetermined variables evaluated at the means of the respective variables. . . . . . . . . . 14 3 38 Ratios of the root mean square error to observed mean value for comparing actual and derived reduced-form estimated values of endogenous variables for specific periods within the data range used in the study ......................... 156 xi

PAGE 12

LIST OF FIGURES Figure 1 Orange juice trade flows for major exporting and importing countries ............... 15 2 Effects of currency devaluation by exporting country on equilibrium price and quantity for a given corninodi ty. . . . . . . . 42 3 Orange juice market structure in the USA ........ 46 4 Orange juice market structure in Canada, the EEC7 and the European non-EEC3 countries ........ 62 xii

PAGE 13

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy AN ECONOMETRIC MODEL OF INTERNATIONAL TRADE OF FROZEN CONCENTRATED ORANGE JUICE By LUIZ JOSE MARIA IRIAS August, 1981 Chairman: Dr. Daniel S. Tilley Cochairman: Dr. Jonq-Ying Lee Major Department: Food and Resource Economics A simultaneous equation econometric model of inter national trade of frozen concentrated orange juice was developed for two major producing and exporting countries (the USA and Brazil) and four major importing countries or groups of countries, i.e., Canada, the USA, the EEC7 countries (West Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark and France) and the European non-EEC3 countries (Sweden, Norway and Finland). The model consists of demand and price equations for FCOJ and other orange juices for the USA retail, wholesale (Florida) and import (FCOJ from Brazil) markets, and for Brazil and the USA FCOJ export markets (Canada and Europe). xiii

PAGE 14

The system of structural-form equations was estimated by two-stage least squares. Florida's wholesale prices, orange production forecasts, USA import and export prices, exchange rates and occurrence of freeze in Florida were found to play major explanatory roles in the various price equations. Florida's wholesale FCOJ price, together with the USA import price of FCOJ from Brazil, determines the price spread. This spread is the key variable in the simultaneous determination of imports by the USA, and the USA export price to Canada and Europe. USA export prices are the major variables in the simulatenous determination of Brazilian export prices to Canadian and European markets. USA imports of FCOJ from Brazil were found to have an elastic response to the price spread (Florida's wholesale price minus import price). Brazilian exports to Canada and USA exports to Europe were found to have an elastic response to respective own-prices. Brazilian FCOJ was found to be a substitute for USA FCOJ in the Canadian market, while in the European markets the two products behaved quite independently of each other. xiv

PAGE 15

CHAPTER I PROBLEM SETTING Statement of the Problem Volume of world trade of processed citrus products has increased dramatically within the last fifteen years. Among processed citrus, orange juice is by far the most important traded product as compared with fresh grapefruit and lemon juice, pulp, oil and other by-products. The majority of orange juice trade has been in the form of frozen concentrated juice with some hot pack and single strength juice also traded. From 1969 to 1978 total orange juice exports increased from 197.6 to 646.1 million gallons of single strength juice equivalent, an increase of nearly 230 percent (Table A.1. Appendix A). In the USA and Canada, orange juice consumption has been increasing. Myers (1978b) reports a consumption increase of over 85 percent in Canada from 1968 to 1976. From 1972 to 1979 Florida's domestic movement of frozen concentrated orange juice (FCOJ) increased more than 50 percent. Canned and chilled orange juice domestic movement increased nearly 70 percent during the same period (Florida Canners Association, 1971-1980). Similar trends in con sumption took place in Europe beginning in the early sixties 1

PAGE 16

2 when Brazil offered FCOJ at low prices (Lingens, 1978). European countries increased imports of orange juice by 60 percent from 1969 to 1978 (Table B.l. Appendix B). Brazil's exports to this region, in the same period of time, increased by more than 750 percent (Table A.2. Appendix A). Brazil, the United States of America (USA), Israel, Spain, Italy, Morocco and Mexico are the leading orange juice producing and exporting countries. These countries have accounted for almost 90 percent of total exports (Table 1). This volume increased considerably from 1975 to 1978. Other exporting countries include Greece, South Africa, Belize and Argentina. The USA is the world's largest producer of oranges and orange juice, and the state of Florida is the primary producing region (Table 2). According to the Florida Crop and Livestock Reporting Service (1979), practically 80 percent of the USA orange production is processed and Florida alone accounts for almost 90 percent of the entire portion which is processed. From 1969 to 1978, the USA average share of total orange juice exports was around 15 percent. Brazil's share of the export volume has increased steadily. Brazil is the largest exporter of orange juice (Table 1) and the second largest orange producer (Table 2). The state of Sao Paulo accounts for more than 62 percent of Brazil's commercial orange production and for practically all of the

PAGE 17

Table 1. Shares of total exports of orange juice for major exporting countries, 1969-1978 Year Brazil USA Israel Spain Italy Morocco Mexico Other Countries 1 Total ---------------------------------Percent----------------------------------------1969 1970 1971 1972 14.94 18.38 32.30 32.31 1973 38.04 1974 35.88 1975 52.84 1976 51. 89 1977 53.52 197 65. 61 Average 44.49 14.58 18.28 15.19 14.65 15.47 16.70 16.13 16.63 17.08 8.98 14.99 12.03 10.01 13.96 12.60 10.74 12.78 9.83 8.12 7.55 6.63 9.85 7.20 10.00 7.85 7.86 6.07 6.02 4.79 8.80 4.26 3.79 6.25 15.24 14.11 9.56 8.90 6.16 5.49 3.83 1.16 1.73 1. 83 5.33 11.77 12.58 4.89 5.97 5.66 4.10 2.45 2.62 2.66 2.08 4.47 .19 .44 .75 2.45 3.27 3.64 1.02 2.10 5.99 4.62 2.89 24. 05 16.20 15.50 15.26 14.59 15.39 9.11 8.68 7.21 6.46 11. 73 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 1 Includes Algeria, Argentina, Belize, Cyprus, Ghana, Greece, Jamaica, South Africa and Trinidad-Tobago. Source: Computed from Table A.l. Appendix A. w

PAGE 18

Table 2. Orange production in the United States of America, Brazil and other major orange-producing countries, 1972-1979 Brazil Year State of Other Sao Pauloa states 1 Totalb Florida USAc Other states 2 Total Other_ d 3 countries Total --~------------------------million boxes of 90 pounds---------------------------1972 60.7 20.1 80.8 137.0 1973 62.4 58.3 120.7 169.7 1974 82.0 70.6 152.6 165.8 1975 84.7 70.4 155.1 173.3 1976 99.6 80.8 180.4 181.2 1977 92.0 83.5 175.5 186.8 1978 150.0 42.4 192.4 167.8 1979 151.5 45.3 196.8 164.0 Total 782.9 471.4 1,254.3 1,345.6 4 Percent 16.75 10.08 26.83 28.78 54.4 191.4 206.0 55.0 224.7 209.9 50.4 216.2 219.5 64.5 237.8 216.8 61.6 242.8 205.3 56.1 242.9 208.2 52.3 220.1 187.8 46.5 210.5 180.9 440.8 1,786.4 1,634.4 9.43 38.21 34.96 478.2 555.3 588.3 609.7 628.5 626.6 600.3 588.2 4,675.1 100.00 1 Includes the states of Bahia, Minas Gerais, Parana, Rio Grande do Sul, Rio de Janeiro, Santa Catarina and Sergipe. 2 Includes the states of Arizona, California and Texas. 3Includes Argentina, Belize, Israel, Italy, Jamaica, Mexico, Morocco, Spain and South Africa. 4 Average percentage of total orange production. Sources: augited States Department of Agriculture, Foreign Agricultural Service, 1980a and 1980b. Food and Agriculture Organization of the United Nations, 1973-1979; and United States Department of Agriculture, Foreign A~ricultural Service, 1980c. cFlorida Crop and Livestock Reporting Service, 1976 and 1979. Food and Agriculture Organization of the United Nations, 1973-1979.

PAGE 19

5 d 1 citrus in ustry processing capacity. More than 80 percent of Sao Paulo's orange production is estimated to be processed into FCOJ (United States Department of Agriculture, Foreign Agricultural Service, 1980a), and almost all of the FCOJ is exported. 2 The Brazilian share of total orange juice exports increased from 14.94 percent in 1969 to 65.61 percent in 1978 with an average share for the period of 44.49 percent (Table 1). Israel ranks third in orange juice exports followed by Spain, Italy, Morocco and Mexico (Table 1). Export markets for the fresh fruit are the primary outlets for Isreali citrus fruit, followed by the domestic fresh market. The citrus processing industry is the third most important market alternative for the Israeli product under the Citrus Marketing Board policies (Melamed,1977). Spain, Italy and Greece are orange juice exporting countries that are entirely dependent on the market in the European Economic Community (EEC). Among the remaining countries that export orange juice, Mexico is the most important. From an insignificant position in the late sixties, Mexico moved to 1 Except for small plants located in the northeastern state of Sergipe and in the southern states of Santa Catarina and Rio Grande do Sul, Brazil's citrus processing industry is located entirely in the state of Sao Paulo. 2 Domestic consumption of processed oranges in Brazil is estimated to be 5 to 7 percent (Myers, 1978a; United States Department of Agriculture, Foreign Agricultural Service, 1980a)

PAGE 20

6 be the fourth most important exporting country in 1977 and 1978, its shares of total exports being 5.99 and 4.62 percent, respectively, in those years (Table 1). From 1969 to 1978, Brazilian USA and Israeli orange juice exports accounted for nearly 70 percent of the total trade. From 1975 to 1978, these countries exported almost 80 percent of the total. Brazil and the USA accounted for an average of 60 percent of the total orange juice exports and from 1975 to 1978 these two countries exported around 70 percent of the total (Table 1). They are the major competing suppliers in world orange juice markets. The major orange juice importing countries are located in Europe and North America. From 1969 to 1978, 12 countries in these two regions imported nearly 90 percent of total orange juice traded (Table 3). Most European imports go to EEC member countries and specifically to West Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark and France (the EEC7). These countries imported an average of 63.07 percent of the total world imports. The European non-EEC countries are the next largest orange juice importing region, accounting for 18 percent of the total volume traded. Within this region Sweden, Norway and Finland (the European non-EEC3) import 9 percent of the world total (Table 3) and have the highest per capita orange juice consumption indices in Europe (Lingens, 1978). North America, as represented by Canada and the USA, is ranked as the third most important orange juice importing

PAGE 21

Table 3. Shares of total imports of orange juice for major importing countries, 1969-1978 EEC European non-EEC Year EEC7 1 Others 2 3 4 non-EEC3 Others Canada USA Other 5 countries Total --------------------------------------Percent-----------------------------------1969 67.20 1970 66.70 1971 67.90 1972 66.65 1973 68.20 1974 64.22 1975 62.20 1976 60.79 1977 61.60 1978 51. 98 Average 63.07 .74 l. 91 3.33 2.63 l. 32 1.12 .43 .19 .17 .29 1.11 9.53 11.68 8.91 7.75 8.58 9.24 8.14 9.80 9.11 7.53 8.89 10.84 10.44 8.04 8.05 8.10 10.52 9.23 8.73 8.89 7.64 8.90 8.89 7.15 6.91 6.51 7.72 8.06 9.49 8.95 9.35 11. 37 8.59 l. 99 .65 3.59 7.32 4.52 3.42 4.80 4.80 6.75 18.92 6.31 .81 l. 47 l. 32 1.09 l. 56 3.42 5.71 6.74 4.13 2.27 3.13 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 1 EEC7 as Belgium, non-EEC3 non-EEC3 include defined in this study includes West Germany, The Netherlands, the United Kingdom, Luxembourg, Denmark and France. 2 other EEC includes Italy and !celand. 3 European as defined in this study includes Sweden, Norway and Finland. Other European includes Austria, East Germany, Iceland, Spain and Yugoslavia. 5 other countries Antigua, Australia, Barbados, Jamaica, South Africa and Trinidad-Tobago. Source: Computed from Table B.l. Appendix B.

PAGE 22

8 region accounting for an average of 15 percent of the total. While Canada's share of total imports has been relatively constant at about 8 percent, the USA import share has fluctuated from .65 percent in 1970 to 18.92 percent in 1978. Most previous studies on production, consumption and trade of orange juices with an international perspective concentrate on one sector or on one country or world region. In the USA a number of studies focused on orange juice import and export trade issues with domestic and/or foreign perspectives (Polopolous, 1968; Ward & Niles, 1975; Ward, 1976a; Myers et al., 1979). Most of these studies viewed the import-export issues from the viewpoint of the Florida citrus industry. Ward's (1976a) study was the first major FCOJ trade model addressing import and export issues in a simplified world context. It provided an understanding of the major economic relationships of Florida's imports and exports. The model consists of a system of simultaneous equations explaining imports to Florida, export price from Florida to Europe, European demand for Florida FCOJ and a definitional equation for the price spread. 3 European demand was found to be highly sensitive to Florida concentrate price changes. 3 Price spread was defined as the difference between the quarterly average Florida domestic FOB price and the quarterly average price of imports to Florida from Brazil, excluding tariffs. The import price was adjusted by the cost of converting bulk concentrate to equivalent gallons of retail pack concentrate (Ward, 1976a, pp. 105-106).

PAGE 23

9 The desirability of a two-pricing system between the domestic and European markets was clear. Price spread changes were found to be the major factor influencing the level of Florida imports of FCOJ. Myers et al. gave an overview of the FCOJ import-export policies and programs by evaluating how different governmental policies, industry programs and economic factors affect the volume and directions of product flows throughout the system. An important contribution of this study was to show the interdependency between price spreads, tariff levels, and duty drawback provisions 4 and their effects on the quantity of Florida imports. It was shown that under the current set of import-export policies a trade-off exists between the percent of imports offset with exports and the magnitude of the price spread. Even at a zero price spread it is possible to import and export using the drawback provisions without losing money. The necessary condition is a relatively higher export price as compared with the domestic price. Studies by Prato (1969), Rausser (1971) and Buckley (1977) addressed the USA FCOJ domestic economic issues. They focused on the estimation and analysis of consumer demand structures within the USA domestic markets with little reference to the role of imports and exports. 4 The duty drawback provisions of the u. S. Tariff Act of 1930 permit processors to import foreign FCOJ in bulk form and to recover 99 percent of the duties by exporting like amounts of FCOJ in bulk or consumer pack within a 5-year period from date of importation (Myers et al., 1979).

PAGE 24

10 Brazilian participation in world trade of orange juice has been studied by Morais and Medeiros (1978), and by Moretti (1978). Morais and Medeiros describe the Brazilian citrus industry in a general fashion. Moretti's work uses an analytical framework to describe and to analyze important market factors affecting Brazilian trade of FCOJ. He used a single-equation model to estimate export demand relation ships. Major study results indicate that concentrate exports were highly responsive to changes in prices and exchange rates. European markets are quite important in orange juice consumption. Priscott (1969), Ward (197Gb), and Nguyen (1977) are the major studies which addressed trade, consumption patterns, and structural consumption relations in these markets. These studies focused on Florida's FCOJ exports to Western European countries. Their results showed highly price-elastic markets in most of the countries with changes in exchange rates affecting export demands. Ward's (197Gb) study also addressed the benefits to the Florida citrus industry of discounting its export price to Europe relative to the domestic price by using the relatively cheaper imported product. Chern (1973), and Tilley and Lee (1981) studied consumer demand for orange juice in Canada. Chern used a single-equation model to estimate Canadian demand for FCOJ, canned single strength orange juice (CSSOJ) and chilled orange juice (COJ). An aggregation of the three products was

PAGE 25

11 found to have a unitary price elasticity, an elastic response to income and a positive but a statistically insignificant effect of advertising. Individual orange juice product estimates indicated that FCOJ was price-inelastic and had a strong consumer response to advertising. CSSOJ and COJ demand was found to be price-elastic with no immediate response to advertising. All products were positively related to income changes. Tilley and Lee (1981) used a system of simultaneous equations to analyze Canadian orange juice consumer demand and import demand from Brazil and from the USA. Canada's imports from Brazil were found to be highly price-elastic, while imports from the USA were quite price-inelastic. Cross-price effects on imports indicate that an absolute change in the import price from the USA would have a greater impact on Canadian imports from Brazil than the impact on Canadian imports from the USA, given the same change in the import price from Brazil. Despite significant research efforts a more complete world trade model for orange juice has not been developed. Such a model should explain levels of consumption, trade flows, inventories and prices for major producing and consuming countries. Also, the model should have the capability of evaluating the impact of the introduction of, and changes in, protective trade barriers, such as tariffs and quotas. These are major trade issues affecting economic, social and political decisions in the citrus industry around the world. A model of world orange juice trade would give a

PAGE 26

12 better understanding of the economic relations and inter relations of these factors under domestic and international conditions. Objectives The overall objective of this study is to provide an economic analysis of domestic and foreign trade of orange juice products (frozen concentrate and other forms of orange juice traded) for major producing and consuming countries. This analysis will address the basic economic relations and interrelations in the determination of levels of consumption, trade flows and prices. An attempt will be made to: 1. identify the most important factors affecting trade of orange juice products: 2. evaluate to what extent changes in domestic and foreign demand for the USA orange juice products affect the USA FCOJ imports: 3. evaluate whether foreign and domestic promotion programs for the USA orange juice products provide an "umbrella" for the market expansion for the same products from other producing-exporting countries: 4. determine the impact of pricing strategies and price changes on exports and imports of orange juice products among producing-exporting and importing countries. In order to accomplish these objectives an economic model for international trade in orange juice products will be developed. This model will focus on two basic orange juice products: frozen concentrated and an aggregate of

PAGE 27

13 other product forms of orange juice. Unless otherwise specified, these products are expressed in gallons of single strength juice equivalent. For modeling purposes the world trade of these orange juice products consists of a simplified geographical scenario of two major producing and exporting countries (the United States of America and Brazil) and four major importing countries or groups of countries (Canada, the United States of America, the EEC7 and the European non-EEC3 countries).

PAGE 28

CHAPTER II AN ECONOMIC PROFILE The purpose of this chapter is to describe and discuss the orange juice industry. The first section analyzed basic economic relations concerning production, consumption and trade of orange juices, while the second section discusses major structural and institutional factors affecting the industry. Trade Flows Figure 1 summarizes orange juice flows for major exporting and importing countries. Solid arrows indicate export flows and broken arrows indicate import flows. Numbers shown are estimated average percentages of exports or imports from 1969 to 1979, except in the case of Florida where the percentages are for the 1972 to 1979 period. The USA, as the largest producer and consumer and the second largest exporter of orange juice, is the primary economic indicator of this industry around the world. The state of Florida is by far the largest USA supplier of orange juices, followed by the states of Arizona, California and Texas. While Florida's orange production is practically all utilized for processed products, the production from 14

PAGE 29

s Stocks f mports I 1-~-.J I Domestic 17 ___ _, supply ... 4 FLORIDA U S A. .. Sao Paulo Other Brazil 79 78 Domestic demand Figure 1 Or~nge ju i ce trade flows for major exporting and importing countries. Note : Solid arrows indicate export flow and broken arrows indicate import flows Numbers are approximated average percentage of exports or imports from 1969 to 1979 (Appendices A, B and C)

PAGE 30

16 other states is mainly for fresh products (Myers, 1978b). Detailed data on processed orange products are not available for the states of Arizona, California and Texas. Therefore most of the discussion on the USA processed orange industry will be related to state of Florida data. In the last eight years (1972 to 1979) an average of approximately 79 percent of Florida's total supply of orange juices was produced from the current crop of domestic oranges. The remaining supply came from stocks from previous years (17 percent) and from imports (4 percent). In the past, Florida has been a net exporter of orange juice. Only in the recent two years (1978 and 1979) have Florida's processors imported more than they have exported. This occurred following a freeze in 1977 that substantially reduced domestic production. Florida accounted for more than 70 percent of USA total imports and 80 percent of total USA exports from 1972 to 1979 (Appendix Band Appendix C). Florida's canned and chilled orange juices exports represent about 50 percent of the USA other orange juices (single strength and hot pack) exports (Table A.4. Appendix A and Appendix C). Florida imports and export data by country are not available. Because of Florida's market dominance, United States Department of Commerce data can be used to identify sources and country of destination. Brazil has been the primary source of the USA imports of FCOJ. Brazilian imports averaged around 84 percent of total USA FCOJ imports from

PAGE 31

17 1969 to 1979. Mexico and Belize accounted for 10 and 1 percent respectively (Figure 1, Table 4). USA FCOJ imports varied from 2.6 million gallons in 1970 to a record of 156.4 million single strength equivalent gallons in 1979 (Table B.2. Appendix B). FCOJ is the leading citrus export product in the USA accounting for approximately 77 percent of total exports. Single strength and hot pack accounted for the remaining 23 percent over the eight-year period shown in Table 5. In the six years preceeding the 1977 Florida freeze, the USA FCOJ exports were increasing rapidly. In 1978, exports decreased by almost 40 percent from those of 1977. Practically 48 percent of all the USA orange exports go to Canada (Figure 1 and Table A.3. Appendix A). The EEC countries, led by The Netherlands and West Germany, accounted for an average of more than 25 percent, and the European non-EEC countries accounted for almost 17 percent of USA exports (Table A.4. Appendix A). In 1974 the Brazilian orange industry faced its first crisis when exports declined by more than 10 percent. The economic recession in Europe, high stock levels in most of the importing countries, and quality control problems of the reconstituted Brazilian juice in those countries are factors contributing to the decline. Prices for the Brazilian product (FOB Santos) declined below the 1974 minimum export price of 560 dollars per metric ton established by CACEX 1 (Morais and Medeiros, 1978). 1 cACEX Carteira de Comercio Exterior, Export Trade Office of the Bank of Brazil, Inc., associated with the Ministry of Finance. CACEX's major role is to implement and enforce official government policy regarding exports and imports.

PAGE 32

Table 4. Market shares for major suppliers of the United States of America imports of orange juice, 1969-1979 Year 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 Total 1 Includes Source: Brazil Mexico Belize Other 1 countries Total -------------------------------Percent-----------------------------------53.30 9.77 23.99 12.94 100.00 49.68 21. 91 28.33 .08 100.00 81.37 6.49 7.47 4.67 100.00 77.61 12.87 2.69 6.83 100.00 46.56 20.43 4.71 28.30 100.00 67.20 24.57 1. 73 6.50 100.00 85.38 10.03 00 4.59 100.00 89.14 4.22 00 6.64 100.00 67.89 27.74 .36 4.01 100.00 92.00 6.99 00 1.01 100.00 95.07 4.72 00 .21 100.00 84.41 10.23 1.29 4.07 100.00 Argentina, Israel, Italy, Japan, Spain, South Africa and eleven other countries. Computed from Table B.2. l'}.ppendix B. t-' co

PAGE 33

Table 5. The United States of America exports of FCOJ and of other orange juices, 1 1972-1979 Year FCOJ Total 2 Volume Percent 0 h 1 t er orange Juices ') Volume~ Percent Volume Percent 1972 34,585 68.89 15,620 31.11 50,205 100.00 1973 47,155 75.42 15,367 24.58 62,522 100.00 1974 45,155 76.68 14,951 23.32 64,106 100.00 1975 56,150 80.05 13,990 19.95 70,140 100.00 1976 69,096 80.89 16,322 19.11 85,418 100.00 1977 69,486 80.29 17,053 19.71 86,539 100.00 1978 42,428 73.11 15,609 26.89 58,037 100.00 1979 48,539 73.57 17,439 26.43 65,978 100.00 Total 416,594 76.73 126,351 23.27 542,945 100.00 1 1ncludes orange juices exported as s j,ngle strength an
PAGE 34

20 Nominal price changes for Brazilian FCOJ exports to major markets (Table 6). Average prices in 1975 were the lowest prices shown in all markets. Price levels recovered following the 1977 Florida freeze. Western Europe countries, and especially the EEC countries and the European non-EEC countries, have been the primary customers for Brazilian FCOJ exports accounting for almost 61 percent of Brazil's total concentrate exports from 1969 to 1979 (Figure 1 and Table A.2. Appendix A). West Germany, The Netherlands and Sweden accounted for more than 40 percent of Brazilian FCOJ exports. The USA bought an average of 23 percent of the Brazilian product. The USA recorded 187.4 million gallons of FCOJ in single strength juice equivalent imports from Brazil in 1978 when it emerged as Brazil's most important buyer. As a result of the 1977 Florida freeze and the subsequent reduced orange production, domestic prices of orange juices increased in the USA. Both wholesale and retail prices increased substantially as compared with 1976 levels (Table 7). In 1978 FCOJ prices increased as much as 71 percent at wholesale and 59 percent at retail from respective price levels of the year before the freeze. Non-frozen orange juice prices increased approximately 30 percent in both markets. The increased price differential between wholesale domestic and imported juice made it profitable for the USA importers and processors to buy foreign orange juice

PAGE 35

Table 6. Brazilian FCOJ weighted average export price (FOB Santos) in major markets, 1972-1979 Year EEC 1 European 2 non-EEC Canada USA Other 3 countries Total --------current cents per gallon of single strength juice equivalent--------1972 36.93 38.57 37.96 38.34 36.56 37.48 1973 41.51 40.54 38.43 44.47 39.41 41.39 1974 43.04 44.61 43.64 40.93 45.41 42.94 1975 35.38 36.85 35.35 35.57 37.00 35.77 1976 38.05 37.66 36.90 37.10 38.25 37.84 1977 62.91 59.82 59.02 75.04 73.54 65.26 1978 77.34 74.20 81.44 77.62 82.95 78.01 1979 75.61 73.90 76.50 75.29 79.35 75.78 Belgium, Luxembourg, 1 Includes West German~, the United Kingdom, The Netherlands, Denmark and France. Includes Sweden, Norway, Finland and 7 other European countries. 3 Includes Israel, Japan, Venezuela and eleven other countries. Source: Computed from Brazil, Servi90 de Estatistica Econ6mica e Financeira, Ministerio da Fazenda, 1970-1979. N I-'

PAGE 36

Table 7. Weighted average wholesale and retail prices of FCOJ and of other orange juices 1 in the United States of America, 1972-1979 Year a Wholesale FCOJ b Retai 0 h a t er orange Juices Wholesalea Retailb -----current dollar per gallon of single strength juice equivalent------1972 .81139 1. 25078 1.09515 1.77025 1973 .75906 1.23319 .99571 1. 71801 1974 .79321 1.27437 1.11686 1.80698 1975 .86490 1. 37840 1. 20589 1.88345 1976 .83684 1. 41075 1.33023 1.94496 1977 1.1494 8 1. 70821 1.54129 2.19211 1978 1.43292 2.23701 1. 84 367 2.69041 1979 1.43489 2.40002 1.97372 2 96256 1 rncludes canned single strength and chilled orange Juices at Florida's wholesale level, and single strength orange juices in cans, glasses, and other containers at the USA retail level. Sources: aFlorida Department of Citrus, Economic Research Department, 197 2. 1980; and bFlorida Department of Citrus and A. c. Nielsen Company, 1972-1980. N N

PAGE 37

23 concentrate, pay the USA import duty 2 and sell the imported product in the domestic market. FCOJ export prices from the USA have been consistently quoted higher than prices for the Brazilian product (Table 8). Brazilian FCOJ exports are in 65 brix concentrate in bulk drums, while most of the USA exports are in consumer retail pack at 42 brix to Europe and Canada. Brazilian FCOJ exports to Canada have expanded substantially during the seventies (Table A.2. Appendix A). From under 6 million single strength equivalent gallons in 1969 and 1970, exports of the Brazilian orange juice to Canada increased by more than 600 percent, reaching a record of 43.0 million gallons in 1978. Total Brazilian FCOJ exports in 1979 declined to 371.4 million single strength juice equivalent gallons or approximately 12 percent less than the record level of 1978. As previously stated, Europe and Canada are the most important importers of orange juice. Per capita import indices and FCOJ market shares to major suppliers were constructed for these regions for the last eight years (Table 9). Scandinavian countries (the European non-EEC3) have the highest per capita imports in Europe. According to Lingens (1978), per capita consumption of orange juice in Sweden is more than five gallons of single strength 2 rmport duty on FCOJ imports has been 34 cents per pound of solids or 34.986 cents per gallon of single strength juice equivalent or 487.22 USA dollars per metric ton of 65 brix concentrate.

PAGE 38

Table 8. Brazilian and the United States of America FCOJ weighted average export prices (FOB) in three major markets, 1972-1979 Year EEC7 1 European non-EEC3 2 Canada Brazil USA Brazil USA Brazil USA ------current USA dollars per gallon of single strength juice equivalent----1972 .36908 .62114 .38503 .82022 .37957 .90133 1973 .41549 .51342 .40967 .74340 .38433 .90879 1974 .43042 .59091 .44630 .81385 .43641 .92286 1975 .35365 .63207 .37995 .86196 .35354 1. 00308 1976 .37413 .56843 .37662 .83142 .36899 1. 00201 1977 .62909 .70637 .61148 .95143 .59024 1.22939 1978 .77343 1.49492 .73833 1. 29239 .81436 1. 71403 1979 .75615 1.47822 .73134 1. 26381 .76502 1.75170 1 West Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark Includes and France. 2 rncludes Sweden, Norway and Finland. Sources: Computed from Brazil, Serviyo de Estatistica Economica e Financeira, Ministerio da Fazenda, 1970-1979; and from United States Department of Commerce, Bureau of the Census, 1969-1979a. N ""'

PAGE 39

Table 9. Orange juice imports per capita and major FCOJ suppliers market share in EEC7 1 European non-EEC32 and Canada, 1972-1979 Im2orts per capita 3 FCOJ market shares 4 Year EEC7 1 European 2 Canada EEC7 1 2 Canada EuroEean non-EEC3 non-EEC3 Brazil USA Brazil USA Brazil USA Others 5 1972 .38 1.00 1. 81 88.71 11. 29 57.23 42.77 36.63 51.05 12.32 1973 .63 1.53 2.22 89.29 10.71 59.56 40.44 39.12 47.00 13.88 1974 .43 1. 89 2.21 90.36 9.64 68.48 31. 52 19.18 60.36 20.46 1975 .68 2.13 2.87 90.98 9.02 72.22 27.78 47.94 46.82 5.24 1976 .93 2.50 2.83 87.99 12.01 77.40 22.60 39.21 54.86 5.93 1977 .75 2.54 2.96 88.07 11. 93 76.64 23.36 39.28 49.31 11. 41 1978 .66 2.42 3.61 94.38 5.62 89.92 10.08 61.07 34.13 4.80 1979 .85 2.54 3.71 94.56 5.44 91. 06 8.94 56.37 40.10 3.53 1 Includes West Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark and France. 2 Includes Sweden, Norway and Finland. 3 Gallons of single strength juice equivalent per capita. In the EEC7 and the E~ropean non-EEC3 countries includes only imports from Brazil and the USA. 4 Percent. Includes Argentina, Belize, Israel, Jamaica, Mexico and nineteen other countries. Sources: Computed from Tables B.3. to B.5., Appendix B; Table A.4., Appendix A and International Monetary Fund (populatio~, 1970-1980. l\J u,

PAGE 40

26 juice equivalent per year. Considerable attention is given to juice quality and packing procedures in Sweden. In EEC7 total orange juice imports in 1979 from Brazil and the USA, were around 169.2 million single strength juice equivalent gallons, an increase of more than 590 percent from the 1969 import level (Table B.4. Appendix B). Orange juice imports per capita more than doubled in that region in the last eight years (Table 9). In Sweden, Norway and Finland, Brazil substantially increased its FCOJ market share from 57.23 percent in 1972 to 91.06 percent in 1979. Even though the USA's participation in this market increased in absolute terms, the USA share of total FCOJ exports declined from 42.77 to 8.94 percent from 1972 to 1979 (Table 9). All orange juice used in Canada is imported. Brazil and the USA are again the major suppliers with a large group of other countries far behind. In 1978 and 1979, Brazil's market share exceeded 50 percent. The USA share dropped below 40 percent in 1978 and was just over 40 percent in 1979. Structural and Institutional Factors The purpose of this section is to describe and discuss major structural and institutional factors affecting production, consumption and trade in the orange juice industry. In Brazil the major government impact on trade of FCOJ is exercised through CACEX and the Citrus Juice Export Committee integrated by government officials (CACEX,

PAGE 41

27 Ministry of Agriculture, and state of Sao Paulo) and by citrus grower and processor representatives. The most important activities of CACEX are to license FCOJ exports, to establish and enforce reference prices for oranges to be processed, to set minimum FCOJ export prices, and to control stocks. The remainder of this section is organized to cover import and export incentives, market promotion programs, exchange rate policies, protective restrictions, and preferential concessions. Import and Export Incentives In the USA the duty drawback provisions of the Tariff Act of 1930 are a very important mechanism influencing FCOJ trade. Under such provisions processors can import foreign FCOJ in bulk form and recover 99 percent of the import tariff (34 cents per pound of solids) by exporting a like amount of product in bulk or consumer pack within a five year period from the date of importation (Myers et al., 1979). The rationale for such trade activity was discussed by Ward (1976a) and expanded by Myers et al. (1979). Import tariff, the price spread (the differential between domestic and import FCOJ prices), duty drawback provisions and export prices are key determinants of USA FCOJ imports and exports. If the price spread is higher than the tariff, then it is profitable to sell imported FCOJ in the domestic markets. However, if the price spread is equal to or less than the tariff on imports, then processors will not be able to

PAGE 42

28 realize profits on the imported FCOJ unless they have an option to export part of the reprocessed product and then recover most of the tariff paid by taking advantage of duty drawback. As long as the price spread is less than the tariff level, more juice needs to be exported under duty drawback for a given export price. As the spread approaches the tariff level, fewer exports are needed to compensate for imports. Under the present set of USA FCOJ import-export policies, a trade-off exists between the percent of imports offset with exports and the magnitude of price spread. Myers et al. (1979) showed that, even at a zero price spread, it is possible to import and export FCOJ and use the duty drawback provisions without losing money. The necessary condition is a relatively higher export price as compared with the USA domestic price. Brazil 3 does not have specific incentives for citrus growers, but does have federal and state support programs for the agricultural sector. Among others, citrus producers are entitled to subsidized credit, research and extension programs. The government also gives some assistance in establishing grower prices for oranges to be processed through the Citrus Juice Export Committee. However, oranges are not included in the minimum price program carried out by the federal government for major agricultural products. The subsidized credit program of zero interest rates on loans 3 Most of the discussion of export incentives in Brazil was organized from United States Department of Agriculture, Foreign Agricultural Service, 1980a and 1980b.

PAGE 43

29 for fertilizer and of 15 to 18 percent a year for other production activities, can be a great incentive in view of Brazil's high rate of inflation in the past years (more than 40 percent). Processors and exporters have various incentives in Brazil. Under a national policy, given the non-existence of similar domestic products, processors are allowed to import plant parts and equipment duty-free. Also available are subsidized credit for investments in new factory equipment (in 1979 the interest rate was 22 percent a year for fiveor six-year loans), and for financing production for exports (in 1979 the interest rate was 8 percent a year discounted at the time of the loan). Under the latter program, an exporter is allowed to borrow up to 30 percent of the value of his exports in the previous year and repay it in no more than a year. Oranges processed for export are not subject to state and federal value-added taxes 4 (ICM and IPI) and profits from FCOJ export sales are exempt from corporate income taxes. A tax credit of 20 percent on value of exports, which could be used to offset taxes on domestic sales, has been reduced since 1979 by one percent per quarter so that it will be completely eliminated by the end of 1983. Initiated late 4 ICM (Impasto de Circula9ao de Mercadorias) is a state levied value-added tax, and IPI (Impasto de Produtos Industrializados) is a federal-levied value-added tax on industrial products.

PAGE 44

30 in 1979, when the Brazilian cruzeiro 5 was substantially devaluated against the USA dollar, an export tax on FCOJ was introduced as a measure intended to support export prices and to guarantee an adequate domestic supply of oranges and processed products. Market Promotion Programs Brazilian citrus industry market promotion activities have been minimal. Domestically, the only information available (Myers, 1978a) was related to industry participation in expanding internal use through the development of a school lunch market for FCOJ in Sao Paulo and in other central and southern Brazil states. Growers, in general, are not directly involved in any demand-creation efforts. In external markets the situation is similar. Only one Brazilian processor (Cutrale), jointly with Coca Cola, was participating in an effort to package and market concentrate in a form which requires no refrigeration. A successful market test was completed in Sweden, and production and distribution began in 1979 (Myers, 1978c). Processors appear to be quite unenthusiastic about market promotion programs, given the unusually strong current demand for their product. Practically all processed products are exported. 5 cruzeiro is the Brazilian m~netary unit. rate policy is discussed with more detail on that topic. Brazil's exchange in the subsection

PAGE 45

31 In the USA, and in particular the Florida citrus industry, a great effort has been under way for several years to promote citrus products from Florida in domestic and in foreign markets. Domestically, Florida Department of Citrus expenditures for promotions on orange juice products are reported to be more than 12.6 million dollars as an average for the last seven fiscal years (Table 10). To this effort one should add specific brand and distributor's advertising programs around the country. Florida's shipments of orange juice to domestic markets has increased by more than 50 percent in the last eight years (Table C.3. Appendix C). In foreign markets, the Florida citrus industry has been promoting orange juices in Western Europe and in Canada. Since 1966-67 their major promotional effort in Europe has been through the Three-Party Market Development Program. 6 The main target of the program, as measured by distribution of total expenditure, appears to be the Scandinavian countries (European non-EEC3). About half of the promotional expenditure has been used in Sweden, Norway and Finland (Table 10). The other half of the promotional expenditures has been used in the seven EEC countries. Effectiveness of the three-party program has been reported by the studies of Lee (1977, 1978) and 6 This program has been supported by Florida Department of Citrus, United States Department of Agriculture, Foreign Agricultural Service, and European distributors. See Lee et al. (1978) for a more complete description.

PAGE 46

Table 10. Promotional expenditures of the three-party program in Europe and of advertising programs in the USA and Canada for orange juices from Florida, 1972/73 1978/79 year 1 Three party program in Europe Fiscal EEC7 2 European 3 Others 4 Total USA Canada non-EEC3 --------------------------------1,000 dollars---------------------------1972-73 636.4 788.7 190.4 1,615.5 9,790.0 311. 4 1973-74 723.3 756.7 137.0 1,617.0 6,743.4 434.1 1974-75 397.2 650.9 121. 2 1,169.3 12,131.4 483.6 1975-76 518.0 733.8 105.9 1,357.7 12,938.5 557.2 1976-77 507.6 680.5 157.1 1.345.2 16,360.4 462.6 1977-78 507.3 444.0 113.9 1,065.2 12,792.9 659.4 1978-79 470.8 404.9 157.5 1,033.2 17,791.9 738.8 Total 3,760.6 4,459.5 983.0 9,203.1 88,548.5 3,647.1 1 Includes expenditure from July to June {twelve months), except for the three-party program in 1975-76 {July 1975 to September 1976 or fifteen months) and from 1976-77 to 1978-79 {October to September or twelve months). 2 Includes Wes~ Germ~ny, The Netherlands, the United and France. Includes Sweden, Norway and Finland. Switzerland. Source: Florida Department of Citrus, 1980. Kingdom, Belgium, Luxembourg, Denmark 4 Includes Austria, Iceland, Italy and w N

PAGE 47

33 Lee et al. (1979). Overall conclusions from those studies are favorable to the program. Additional dollar sales generated by the program are far in excess of its costs. These studies did not identify any direct link of the three-party program expenditures to Florida's imports of FCOJ. In Canada, the Florida Department of Citrus advertising program for orange juices has also increased substantially. From 1972/73 to 1978/79 promotional expenditures in Canada increased by practically 140 percent, achieving in that period an average fiscal year expenditure of more than one-half million dollars (Table 10). A multi-level study by Tilley and Lee (1981) on Canadian demand for orange juice found a weak positive sales response as advertising expenditure increased. Exchange Rate Policies Exchange rates express the monetary value of a national currency per unit of another country's currency. Changes in this relation reflect changes in the domestic currency's purchasing power as compared with the foreign currency. In terms of commodity trade, an exporting country's devaluation is equivalent to a commodity price cut viewed by the foreign importer. A devaluation of the Brazilian cruzeiro relative to the USA dollar means that fewer dollars are needed to buy a unit of a given commodity from Brazil for any given price expressed in cruzeiros. The final effect of exchange rate changes on the amount of trade

PAGE 48

34 depends on other factors such as commodity elasticities, government import-export policy and adjustments of the international monetary system. These factors will be incorporated in the theoretical discussion of exchange rates in the next chapter. The remaining part of this subsection will discuss exchange rate changes and exchange rate policies in Brazil and in the USA. In Brazil the official exchange rate has been historically controlled by the Government to a greater or lesser degree, depending on the time period analyzed (Veiga, 1974). Since 1968, Brazil has been using a system of mini adjustments of the cruzeiro against the USA dollar. These adjustments, most of them devaluations, have been made at short and random intervals (one or two months) to minimize the speculative demand for dollars. The cruzeiro is devalued at a rate approximately equal to the difference between Brazil's internal inflation rate and the average inflation rate of Brazil's major trading partners, basically the USA, Western Europe and Japan (Alves & Pastore, 1978, p. 866). A departure from the mini-devaluation policy by the Brazilian Government took place in December 1979 when the cruzeiro was devalued by 30 percent (Funda9ao Getulio Vargas, 1972-1980). This was part of an overall change of the country's economic programs in an expectation of better economic results that would help it pay its increasing international debt (especially oil bills and loans), as well as to fight the high and increasing levels of internal

PAGE 49

35 inflation. The annual rate of inflation in 1979, estimated by Funda9ao Getulio Vargas (Conjuntura Economica 34, February,1980 p. 14), was 77.2 percent or almost twice as much the 40.8 percent level in 1978. Foreign exchange market forces are the basic determinants of exchange rate values for the dollar in the USA. The change from a fixed to a floating system of exchange rates by the international monetary community in early 1973 (Abrams, 1980) was another factor that can be used to explain increasing variability of the USA dollar against major foreign currencies. Schuh (1974) argued that the exchange rate has had an important role in explaining depressed prices of agricultural products associated with an overvaluation of the dollar prior to 1971, as well as the increases in the same prices in 1973-74 associated with a devaluation of the dollar. Studies by Vellianitis Fidas 1976, and by Johnson et al., 1977 that attempt to empirically evaluate the effects of exchange rate variability on the USA trade in agricultural commodities fail to support the hypothesis that devaluation had a substantial effect on prices. Both studies found that the exchange rate was found to be less important than other factors in explaining high USA agricultural products prices. Trade Restrictions and Concessions Trade restrictions include a variety of protective import tariffs, quotas and non-tariff barriers designed to

PAGE 50

36 shield domestic industry and/or to enforce quality control measures as a way to protect consumers. Moretti (1978) gave a complete description of duties on citrus and citrus products entering the major importing countries. Great variation in those duties is apparent. For example, FCOJ is free of tariffs in Norway, however it is subject to 200 percent ad valorem duty in Venezuela. Orange juice entering the EEC is subject to a common external ad volorem tariff of 12 to 19 percent according to the specific gravity of the juice. However, the EEC allows for a partial or total suspension of imports and the establishment of a minimum import price, if the market for certain processed fruits, including orange juices, experiences or is threatened with serious distrubances (Taylor, 1978). Canada has an overall five percent ad valorem duty on orange juice. However, the product is free of duty to commonwealth suppliers and to all other supply sources of unsweetened orange concentrate if not less than 58 brix. Finland has a 30 percent duty on orange juices that can qualify for a substantial reduction if it is imported for industrial manufacture. In Sweden there is a duty of 5.00 to 7.50 krona per 100 kilograms for orange juice without added sugar, according to the container size, and 30.00 krona per 100 kilograms for orange juices with added sugar. In the USA FCOJ imports are subject to a tariff of 34 cents per pound of solids and, as previously discussed in this chapter, FCOJ importers can use the duty drawback provision to recover 99 percent of the duty paid

PAGE 51

37 if a similar quantity is exported. Japan uses quotas as a way to protect its domestic industry. Preferential concessions are forms of favorable treatment (tariff reductions or other forms) given to one country or to a group of countries. The EEC allows a 70 percent preferential duty reduction on orange juice imports from Algeria, Israel, Morocco and Tunisia, while other suppliers must pay the full amount. Italy and Greece, as producers of citrus and citrus products within the community, have benefited by the condition of free trade between members. This treatment will soon be extended to Spain if it becomes a member of the EEC in the near future. Other regional organizations and trading blocs 7 with a large variety of agreements and commitments may also affect trade of orange juices. Cuba, for example, has been reported as a potential producer of orange juice in the future (Wolff, 1978) and, as a member of the COMECON will have preferential treatment in that market. 7 European Free Trade Association (EFTA) includes Austria, Norway, Portugal, Sweden, Finland, Iceland and Switzerland. Council for Mutual Economic Assistance (COMECON) includes Eastern European nations, the Soviet Union, Mongolia and Cuba. Yugoslavia does not participate fully and Finland, Iraq and Mexico are "cooperant" members of COMECON. Other groups are the Latin America Free Trade Area (LAFTA), the Caribbean Free Trade Area (CARIFTA) and the Central American Common Market (CACM; Ryan & Tontz, 1978).

PAGE 52

CHAPTER III THE ECONOMIC MODEL In this chapter efforts are devoted to setting up the conceptual development of the economic model of the orange juice industry. The goal is to develop the reasoning for a set of equations that summarize the most important market segments of the industry. The first section gives an overview of the underlying economic theory. A brief discussion of theory of consumer behavior is followed by a review of recent theoretical developments of exchange rate effects on commodity prices and trade. In the second section the economic model is formally specified with flow charts and sets of equations. This specification results from putting together important economic considerations from Chapter II, and theoretical guidelines, logical reasoning, experience and basic knowledge of the orange juice industry. Finally, the third section summarizes the estimation of the empirical model. Theoretical Background The analytical framework of this study can be described as a system of demand and price equations for orange juice. The set of demand equations (retail, wholesale, import or export) are derived, in the sense that they reflect consumer 38

PAGE 53

39 1 preferences. In some markets the derived concept will also include a chain effect, for example, retail variables may be in a wholesale equation, or retail and wholesale variables may be in an import or export equation. In such cases, the derived demand relations are also partial reduced-form equations derived from a more complete structure. The price equations stem from the simultaneous nature of the system. This formulation attempts to capture the price transmission effects within different levels of a given market and across markets. Bredahl et al. (1979) emphasize the role of elasticity of price transmission in estimating the elasticity of export demand for a given commodity. Their major concern relates to the implications of government policies that insulate domestic consumers and producers from external price fluctuations. The basic economic foundations for the price equations f d h h h 1 d h' 2 or pro uct sources ot er tan t e USA is price ea ers ip. The USA sets the price and prices from other sources are derived from it. The USA price equations are basically derived from Florida's wholesale prices which have as major determinant measures of orange juice stocks and crop sizes. Exchange rates are included as price equation shifters in 1 The underlying theories of consumer behavior (utility function and revealed preference) will not be discussed in detail in this study. Most Microeconomics textbooks discuss these theories. For example, see Henderson and Quandt, 1980; Russell and Wilkinson, 1979; Phlips, 1974; and Silberberg, 1978. For modern treatments of revealed preference theory see Uzawa, 1960; and Richter, 1966. 2 For more detail on price leadership, see for example Russell and Wilkinson, 1979; Gould and Ferguson, 1980. For application of price leadership in the citrus industry see Buckley, 1977.

PAGE 54

40 most of the structure. The role of exchange rate should be reviewed, not only for its importance as discussed in Chapter II of this study, but also because of the recent attempts to develop a theoretical framework explaining its effects on commodity prices and trade. Most of the work on exchange rates, as economic policy variables, has been in the context of their impact on the balance of payments of a country. In recent years attempts have been made to formulate a theoretical framework that incorporates the role of exchange rate policy on commodity analysis in a less aggregate basis (Schuh, 1974; Kost, 1976). Both studies use the same economic framework (supply and demand analysis in an industry with export potential). ( \ Schuh concentrated relatively more on the role :fi.._<_? the exchange rate within the USA economy with particular interest in the distribution of benefits of economic progress and adoption of new technologies. Kost, on the other hand, was more general in the sense that his discussion attempted to cover the effects of changes in exchange rates on commodity production, consumption, trade levels, and prices for any 3 two trading partners. Only the effects of devaluation by exporting countries will be discussed here under Kost's theoretical framework. 4 3 Empirical works have, in general, provided support for ex change rate effects on commodity analysis with their major differences on its magnitude. See for example Greenshields, 1974; Vellianitis-Fidas, 1976; Johnson et al., 1977; and Collins et al., 1980. 4 For more detailed analysis and effects of other sources of exchange rate changes see Kost's paper.

PAGE 55

41 Figure 2 summarizes these major effects under a simplified set of assumptions 5 for a given commodity. figure has three basic segments describing the economic conditions for this special commodity in the exporting This country (exporter), in the importing country (importer), and at the international level (trade sector). Its horizontal axes measure quantity in any unit and its vertical axes measure prices. Solid lines in Figure 2 show the classical equilibrium in these two economies after trade, under the assumption that currency ratios are one to one. PE 0 and QT 0 are equilibrium price and traded quantity, respectively. Broken lines show a situation of equilibrium after trade associated with a devaluation of the exporting country currency (say the Brazilian cruzeiro) relative to the importing country currency (say the USA dollar). These exchange rate changes (from a ratio of one to one to a ratio of two cruzeiros to one dol l ar) are shown in the importer segment on Figure 2 Its vertical axis was stretched out with the devaluation of the exporter's currency. Under a new price scale the structure of demand 5 These assumptions are: (1) The model consists of a two country world or one country-rest of the world (potential exporter and importer), (2) competitive economic systems exist in both countries, (3) a single homogenous coI!lITlodity is traded, (4) there are no transport costs and no trade barriers, and (5) the market for the single homogenous commodity can be specified by a single downward-sloping demand function and a single upward-sloping supply function for each country (Kost, 1976).

PAGE 56

Cr$ 8 7 6 5 3 PEo 2 8 7 6 5 PEo 2 Cr$ \ \ \ \ \ \ \ \ \ \ \ SE 8 7 6 5 4 3 Pio 2 o..,_ ___________ _,.o..,.. ______________ o 'O 0 Exporter 0 m, Trade Sector 0 0 $ I -14.0 I I J 3.5 I I I l 3.0 I I -I 2 5 I / I J 2 0 / I \ I ,, \ I \ I \ I \ I \ I \ I \ L __ _J __ 1 1 I / ---\ I 1.5 / I I /,._ '11.0 / -, I / I sd / I l 0.5 I I Importer \ I"' \ \ Del 0 Figure 2 Effects of currency devaluation by exporting country on equilibr i um price and quantity for a given commodity.

PAGE 57

43 and supply functions for the importing country will pivot up from their respective intersection with the horizontal axis such that internal equilibrium price (dollars) and quantity will remain as before. Demand for imports (DI) function in the trade sector also moves in the same fashion, given that it is derived from importer demand and supply structure. Since this structure did not change in the exporting country, then the derived supply for exports (SE) in the trade sector segment does not change either. The new equilibrium implies a relatively large traded quantity of of the commodity (QT 1 ), a higher export price in the exporting country currency (PE 1 ) and a lower import price expressed in the importing country currency (PI 1 ). The impact of exchange rate effects on prices and quantities is determined in large degree by the elasticity of supply for exports and by the elasticity of demand for imports 6 which in turn depend on the respective domestic demand and supply elasticities in the exporting and importing countries. In this particular case of devaluation by the exporting country, the role of the export supply elasticity is quite important. One can see that the effects of currency devaluation by an exporter will be relatively more on prices (or traded quantity) if the supply export elasticity is relatively more inelastic (or elastic). Since agricultural products are 6 Another important determinant is the magnitude of the exchange rate changes.

PAGE 58

44 traditionally quite inelastic, then it is very likely that exchange rate effects will be larger on prices than on quantity. In this study it is assumed that total orange juice supply is relatively fixed for both sources (Brazil and the USA). Given previous discussion and the fact that domestic demand for orange juice in Brazil is very small, as compared with large demand in the USA, it is reasonable to expect that a devaluation of the Brazilian cruzeiro will have a relatively larger effect on export price than on export quantity. The same change in the USA dollar is expected to result in a relatively larger effect on export quantity than on export price if the orange juice export supply elasticities are relatively more inelastic in Brazil and relatively less inelastic in the USA. The Economic Model The economic model is essentially a model that explains the trade flows and prices. Supply of fruit is assumed to be exogenous. 7 Variables expected to influence the demand structure, but not incorporated in previous discussions, will also be included in the proper segment of the economic model. Two flow charts, together with four sets of 7 This assumption stems from the nature of orange production. New orange groves may require four to five years before reaching commercial production capacity. This is certainly a major constraint on the USA and Brazil supply responses from changes in orange juice market conditions in a given year. Thus, a simplified assumption of fixed or practically fixed supply is quite reasonable, especially in this study that uses bimonthly data.

PAGE 59

45 equations, describe the economic model for orange juices in the USA, Canada, the EEC7 (seven countries of the European Economic Community), and the European non-EEC3 8 (three European countries non-EEC members) markets. FCOJ and Other Orange Juices Market Structure in the USA Relations and interrelations of the major variables of the model describing the orange juice market structure in the USA are shown in Figure 3 and Equations 1 through 16. 9 The set of equations formally specifies the derivation of 16 endogenous variables 10 in behavioral relations (Equations 1 through 4, Equation 7, Equations 9 through 11, and Equations 13 and 14) or in identity relations 8 The Brazilian domestic market is not included in the model because of its small size (less than 5 percent of the orange juice production) and lack of systematic data. 9 A variable superscript sign plus (+) or minus (-) indicates the expected directional relationship between variables of the equation. Variables without a sign do not have a basis for hypothesizing the relationship. The subscript t stands for time period (December-January 1971/72 to February-March of 1979). Unless otherwise specified, most of the non monetary variables are expressed in gallons of single strength or reconstituted juice equivalent per thousand persons. Monetary variables are expressed in USA dollars in terms of 1975 prices. For more details see Appendix E. 10 Endogenous, or current endogenous or jointly dependent variables,are those variables whose values are determined within a set of equations. Predetermined variables, which can be subdivided into exogenous, lagged exogenous, or lagged endogenous, are those variables used as explanatory variables in the set, but which are completely determined outside the system (exogenous) or are past values of current endogenous variables

PAGE 60

LEGEND Endogenous variables 0 Predetermined variables D Endogenous variables which ,, Solid arrows Indicate derivation Is not shown \_J functional relations ---+ Figure 3. Orange juice market structure in the USA Broken arrows indicate Identity relations _. I

PAGE 61

47 (Equations 5, 6, 8, 12, 15 and 16). (01) RPFSt (02) + + + WQFFlt = 3 (WPFFlt' WPOFlt, RQFSt' RQOSt' (03) WPFFlt = 4 (WSkFS~, WSkFS~-l' PdS~, PdS~-l' (04) (05) WQkFSt = [(WQkFSt-l + WQsFFlt + MQFSBt + MQFSOct (06) + + + MQFSBt = 5 (SpBt' WQOFlt' MPFSOct, WSkFSt' WSkFS~-l' WSkOS~, WSkOS~-l' Zlt' Z2t' ( 07) (08)

PAGE 62

48 + + + + MPFSBt = f6 (WPFFlt' WPFFlt-l' WPOFlt' WPOFlt-l' PdSp~_ 6 RxBS~) ( 0 9) f7 (RPOS~, + + + + + RQOSt = RPShSt' RPFSt' !St' ASt' ASt-1 Zlt' Z2t' Z3t' Z4t' Z5t) ( 10) fa + + RpOSt) RPOSt = (WPOFlt' WPOFlt-l' (11) RpOSt = RQObSt/RQOSt (12) WQOFlt = f 9 (WPOFl~, + WPFFlt, + RQOSt' Zlt' Z2t' Z3t' Z4t, ZSt) (13) WPOFlt = flO (WSkOS~, WSkOS~-l' PdS~, PdS~-l' + Dt' T~) (14) WSkOSt = WQkOS/WQOFlt (15) WQkOSt = [(WQkOSt-l + WQsOFlt WQOFlt) PlSt (16) A description of all variables in the specification of the USA market relationships is presented in Table 11. The demand structure for orange juices in this market is formulated separately for FCOJ (Equations l through 9) and for other orange juices (Equations 10 through 16). The underlying reasons stem from the importance of these two forms of orange juice marketed in the USA, as discussed

PAGE 63

49 Table 11. Description of variables in the orange juice market structure in the USA Variables 1 Description Endogenous variables RQFSt RPFSt WQFFlt WPFFlt WSkFSt WQkFSt MQFSBt SpBt Retail quantity of FCOJ in the USA Retail price of FCOJ in the USA Wholesale quantity of FCOJ in Florida Wholesale price of FCOJ in Florida Wholesale stock of FCOJ in the USA, expressed in bimonths of Florida supplies Wholesale quantity in stock of FCOJ in the USA Import quantity of FCOJ by the USA from Brazil Price spread between Florida's FCOJ wholesale price (WPFFlt) and the USA FCOJ import price from Brazil tMPFSBt) Import price of FCOJ by the USA from Brazil Retail quantity of other orange juices in the USA Retail price of other orange juices in the USA Retail proportion of other orange juices marketed in cardboard containers in the USA Wholesale quantity of other orange juices in Florida Wholesale price of other orange juices in Florida Wholesale stock of other orange juices in the USA, expressed in bimonths of Florida supplies Wholesale quantity in stock of other orange juices in the USA (continued)

PAGE 64

50 Table 11 (continued) Variables 1 Description XQFCSt Export quantity of FCOJ to Canada from the USA XQFESt Export quantity of FCOJ to the EEC7 countries from the USA XQFNSt Export quantity of FCOJ to the European nonEEC3 countries from the USA Predetermined variables MQFSOct MPFSOct XQFOcSt RPShSt RQObSt WQkFSt-l WQsFFlt WQkOSt-l WQsOFlt WPFFlt-l WQOFlt-l WSkFFlt-l Import quantity of FCOJ by the USA from other countries Import price of FCOJ by the USA from other countries Export quantity of FCOJ to other countries from the USA Export quantity of other orange juices to Canada from the USA Retail price of frozen orange-fiavored synthetics and drinks in the USA Retail quantity of other orange juices in cardboard containers in the USA Wholesale quantity in stock of FCOJ in the USA in the previous period Wholesale quantity supplied of FCOJ in Florida Wholesale quantity in stock of other orange juices in the USA in the previous period Wholesale quantity supplied of other orange juices in Florida WPFFlt lag by one period WPOFlt lag by one period WSkFFlt lag by one period (continued)

PAGE 65

51 Table 11 (continued) Variables 1 WSkOFlt-l RxBSt Description WSkOFlt lag by one period Brazil exchange rate expressed in cruzeiros per USA dollar Income per capita in the USA, seasonally adjusted at annual rates, expressed in thousand dollars Current and lagged advertising expenditures (generic and brand) by the Florida Department of Citrus in the USA, expressed in dollar per thousand persons PdSt, PdSt-l Current and lagged estimates of orange production in the USA, expressed in billion boxes of 90 pounds PdSpt_ 6 Estimates of orange production in the state of Sao Paulo (Brazil) lag of six periods, expressed in billion boxes of 90 pounds PlSt, PlCt, PlEt, PlNt and PlOct Estimates of population in the USA, Canada, the EEC7, the European non-EEC3 and in other countries, respectively, expressed in million persons Dummy variables to account foE the effects of the USA freeze in 1977. Dt is equal to one for observations from February-March bimonth of 1977 on, and zero otherwise Linear time trend to account for time-related changes in prices (continued)

PAGE 66

52 Table 11 (continued) bl 1 Vari.a es Description Five (i=l, .. ,5) dummy variables to account for bimonth seasonal effects on demand. Zl is equal to one for December-January bimontSs, and zero otherwise. Similarly, Z2t through zst is equal to one for February-March, April May, June-July and August-September bimonths, respectively, and zero otherwise 1 unless otherwise specified non-monetary variables are expressed in gallons of single strength or reconstituted juice equivalent per thousand persons. Monetary variables are expressed in the USA dollar in terms of 1975 prices. Price variables are expressed in the USA dollars per gallon of single strength or reconstituted juice equivalent. A "t" subscript on a variable stands for "time period," which is from December-January bimonth of 1971/1972 to February-March bimonth of 1979. For more detail see Appendix E.

PAGE 67

53 previously (Chapter II), and the existence of systematic data for both of them. There is no doubt about the significance of estimating demand relations for FCOJ. However, the quantity of other forms of orange juice marketed in the USA has increased considerably. For example, Florida's wholesale domestic sales of canned and chilled orange juices increased by almost 70 percent from 1972 to 1979 (Table C.2. Appendix C). Thus, the inclusion of a set of demand and price equations for other orange juice is required to complete the model in this market. The retail and wholesale sectors for FCOJ and other orange juices are described by Equations 1 through 6 and by Equations 10 through 16. The wholesale sector is specified only for Florida since this state accounts for almost 90 percent of the total oranges processed in the country (Florida Crop and Livestock Reporting Service, 1979) and detailed data are not available for the other states. USA imports 11 are described by Equations 7 through 9. In the import segment only the demand structure for FCOJ imports is specified, since the USA does not import significant quantities of other forms of orange juices. As shown by the flow chart in Figure 3 all sectors of the USA market are linked together so that changes in any variable are likely to have some effects on all variables determined by the system. It should be pointed out that 11 USA export components will be described in the formulation of the demand structure of the other markets as defined in this study.

PAGE 68

54 wholesale prices, especially for FCOJ, and the FCOJ import price from Brazil, are variables that play major roles in the model. Wholesale prices are important determinants of retail prices in both forms of orange juices and wholesale prices of FCOJ together with the FCOJ import prices from Brazil determine the price spread, a key variable in this model. Another role of the domestic-import price relation would be in the derivation of the USA export prices which will be discussed in the next sub-section, where the FCOJ demand structures for other markets are conceptually formulated. The USA market structure for orange juices is conceptually quite similar to the model developed by Ward (1976a) 12 However, it is different in the sense that it also includes the retail sector of FCOJ and other orange juices. In addition, the wholesale sector also addresses the demand structure for other orange juices, and domestic and import prices are endogenous. Another important difference is the simultaneous linkage of major supply sources of orange juices in the most important markets (the USA, Canada, the EEC7 and the European non-EEC3). The remaining part of this sub-section will discuss the formulation of each equation in the system, as well as the expected directional effects of each variable. 12 ward's FCOJ model endogenously determines Florida's domestic demand and demand for imports (domestic and import prices are exogenous), and Florida's export quantities and prices to Canada and Europe.

PAGE 69

55 Retail FCOJ Retail quantity of FCOJ in the USA (RQFSt) is formulated as a function of its own price (RPFSt), prices of orange flavored synthetics and drinks and other orange juices as substitute products (RPShSt and RPOSt), income (ISt), advertising (ASt and ASt_ 1 ), and seasonal variation in consumption (Zlt through Z5t) (Equation 1). The hypothesized signs are consistent with traditional neoclassical demand theory and previous FCOJ demand research by Tilley (1979) and Ward and Tilley (1980). The impact of advertising on demand is assumed to have a lag structure of one time period. The underlying assumption is that some of the effects of advertising on consumption are expected to be reflected at later periods. Lee (1980) reviewed the Florida Department of Citrus, Economic Research Department research program in the area of measuring advertising effectiveness. A common agreement is that.the effects of advertising are not immediately perceived. That is, consumer responses to advertising have a decay or carryover effect over time. Since, in this model a time period aggregates two months (bimonthly data base), the specified structure of the advertising variables (Ast and ASt_ 1 ) will include four months, so that most of the effects of advertising on FCOJ retail demand are expected to be captured. FCOJ retail price (Equation 2) is assumed to be a positive fucntion of current and lagged wholesale price (WPFFlt and WPFFlt_ 1 ). Within this time period (four months)

PAGE 70

56 it is expected that market agents will have sufficient time to fully adjust retail prices to most changes in wholesale prices. Wholesale FCOJ The wholesale FCOJ demand structure is described by Equations 3 through 5. Florida's wholesale quantity {WQFFlt) is hypothesized to be negatively related to its own wholesale price (WPFFlt), and positively related to wholesale price of other orange juices (WPOFlt), retail quantity of FCOJ {RQFSt), and retail quantity of other orange juices {RQOSt). Seasonal variation (Zlt through Z5t) in the wholesale demand for FCOJ is also expected. Underlying most of this formulation is the assumption that wholesale demand is derived from retail demand. Retail quantities {RQFSt and RQOSt) are expected to translate consumers' preferences to the wholesale sector. Retail quantity of other orange juices {RQOSt) is hypothesized to affect FCOJ wholesale demand (WQFFlt) positively, since it can be reconstituted from bulk FCOJ. 13 Wholesale price of FCOJ in Florida (Equation 4) is assumed to be determined by an estimate 14 of wholesale stocks of FCOJ in the USA (WSkFSt), orange production in the USA (PdSt), a dummy variable to account for the effects 13 rt should be recalled that FCOJ, in this study, is defined so as to include all container forms of marketing the product. Florida's processors supply FCOJ in retail-size, bulk and institutional containers. 14 An estimate because data on WSkFSt determinant variables are available only for Florida processors.

PAGE 71

57 of the January of 1977 freeze (Dt), and a linear time trend variable expected to capture the effects of time-related changes (Tt). WSkFSt and PdSt are assumed to be negatively related to wholesale prices under a lag structure. The FCOJ stock variable (WSkFSt) is defined by equations 5 and 6 (identities) as the ratio of FCOJ quantity in stocks in the USA at the end of the bimonth 15 to Florida's domestic move ment (WQFFlt) for the same time period. Under this definition, WSkFSt is expressed in bimonths of domestic (Florida) supplies, such that the seasonal effect of stocks on wholesale prices can be accounted for. A specification that includes current and lagged FCOJ stock variables (WSkFSt and WSkFSt_ 1 ) is expected to capture most of its effects on FCOJ wholesale prices. Large stocks are expected to cause processors to reduce current prices in order to increase sales. Current and lagged orange production estimates (PdSt' PdSt_ 1 ) in the USA 16 are also assumed to affect WPFFlt. The reasoning for it is that changes in expected production would lead processors to adjust current prices, given anticipated juice supplies. The freeze in i977 (Dt) is expected to have a positive impact on wholesale prices, reflecting current and/or future reductions on supply. 15 FCOJ quantity in stocks in the USA at the end of the bimonth (Equation 6) is determined by the quantity in stocks in the USA from previous bimonth plus current Florida domestic supply and the USA imports from Brazil and from other countries minus current Florida domestic demand and the USA exports to Canada, the EEC7, the European non-EEC3, and other countries. 16 rt is assumed that the wholesale price equation for FCOJ (Equation 4), as well as the wholesale price equation for

PAGE 72

58 FCOJ Imports USA import demand for FCOJ from Brazil is specified in Equation 7 through 9. In this formulation, Florida's wholesale prices (WPFFlt and WPOFlt) and the import price of FCOJ from Brazil (MPFSBt) are the major factors influencing the USA import demand equations. The price spread (SpBt) between the FCOJ domestic wholesale price in Florida (WPFFlt) and the import price from Brazil (MPFSBt) plays a major role, as suggested by Ward (1976a), in the determination of USA imports from Brazil (Equations 7 and 8). SpBt is expected to be positively related to imports, meaning that, as this price differential increases (decreases), quantity of FCOJ imported from Brazil is likely to increase (decrease). While the preceding relationship determines the USA FCOJ import demand, the other variables in Equation 7 are assumed to be major import demand shifters. Wholesale quantity of other orange juices (WQOFlt) is expected to be positively reflected to FCOJ imports under the assumption that pressure from domestic wholesale sales of other forms of orange juices may lead to more FCOJ imports. other orange juices (Equation 14), are both affected by changes on orange production in the USA as whole, even through the equations are for Florida's orange juices. The reasoning for such formulation is as follows. If Florida's orange production changes, the state supply of orange juices, and consequently their prices, is very likely to change, given that, in Florida, practically all orange production is processed. If orange production in other states changes, the most likely result would be changes in the market for fresh oranges.

PAGE 73

59 Import price of FCOJ from other sources (MPFSOct), as the price of an expected substitute product, is hypothesized to be positively related to imports from Brazil (MQFSBt). In 17 this formulation stocks play a direct role on FCOJ imports. Imports are seen as the immediate source to raise stocks to some desired level, independent of future allocation. In this context stocks and imports move in opposite directions. Larger stocks should lessen the need for imports, while as stocks fall below some critical level, imports are viewed as immediately available supply. Even with stocks above the critical level, strong pressure from Canadian and European buyers of the USA product may lead to more imports in order to average down export prices. Bimonthly dummy variables (Zlt through Z5t) are included to account for seasonal influences on FCOJ imports. It is assumed that current and lagged Florida FCOJ wholesale price (WPFFlt, WPFFlt_ 1 ) are the major determinants of import prices from Brazil (MPFSBt) (Equation 9). In other words, to determine MPFSBt' Brazilian export agents and institutions first consider FCOJ market conditions in Florida, directly through market information and indirectly through actions of importers. WPOFlt and WPOFlt-l are included in the import price equation because pressure on the domestic wholesale market of other orange juices is 17 The indirect role of stock on imports comes from its effects on domestic prices (WPFFlt) and, consequently, on price spreads (SpBt) and on the level of imports.

PAGE 74

60 expected to put upward pressure on prices in that market. Brazilian FCOJ exporters may react to this movement by arguing for higher prices, assuming that some FCOJ imports are reconstituted and sold in that market. Sao Paulo's orange production (PdSpt) 18 enters the import price equation only under a lag of six periods (PdSpt_ 6 ), which is assumed to be the best formulation to capture the traditional negative effect of changing production on prices. This lag structure is directly associated with the fact that most Brazilian crop official estimates--as in the case of orangesare done in a yearly basis. Production reports are more likely to influence prices in the next period of production and marketing. The last determinant of import prices, as in Equation 9, is expected to be cruzeiros per USA dollar (RxBSt). negatively related to MQFSBt. Retail and Wholesale Other Orange Juices Most of the discussions with regard to FCOJ in the USA domestic markets can be fully extended to the formulation of demand structures for other forms of orange juices as described by Equations 10 through 16. In Equation 11, retail prices of other orange juices (RPOSt) are functions of WPOFlt, WPOFt-l and of the proportion of orange juices sold in cardboard containers (RfOSt), as defined by Equation 12. 18 h 1 d ~ 1 Int is case, on y orange pro uction in Sao Pau o is included as that state is the source of all FCOJ exports.

PAGE 75

61 It is expected that as this proportion increases retail prices will decrease because cardboard containers are cheaper than glass and cans. It should also be pointed out that the USA does not import other orange juices. In the determination of the wholesale stocks of other orange juices in the USA (SWkOSt), as described by Equation 15, the supply side (see Equation 16) is comprised only of the USA quantity in stocks from the previous period and current domestic supply (WQkOSt-l and WQsOFlt). The demand side includes Florida's domestic demand (WQOFlt) and exports to Canada (XQOCSt). USA exports of other orange juices to Europe are of relatively minor importance. FCOJ Market Structure in Canada The conceptual structure of the economic model showing the relations and interrelationships of variables for the Canadian and European markets is described by the flow chart in Figure 4. These segments of the model were simplified to include only the international trade sector and, specifically, only FCOJ exports to these markets from Brazil and from the USA. In most of these markets the domestic producing sector of orange juices is quite small or non-existent so that it is reasonable to assume that all domestic demand forces can be captured by the trade demand structures. Thus, the system of equations resulting from this formulation is a partially reduced form system. It is reduced in the sense that variables, characteristically defined as most likely to have a direct effect on wholesale (orange production) or on

PAGE 76

MPFCOct LEGEND Endogenous 0 Predetermined D variables variables Zit i=l, ... ,5 ,-/ WPFFlt / \ / ,_ / I \ MPFSBt \ / __ RxSEt Endogenous variables which derivation is not shown or or ... Solid arrows indicate \ I functional relations Figure 4. Orange juice market structure in Canada, the EEC7 and the European non-EEC3 countries. O'I N

PAGE 77

63 retail (income and advertising expenditures) sectors, are included in the international trade equations. It is partial because the set of equations still contains endogenous variables. The simplification that only Brazil and the USA export FCOJ to these markets is possible because other sources of orange juices have been supplying relatively small quantities. This formulation of Canadian and European demand for FCOJ from Brazil and the USA is in some way similar to the model developed by Tilley and Lee (1981) for the Canadian retail and import level purchases and prices for orange juices. The similarities are translated in the common way of treating some variables as endogenous or as pre determined. For example Canadian FCOJ trade with the two suppliers (quantities and prices) is determined by the system. However, in the Tilley and Lee model Canadian import prices from Brazil and the USA are exogenous. Another difference is that their model includes retail demand structures (quantity and price) for all orange juices, as well as an aggregated demand for other sources of Canadian imports. Major roles in the specification are again played by Florida's FCOJ wholesale price (WPFFlt) and by the import price of FCOJ from Brazil (MPFSBt). Together their important role is through the price spread (SpBt) in the determination of USA imports from Brazil and the impact of these imports

PAGE 78

64 on the USA export prices. It was shown that widening price spreads associated with declining import prices from Brazil or higher domestic prices are likely to result in more imports into the USA. Higher price spreads may lead to a greater export price discounting because more imports are available to average down export prices, and Brazilian imports also have a lower relative price. The net effect of a wider spread on USA exports will thus depend on the degree of discounting and the substitution relationships in the import markets. In the Canadian market four endogenous variables are determined within the structure (Equations 17 through 20) and there are 16 other variables in the system. Among these variables, WPFFlt and MPFSBt are endogenously determined in the USA market. The other variables are predetermined. The following equations represent the FCOJ market structure in Canada and Table 12 has a description of all variables of the structure. + + + XQFCBt = fll (XPFCBt' XPFCSt' MPFCOct, XPOCSt' XQFCBt = f 12 (XPFCS~, PdSp~_ 6 RxBC~) (18) XQFCSt fl3 (XPFCS~, + + + + = XPFCBt' MPFCOct' XPOCSt' !Ct' + ACt, + ACt-1' Zlt' Z2t' Z3t' Z4t' Z5t (19) fl4 + + + RxSC~) (20) XPFCSt = (WPFFlt' MPFSBt' MPFSOct,

PAGE 79

65 Table 12. Description of variables in the FCOJ market structure in Canada Variables 1 Description Endogenous variables XQFCBt XPFCBt XQFCSt XPFCSt WPFFlt MPFSBt Predetermined XPOCSt RxBCt RxCSt Export quantity of FCOJ to Canada from Brazil Export price of FCOJ to Canada from Brazil Export quantity of FCOJ to Canada from the USA Export price of FCOJ to Canada from the USA Wholesale price of FCOJ in Florida Import price of FCOJ by the USA from Brazil variables Import price of FCOJ by the USA from other countries Import price of FCOJ by Canada from other countries Export price of other orange juices to Canada from the USA Brazil exchange rate expressed in cruzeiros per Canadian dollar The USA exchange rate expressed in cents of dollar per Canadian dollar Income per capita in Canada, seasonally adjusted at annual rates, expressed in thousands of Canadian dollar Current and lagged advertising expenditures of the Florida Department of Citrus in Canada, expressed in dollars per thousand persons (continued)

PAGE 80

66 Table 12 (continued) Variables 1 PdSpt_ 6 Description Estimates of orange production in the state of S~o Paulo (Brazil) lag of six periods expressed in billion boxes of 90 pounds Five (i=l, ... ,5) dummy variables to account for bimonth seasonal effects on demand. Zlt is equal to one for December-January bimonths, and zero otherwise. Similarly Z2t through ZSt is equal to one for February-March, April-May, June-July and August-September bimonths, respectively, and zero otherwise. 1 unless otherwise specified non-monetary variables are expressed in gallons of single strength juice equivalent per thousand persons. Monetary variables are expressed in the USA dollar in terms of 1975 prices. Price variables are expressed in the USA dollars per gallon of single strength juice equivalent. A "t" subscript on a variable stands for "time period," which is from December-January bimonth of 1971/1972 to February-March bimonth of 1979. For more details see Appendix E.

PAGE 81

67 Brazilian FCOJ Exports to Canada In Equation 17, quantity of FCOJ exported to Canada from Brazil (XQFCBt) is hypothesized to be negatively related to export price (XPFCBt), and positively related to export price of FCOJ to Canada from the USA (XPFCSt), the import price of FCOJ by Canada from other countries (MPFCOct), the export price of other orange juices to Canada from the USA (XPOCSt), disposable income in Canada (ICt), and adver tising expenditures of the Florida Department of Citrus in Canada (ACt). XPFCSt' MPFCOct and XPOCSt enter the equation as competing substitute products differentiated by the supply source and/or the product characteristics. Florida's adver tising in Canada (ACt) is designed to promote orange juice from Florida. If, due to Florida's advertising, Canadian imports from the USA increase, the likely result would be a reduction in Canada's imports from other sources, assuming that other factors do not change. However, if the net effect of Florida's product advertising programs in Canada promotes orange juice in general, then it is possible that Florida advertising may have a positive effect on Canadian demand for FCOJ from other suppliers. Current and lagged advertising expenditures (ACt and ACt_ 1 ) are included to test these hypotheses. The remaining variables (Zlt through zs 5 account for seasonal effects. In the Brazilian export price equation to Canada (Equation 18), the USA FCOJ export price to Canada (XPFCSt) is included because it is expected that the USA acts as price

PAGE 82

68 leader. As the price changes, the Brazilian price (XPFCBt) changes in the same direction. Sao Paulo's orange production (PdSpt) and Brazil's exchange rate for the Canadian dollar (RxBCt) are expected to be negatively related to XPFCBt. Orange production in Sao Paulo enters the equation as a six-period lag (PdSpt_ 6 ) for the same reasons as in previous discussions. USA FCOJ Exports to Canada Export quantity of FCOJ to Canada from the USA (Equation 19) is expected to be negatively related to its own price (XPFCSt). It also has six export demand shifter variables (XPFCBt' MPFCOct' XPOCSt' ICt' ACt' ACt-l) that are expected to be positively related to XQFCSt. All the export demand shifter variables are expected to behave as in previous discussions. XPFCBt, MPFCOct and XPOCSt are prices of competing substitute produ~ts positively related to XQFCSt. ICt is disposable income in Canada, and ACt and ACt-l are Florida's promotional expenditures in Canada, both positively shifting the export demand. In Equation 20, the USA FCOJ export price to Canada (XPFCS) is assumed to be positively related to Florida's domestic wholesale price of FCOJ (WPFFlt) and to the import prices of FCOJ by the USA from Brazil (MPFSBt) and from other countries (MPFSOct). This specification recognizes, explicitly, the role of USA imports of FCOJ in export price determination and that USA prices may be affected by competitive prices of product from other sources.

PAGE 83

69 FCOJ Market Structure in Europe The European market structure is hypothesized to be similar to the Canadian market structure. There are only two suppliers (Brazil and the USA) and the structure is formulated only for FCOJ. The structure embraces the same partial reduced form concept and the key variables are the same (WPFFlt, MPFSBt and SpBt). The economic background on the relevant markets for orange juices presented in Chapter II stressed important differences in major European FCOJ markets. On one side the European Economic Community (EEC) countries were said to be quite important due to their market size and specific economic policies with respect to orange juice imports into those countries. On the other side, European countries not EEC members were said to be also important because of their high rate of per capita consumption o~ orange juices, market sophistication as far as juice quality and packing procedures are concerned, and the fact that, as non-EEC members, they do not have the same structure of protective restrictions and preferential concessions as the EEC members. In order to deal with these different economic aspects, the FCOJ demand structures in Europe are formulated for two groups of countries. The EEC7 group (West Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark and France) represents EEC members. The European non-EEC3 group (Sweden, Norway and Finland) represents other European non-EEC members.

PAGE 84

70 The EEC? and European non-EEC3 market structures are described by Equations 21 through 28. XQFEBt = fl5 (XPFEB~, + XPFESt, + Tt, + AEt, Zlt, Z2t, Z3t, Z4t, Z5t) ( 21) fl6 + PdSp~_ 6 RxBE~) (22) XPFEBt = (XPFESt, fl7 (XPFES~, + + + Zlt, Z2t, XQFESt = XPFEBt, Tt, AEt, Z3t, Z4t, Z5t) ( 23) fl8 + + RxSE~) ( 2 4) XPFESt = (WPFFlt, MPFSBt, MPFSOct, fl9 (XPFNB~, + + + Zlt, Z2t, XQFNBt = XPFNSt, Tt, ANt, Z3t, Z4t, z5t ( 25) f20 + PdSp~_ 6 + (26) XPFNBt = (XPFNSt, RxBNt) f21 (XPFNS~, XPFNB+, + + Zlt, Z2t, XQFNSt = Tt, ANt, Z3t, Z4t, z5t (27) XPFNSt = f22 + (WPFFlt, + MPFSBt, + MPFSOct, RxSN~) (28) Eight endogenous variables are determined by the structure. Wholesale prices in Florida (WPFFlt) and the USA import price from Brazil (MPFSBt) are endogenous variables in the system determined in the USA market. The other variables are predetermined. A description of all variables of the FCOJ market structure in the EEC7 and the European non-EEC3 countries is presented in Table 13.

PAGE 85

71 Table 13. Description of variables in the FCOJ market structures in the EEC? and the European non-EEC3 countries Variables 1 Description Endogenous variables XQFEBt XPFEBt XQFESt XPFNSt Export quantity of FCOJ to the EEC? from Brazil Export price of FCOJ to the EEC? from Brazil Export quantity of FCOJ to the EEC? from the USA Export price of FCOJ to the EEC? from the USA Export quantity of FCOJ to the European non EEC3 from Brazil Export price of FCOJ to the European non-EEC3 from Brazil Export quantity of FCOJ to the European non EEC3 from the USA Export price of FCOJ to the European non-EEC3 from the USA Wholesale price of FCOJ in Florida Import price of FCOJ by the USA from Brazil Predetermined variables MPFSOct RxBEt Import price of FCOJ by the USA from other countries Brazil exchange rate index relative to the EEC7 countries Brazil exchange rate index relative to the European non-EEC3 countries (continued)

PAGE 86

72 Table 13 (continued). Variables 1 ANt PdSpt-G Description The USA exchange rate index relative to the EEC7 countries The USA exchange rate index relative to the European non-EEC3 countries, expressed in dollars per thousand persons Advertising expenditures of the Three-Party Program in the European non-EEC3 countries, expressed in dollars per thousand persons Estimates of orange production in the state of Sao Paulo (Brazil); lag of six periods expressed in billion boxes of 90 pounds. Time trend variables as a proxy variable for income per capita in the EEC7 and the European non-EEC countries Five (i=l, .. ,5) dummy variables to account for bimonth seasonal effects on demand. Zlt is equal to one for December-January bimonths, and zero otherwise. Similarly Z2t through Z5t is equal to one for February-March, April-May, June-July and August-September bimonths, respectively, and zero otherwise. 1 unless otherwise specified non-monetary variables are expressed in gallons of single strength juice equivalent per thousand persons. Monetary variables are expressed in the USA dollar in terms of 1975 prices. Price variables are expressed in USA dollars per gallon of single strength juice equivalent. A "t" subscript on a variable stands for "time period" which is from December-January bimonth of 1971/1972 to February-March bimonth of 19 79. For more details see Appendix E.

PAGE 87

73 Brazilian FCOJ Exports to the EEC7 Export quantity of FCOJ to the EEC7 countries from Brazil (Equation 21) is hypothesized to be negatively related to its own price (XPFEBt) and positively related to the USA export price to the same market (XPFESt). Tt is a time trend variable used as a proxy for disposable income in the EEC7 market, which is also expected to be positively related to demand for FCOJ from Brazil. AEt stands for advertising expenditures in the EEC7 market and it includes only financial support of the Three-Party Program which is designed to promote Florida's citrus in Europe. Three-Party Program data are available only on a yearly basis and AEt accounts for advertising expenditures in the seven countries of the EEC7 market. The impact of the program expenditures may be positive or negative, depending on whether the promotion of FCOJ from Florida causes a substitution of the Florida product for the Brazilian product, or creates an umbrella effect that causes orange juice from both the USA and Brazil to be purchased. Zlt through zst are seasonal measures of demand with the same objectives and expected behavior as in previous market demand structure specifications. Equation 22 describes the derivation of Brazil's FCOJ export price to EEC7 (XPFEBt). XPFEBt is expected to be positively related to the USA export price of FCOJ to the same market. Orange production in Sao Paulo enters the equation under the same specification as in the previous case (PdSpt_ 6 ) and with the same negative expected relationship.

PAGE 88

74 Exchange rate in this case is summarized as an index value of the Brazilian cruzeiro against currencies of EEC7 countries. 19 Its effect on the export price from Brazil to EEC7 (XPFEBt) is expected to be positive, since the index is an inverse function of the exchange rate values of individual countries. For example, if exchange rates associated with individual countries increase (devaluations), then the index is expected to decrease as would the corresponding export price. USA FCOJ Exports to the EEC7 The USA export quantity of FCOJ to the EEC7 (XQFESt), as specified by Equation 23, has the same logical formulation as discussed in the Brazilian case, with the only difference being that AEt now is expected to have a positive effect on the USA FCOJ export demand to the EEC7 market. USA export price to the EEC7 (Equation 24) is hypothe sized to be positively related to Florida's FCOJ wholesale prices (WPFFlt), to the import prices of FCOJ by the USA from Brazil and from other countries (MPFSBt and MPFSOct), and to the USA exchange rate index for the EEC7 countries (RxSEt). Florida's FCOJ wholesale price (WPFFlt) has, as already discussed, a direct and an indirect role in the derivation of export prices. Directly, since an increase in the wholesale price is likely to lead to an increase in export 19 For more details on the construction of the exchange rate index see Appendix D.

PAGE 89

75 prices, and indirectly through price spreads in the derivation of FCOJ imports into the USA that may be used to average down export prices. This last rationale also explains the role of FCOJ import prices by the USA in the export price equation. Brazilian and USA FCOJ Exports to the European non-EEC3 The formulation of the European non-EEC3 market structure for FCOJ from Brazil and USA is conceptually the same as in the case of the EEC7 market. Equations 25 through 28 are the same as 21 through 24 except that Nin these formulations means the European non-EEC3 countries. Estimation h d b f d. 20 f Tis section provi es a rie iscussion o a system of simultaneous linear equations and of its major elements, the procedure used to estimate the empirical model, and a test of model validation. From previous discussions in this chapter with regard to the theoretical setting and model formulation the simultaneous nature of the economic model of this study is quite clear. Most equations in the system include more than 21 one endogenous variable. This condition introduces a 20 Most of this discussion is organized from Goldberger, 1964; Kennedy, 1979; Kmenta, 1971; and Maddala, 1977. 21 rn the economic model there are 28 endogenous variables as defined by equations 1 through 28 (left hand side variables) and, among the explanatory variables (right hand side variables), 41 are predetermined variables (the ones that are explanatory, but not current endogenous variables).

PAGE 90

76 violation in the set of assumptions of the classical linear regression model. If endogenous variables are used as explanatory variables it can be proved (Goldberger, 1974, p. 305-306) that they are contemporaneously correlated with the error term in all other equations in the system. As a result, the ordinary least squares (OLS) estimator would be biased, even asymptotically, and, consequently, an alternative estimator should be used. By adding an error term to each behavioral equation of the system formulated in the previous section (called economic model), the econometric model of this study is formally specified in the so-called structural form. In matrix notation, the whole structure can be written as: SY(t) + rx(t) = U(t) (29) where Sis the M by M matrix of coefficients of the endogenous variables; Y(t) is the M by 1 row vector of the t th observation on the endogenous variables; r is the M by K matrix of coefficient of the predetermined variables; X(t) is the K by 1 row vector of the tth observation on the predetermined variables; and U(t) is the M by 1 row vector of the tth (unobserved) values of the error terms. Each error term should satisfy the assumptions of the classical linear regression model, that is, in matrix notation, U(t) = N(0,), E [U ( t ) u' ( s) ] = 0, t :/ s ( 3 0) (31)

PAGE 91

77 and E [ U ( t) U ( s) ] = t= s where is the M by M variance-covariance matrix (non negative definite) 22 of the structural error term. The underlying assumptions imply that each structural error vector (a) is normally distributed with mean zero and constant (homoskedastic) unknown variance and (b) is nonautoregressive, implying that error terms are contemporaneously uncorrelated. (32) If the structural-form of the system (Equation 29) is solved for the endogenous variables, that is, by expressing endogenous variables in terms of predetermined variables and error terms, the result is the so-called reduced-form of the system (Equation 33 or 34). 23 ( 3 3} Y(t} = ITX(t} + V(t} (34) Equations 33 and 34 are called, respectively, restricted and unrestricted versions of reduced-form equations of the system. If structural-form estimates are used to derive the reduced-form (Equation 33), restrictions imposed on the structural-form formulation are accounted for through the estimates of Band r. The same is not necessarily true if 22 In the case of existence of identities, refers only to the equations that are not identities. 23 In solving for reduced form, the matrix Bin Equation 29 is implicitly assumed to be a nonsingular matrix.

PAGE 92

78 the reduced-form parameters are estimated directly on the predetermined variables (Equation 34). From a direct comparison of these two equations, if follows that: and -1 TI= -S (35) ( 3 6) where IT is the M by K matrix of reduced-form coefficients of predetermined variables, and V(t) is the M by 1 row vector of the t th (unobserved) values of the reduced-form error terms. The variance-covariance matrix (~) of the reduced-form error terms can be proved to be given by (37) It should be pointed out that in the reduced-form system each endogenous variable is expressed in terms of only predetermined variables and an error term, after accounting for the interdependence among current endogenous variables. Thus, a given reduced-form coefficient indicates the total effect of a change in its respective predetermined variable on the corresponding endogenous variable, assuming that other variables are held constant. These reduced-form coefficients are called long-run multipliers associated with the model (Kennedy 1979, p. 107). In contrast, a structural form coefficient indicates only its variable partial direct

PAGE 93

79 effect on the respective dependent endogenous variable. However, estimates of the structural-form coefficients are quite important for identifying the underlying economic hypotheses used as major support in developing the model. Thus, depending on a study's objectives, different strategies can be used to estimate reduced-forms--directly from values of predetermined variables as in Equation 34, or deriving it, as in Equation 33, by first estimating structural-form parameters. Associated with these strategies is the identification problem in simultaneous equation systems. It is a mathematical problem of going from estimates of the reduced-form structure back to meaningful estimates of the structural-form parameters. An equation can be said to be overidentified, or exactly identified, or underidentified by zero restrictions on certain parameters, if there are more than enough, or exactly enough, or less than enough predetermined variables excluded from the given equation to act as instruments of endogenous explanatory variables used in that equation. By applying the usual tests for identification (rank and order conditions), the structural form equations of this study were found to be overidentified. Various procedures are suggested in the literature to estimate a system of simultaneous equations. Based on the objectives of this study, the problems associated with the 24 A summary of these procedures is presented by Kennedy, 1979, p. 112-126.

PAGE 94

80 formulation of its model, 25 the nature of the model identification problem, and major characteristics of available procedures, the two-stage least squares (2SLS) procedure was chosen to estimate the parameters of the over-identified structural-form equations of the system. This procedure not only gives consistent estimators of the structural parameters, but studies have also shown it to have small-sample properties superior on most criteria to all other estimators, in addition to being quite robust (Kennedy 1979, p. 115). However, 2SLS estimators are not invariant with respect to the endogenous variable which is normalized. One would like this property in a simultaneous equation estimator. Basically, the 2SLS procedure requires that, in the first stage, each endogenous variable acting as explanatory variable be regressed on all the predetermined variables of the system (estimation of the reduced-form version) and their estimated values are calculated. In the second stage these estimated values (called instrumental variables) of endogenous variables and the included exogenous variables are used in an OLS regression. 25 h f f. h f 1 T ere are some errors o speci 1cat1on int e ormu ation of the structural form version of the model, especially in the case of Canadian and European market structures, where lack of systematic data for retail and wholesale sectors forced the formulation of only import equations. Available retail and/or wholesale variables were used as explanatory variables of import equations. Single-equation estimation procedures are said to keep these misspecification problems within their source in the system, rather than to incorporate them in the estimates of all the structural parameters as in the case of systems methods procedures (Kennedy, p. 116).

PAGE 95

81 For testing how well values predicted by the model conform to observed data, a test based on the ratio of the root mean square error to the respective observed mean value of endogenous variables for different periods within the data range is used. For a given endogenous variable and the ith time period, the ratio (RM.) is given by l. RM. l. = /M. l. (38) Where, Pt and At stand, respectively, for predicted (fore casted) and actual (observed) values; N. is the number of l. observations in the ith time period; and M. is the corre1. spending observed mean value of the endogenous variable. This ratio will give an idea of the magnitude of the fore cast error relative to the respective actual mean value for the endogenous variable in consideration.

PAGE 96

CHAPTER IV EMPIRICAL RESULTS AND IMPLICATIONS This chapter is organized in two major parts with the common goal of discussing the empirical results and impli cations of estimates of the statistical model. In the first section, 2SLS estimates of the structural-form equations are presented and discussed. In the second section, estimates of the derived reduced-form equations are presented and discussed. Validation tests of the model in its derived reduced-form version are also presented and discussed in the second section. Estimates of the Structural-Form Equations In this section, th~ structural-form estimates of the model are presented and discussed for each of the four markets in this study. Special emphasis is given to the implications of these results within a given market context and to the market supply sources. Economic analysis in this section is intended to evaluate the underlying theoretical framework of the model, as well as particular implications of the results. Comparisons with other findings are also made, whenever possible. Explanatory variables are listed in the tables as the respective structure where theoretically specified in the 82

PAGE 97

83 model. This, however, does not necessarily mean that all of those variables are included in the final estimates of the equation. The USA Market Most of the structural-form equations were estimated as stated in the theoretical formulation in Chapter III. How ever, some variables were deleted from specific equations due to multicolinearity, 1 plausibility of the respective signs, and/or the magnitude of a given estimate relative to its standard error. In this particular market, only the equations concerning FCOJ imports from Brazil (Equations 8 and 9), and the equations for the wholesale level of other orange juices in Florida (Equations 13 and 14) deviate from the theoretical specification. Practically, all signs of the estimated coefficients were as theoretically hypothesized. 2 Most of the estimated coefficients are considerably larger than their estimated standard errors. Retail FCOJ Structural-form estimates for retail FCOJ demand (RQFSt) and the price (RPFSt) equations are presented in Tables 14 and 15. Except for RPOSt in the FCOJ retail demand equation, 1 Multicolinearity was detected based on the estimates of the correlation matrices of the parameters. Only extremely high levels of multicolinearity were regarded as major problems, given that 2SLS has been shown to be quite insensitive to this problem (Kennedy 1979, p. 115). 2 Means, standard deviations and coefficients of variation for variables used in the model are presented in Appendix F.

PAGE 98

84 Table 14. Structural-form estimates of the FCOJ retail demand equation in the USA market (RQFSt, Equation 1) Explanatory Coefficient Standard variables a error Intercept 204.6033 238.2164 Endogenous variables RPFSt -98.6950 68.8457 RPOS~ -112.2447 59.4951 Predetermined variables RPShS~ 128.9164 73.3681 IS+ t 62.1050 28.7421 AS+ t 1.4025 1.1662 + ASt-1 .8982 .8695 Zlt 17.5453 8.4340 Z2t 14.2254 8.2063 Z3t -10.2295 7.4022 Z4t -20.6681 8.3289 z5t -10.9591 7.3398 1 variable superscript indicates the a priori expected sign.

PAGE 99

85 Table 15. Structural-form estimates of the FCOJ retail price equation in the USA market (RPFSt, Equation 2) Explanatory 1 Coefficient Standard variables error Intercept .4532 .0801 Engodenous variables WPFFl~ .8201 .1780 Predetermined variables + WPFFlt-l .2638 .1759 1 variable superscript indicates the a priori expected sign.

PAGE 100

86 all other coefficients were consistent with expectations. The negative effect of prices of other orange juices (RPOSt) on retail FCOJ demand (Table 14) does not necessarily imply a departure from the theoretical setting where a positive response was expected. Similar problems of consistency in orange juice demand estimates were reported by Tilley (1979), Ward and Tilley (1980), and Malick (1980). A common agreement of these studies is that no statistical substitution evidences were found between FCOJ and chilled orange juice (COJ). Tilley, and Ward and Tilley also found that canned single strength orange juice (CSSOJ) was negatively related to FCOJ 3 demand. The net effect of a change in the price of any commodity (P.) on the quantity demand of any other good (Qi) J can be decomposed into a substitution effect and an income 4 effect (Phlips, 1974, p. 40-47). Since the income effect of price changes is not necessarily symmetric and that it will be larger for a commodity which takes a larger proportion of total expenditures (Tomek & Robinson, 1981, p. 56), then the negative effect of RPOSt on the demand for FCOJ does not necessarily invalidate the expected substitute characteristic between the two products. This result should also be 3 rt should be recalled that in this study orange juices are grouped as frozen concentrated (FCOJ) and other orange juices. The latter includes all other forms of orange juices but FCOJ. At retail level, other orange juices group is basically chilled and canned single strength (COJ and CSSOJ). 4 rn terms of partial derivatives such decomposition can be written as Qi/oPj = Kij Qj (oQi/oY), where Kij is the substitution effect and Qj (oQi/oY) is the income effect. Note that Qj (oQi/oY) is not necessarily equal to Qi (aQj/oY1.

PAGE 101

87 independent of secondary reasons leading consumers to purchase 5 the two products. To rationalize such results, it is necessary to have a larger absolute value of the income effect relative to the substitution effect and that the FCOJ share of total expenditures be relatively larger than the share of other orange juices, which seems to be the case. Among the other explanatory variables of retail FCOJ demand in the USA, income (ISt) and seasonality (Zlt through Z5t) seem to be quite important. The estimated income coefficient is more than twice its standard error, suggesting that FCOJ consumption is quite sensitive to changes in real income per capita. Seasonal estimates suggest that FCOJ consumption increases during relatively colder months (October to March) and declines during those months that are relatively warmer (April to September). Estimated coefficients for advertising variables (ASt and ASt_ 1 ), even though positive are only slightly greater than their respective standard errors. Price of frozen orange-flavored synthetics and drinks (RPShSt) was found to be positively related to RQFSt, as expected, suggesting a substitution relation between the two products. Retail price of FCOJ (Table 15) was found to be highly sensitive to current and lagged wholesale prices of FCOJ in Florida (WPFFlt and WPFFlt-l). 5 ward and Tilley (1980, p. 10) argued that the failure of their estimates to show substitution between FCOJ and COJ is likely to be related to consumer convenience (by purchasing COJ) versus storability (by purchasing FCOJ).

PAGE 102

88 Estimated elasticities and flexibilities evaluated at the means of the variables from the structural-form estimates of the model are presented in Appendix G. It should be emphasized, however, that these structural-form elasticity and flexibility estimates are partial, in the sense that they indicate only the respective direct effect within a given equation. Total effects of changing predetermined variables on endogenous variables and the consequent effects among endogenous variables will be presented in the discussion of the estimates of the derived reduced-form equations. The value of the own-price elasticity is estimated to be -.4180 and the income estimate is .8948. In the derivation of retail prices (RPFSt), changes in current and lagged FCOJ wholesale prices in Florida (WPFFlt and WPFF1t_ 1 ) are quite important with current and lagged price transmission elasticities estimated to be .5328 and .2642. Wholesale FCOJ Tables 16 and 17 present the estimated coefficients of the wholesale FCOJ structure in Florida. All estimates have signs consistent with the conceptual formulation and most of them are large relative to their respective standard errors. Among the major determinants of wholesale FCOJ demand in Florida (Table 16), retail consumption of other orange juices (RQOSt) seems to play a significant role, given that the estimated coefficient is almost four times larger than

PAGE 103

89 Table 16. Structural-form estimates of the FCOJ wholesale demand equation in Florida (WQFFlt, Equation 3) Explanatory 1 variables Intercept Endogenous variables WPFFlt WPOFl: RQFS: RQos: Predetermined variables Coefficient 136.0881 -358.6431 245.0809 .5817 1.4439 60.1773 41. 7583 -11.8118 18.1411 32.2708 Standard error 323.6901 171.2662 117.8307 .6550 .4033 27.9097 28.7496 25.9771 29.7190 29.1060 1 variable superscript indicates the a priori expected sign.

PAGE 104

90 Table 17. Structural-form estimates of the FCOJ wholesale price equation in Florida (WPFFlt, Equation 4) Explanatory 1 variables Intercept Endogenous variables WSkFS~ Predetermined variables WSkFS~-l PdS~ PdS~-l D+ t Tt Coefficient 2.3476 -.0138 -.0275 -3.4918 1.8793 .2874 -.0050 Standard error .1845 .0165 .0168 1. 2589 1.2442 .0560 .0020 1 variables superscript indicates the a priori expected sign.

PAGE 105

91 its standard error. As retail sales of other orange juices increase the wholesale demand for FCOJ also increases, since wholesale FCOJ in this formulation includes the product in bulk form that can be reconstituted into chilled and canned single strength orange juices. The poor response of Florida's wholesale FCOJ sales to changes in current FCOJ retail consumption reflects the relative stability of FCOJ consumption rate in the USA and the ability of retailers to sell from accumulated inventory. Seasonal variations in Florida's wholesale FCOJ demand are quite consistent with the retail market where greater sales are in colder bimonths (December-January and February-March). All of Florida's wholesale FCOJ price function variables (Table 17) had coefficients more than 1.5 times their respective standard errors, except current stocks (WSkFSt). Orange production forecasts seem to have a major role in the determination of Florida's wholesale FCOJ prices. Current and previous crop estimates (PdSt and PdSt_ 1 ) have a strong effect on Florida's wholesale FCOJ prices, showing that processors anticipate future orange supplies. The effects of the freeze on Florida's wholesale FCOJ price as measured by the estimated coefficient of the dummy variable Dt are quite significant. The result indicates that after a freeze, WPFFlt is likely to increase by 29 cents per single strength equivalent gallon, assuming that all other price determinants do not change. The linear time trend variable (Tt) accounting for time-related changes in the wholesale FCOJ

PAGE 106

92 price had a negative estimated coefficient, meaning a net decrease in the WPFFlt over time. Structural-form elasticity and flexibility estimates for Florida's wholesale FCOJ sector are in Appendix G. The wholesale FCOJ own-price elasticity estimate is -.6618 which is relatively more elastic than the corresponding elasticity at retail level (-.4180). Ward's (1976a) estimate of Florida's wholesale FCOJ domestic demand elasticity (-.5230) is not greatly different from the estimate in this study. The cross-price elasticity with respect to wholesale price of other orange juices in Florida (WPOFlt) is estimated to be .6141, indicating that if processors raised WPOFlt by 1 percent, FCOJ wholesale movement is likely to increase by a little more than .5 percent. FCOJ and other orange juices retail price transmission elasticities are similar in magnitudes (.3901 and .3621, respectively). Wholesale FCOJ prices in Florida are practically insensitive to changes in current and lagged FCOJ stocks in the USA (WSkFSt and WSkFSt-l) as measured by the respective values of stock flexibilities (-.0374 and -.0742). Orange production forecast flexibilities indicate that a one percent increase (decrease) in current and previous crop estimates (PdSt and PdSt_ 1 ) will be associated with a .7959 and .4276 percent decline (increase) in the wholesale price of FCOJ.

PAGE 107

93 FCOJ Imports from Brazil In the structural-form equations (Tables 18 and 19) for the USA FCOJ Brazilian import (quantity and price), most of the variables have estimated coefficients larger than their associated standard errors Stock variables (WSkFSt, WSkOSt and lags) were deleted from the final estimates due to their persistent sign inconsistencies and multicolinearity with other variables in the model. Price spread (SpBt) carries in its definition (WPFFlt MPFSBt) the effects of FCOJ stocks on imports through the derivation of the wholesale FCOJ price in Florida (WPFFlt). Wholesale demanded quantity of other orange juices (WQOFlt) includes the effects of the respective level of stocks on FCOJ imports through the specification of wholesale price of other orange juices. The FCOJ import price for the USA from other countries (MPFSOct) was excluded, given its high correlation with the import price from Brazil (MPFSBt) already included in the formulation of the price spread (SpBt). Current and lagged wholesale price of other orange juices in Florida (WPOFlt and WPOFlt_ 1 ) were deleted from the USA FCOJ import equation (Table 19) due to their very low coefficients relative to the corresponding standard errors. The price spread (SpBt) coefficient in the structural form import equation indicates that a 10 ~ent increase in the difference between Florida's wholesale FCOJ price (WPFFlt) and the import price from Brazil (MPFSBt) is likely to increase FCOJ imports from Brazil by practically 11 single strength juice equivalent gallons per thousand persons.

PAGE 108

94 Table 18. Structural-form estimates of the USA FCOJ import demand equation from Brazil (MQFSBt' Equation 7) Explanatory 1 variables Intercept Endogenous variable + SpBt WQOFl~ WSkFS~ WSkOS~ Predetermined variables WSkFS~-l WSkOS~-l + MPFSOct Zlt Z2t Z3t Z4t z5t Coefficient -188.1535 107.4650 1. 2070 11.7234 -16.3451 -21. 5905 -29.9382 -22.6145 Standard error 52.7638 61.1746 .3265 21. 5108 21.6607 20.8051 21.0097 21. 0708 1 variable superscript indicates the a priori expected sign.

PAGE 109

95 Table 19. Structural-form estimates of the USA FCOJ import price equation from Brazil (MPFSBt' Equation 9) Explanatory 1 variables Intercept Endogenous variables WPFFl~ WPOFl~ Predetermined variables + WPFFlt-l + WPOFlt-l PdSp~_ 6 RxBSt Coefficient -.1058 .2332 .4637 -.7697 -.0095 Standard error .4830 .1715 .1683 .5799 .0579 1 variable superscript indicates the a priori expected sign.

PAGE 110

96 Widening the price spread for a given import tariff, or reducing the tariff level for a given price spread would result in an increase in FCOJ imports into the USA. Thus the effects of price spread changes are equivalent to the effects of changing tariff on imports. However, since a change of tariff has associated with it some change in the structure of exports related to a given duty drawback provision, then the effects of changing tariff may differ from the effects of changing price spreads (Ward, 1976a). Estimated coefficients for seasonal variables do not seem to have significant impact on FCOJ imports from Brazil. Import price equation coefficient estimates (Table 19) support the conceptual hypotheses that the USA FCOJ import price from Brazil is determined by Florida's wholesale price and Sao Paulo state orange production. If other factors are kept constant, a 10 cent change in Florida's current whole sale price for FCOJ (WPFFlt) is likely to result in a 2.3 cent increase in the import price (MPFSBt). Likewise, a 10 cent increase in Florida's previous wholesale price (WPFF1t_ 1 ) will result in a 4.6 cent increase in MPFSBt. Thus, the combined current and lagged effect coefficients indicate that approximately 7 cents of any 10 cent change in Florida prices will be reflected in a change in the Brazilian price. So, given a change in WPFFlt' Brazilian exporters do not fully readjust their prices immediately. orange production in the state of Sao Paulo in the previous year (PdSpt_ 6 ) had an estimated coefficient of -.7697.

PAGE 111

97 Brazilian exchange rate policy appears to have no significant impact on FCOJ import price in the USA, given the absolute magnitude of the estimated coefficient for RxBSt (cruzeiros per USA dollar) and its size relative to its standard error. However, as was pointed out in the model formulation, such results were likely to come about, since exchange rates are directly controlled by the government in Brazil. Current rates of cruzeiros per dollar are set almost every month so as to keep this relation at some constant value. Structural-form elasticity and flexibility results are shown in Appendix G. A price spread elasticity 6 of 1.6607 indicates that for each 1 percent increase in the spread of imports into the USA from Brazil should increase by nearly 1.7 percent. Wholesale other orange juice transmission elasticity is 5.0697 indicating that a 1 percent increase in WQOFlt will increase USA FCOJ imports from Brazil by more than 5 percent. Current and lagged FCOJ wholesale price transmission elasticities (.5123 and 1.0201) indicated that current import price adjustment due to changes in WPFFlt-l is almost twice as large as adjustment relative to changes in current wholesale prices in Florida (WPFFlt). That is, a 1 percent increase in WPFFlt in the current bimonth will result in 6 ward's (1976a) study using double-log forms for a system of simultaneous equations and quarterly data estimated the price spread elasticity for Florida's FCOJ imports from Brazil as 2.6986.

PAGE 112

98 nearly .5 percent increase in MPFSBt' while the same change in Florida's FCOJ wholesale price in the previous bimonth will result in more than a 1 percent increase in current import price from Brazil. Exchange rate transmission elasticities (-.1642) indicate that for each 1 percent increase in rate of cruzeiros per dollar, the USA import price from Brazil will decrease by less than .2 percent. Likewise, an increase in orange production in the state of Sao Paulo in the previous year is likely to reduce MPFSBt by only .1 percent. Retail and Wholesale Other Orange Juices Structural-form estimates for other orange juices at retail level in the USA and at wholesale level in Florida are presented in Tables 20 through 23. Most of the variables in the final estimated structures are large relative to their respective standard errors. Wholesale FCOJ price was excluded from Equation 13 (Table 22) and lagged wholesale stock of other orange juices was excluded from Equation 14 (Table 23), due to their small coefficient values relative to their respective standard errors. As expected from the conceptual formulation, retail FCOJ prices are positively related to retail quantity of other orange juices indicating that FCOJ is a substitute for other orange juices (.Table 20). This result supports the argument developed in the discussion of retail FCOJ demand estimates in the sense that other orange juices are expected to take a relatively small share of total expenditures and

PAGE 113

99 Table 20. Structural-form estimates of the other orange juices retail demand equation in the USA market (RQOSt, Equation 10) Explanatory 1 variables Intercept Endogenous variables RPOSt RPFS~ Predetermined variables RPShS~ IS+ t AS+ t + ASt-1 Zlt Z2t Z3t Z4 t zst Coefficient -425.7334 -136.6758 101. 2446 140.3181 89.8842 .2961 .3790 -2.1561 -2.8706 -7.7425 -1. 6839 4.0167 Standard error 190.1408 47.4881 54.9516 58.5613 22.9415 .9308 .6941 6.7319 6.5502 5.9083 6.6480 5.8585 1 variable superscript indicates a priori expected sign.

PAGE 114

100 Table 21. Structural-form estimates of the other orange juices retail price equation in the USA market (RPOSt' Equation 11) Explanatory 1 Coefficient Standard variables error Intercept .8159 .1128 Endogenous variables WPOFl~ .1447 .2267 RpOSt -.2241 .1163 Predetermined variables + WPOFlt-l .8336 .2134 1 variable superscript indicates the a priori expected sign.

PAGE 115

101 Table 22. Structural-form estimates of the other orange juices wholesale demand equation in Florida (WQOFlt, Equation 13) Explanatory 1 variables Intercept Endogenous variables WPOFl~ WPFFl~ RQOS~ Predetermined variables Coefficient 91. 9673 -24.9021 .5831 15.3357 14.4719 6.7412 13.0681 12.1645 Standard error 9.4146 6.7257 .0303 3.2232 3.2403 3.2245 3.2259 3.2198 1 variables superscript indicates the a priori expected sign.

PAGE 116

102 Table 23. Structural-form estimates of other orange juices wholesale price equation in Florida (WPOFlt, Equation 14) Explanatory 1 variables Intercept Endogenous variables WSkOS~ Predetermined variables WSkOS~-l PdS~ PdS~-l D+ t Tt Coefficient 2.3323 -.0209 .3179 -4.4162 .3163 -.0056 Standard error .1961 .0334 1.3543 1. 3513 .0549 .0023 1 variable superscript indicates the a priori expected sign.

PAGE 117

103 consequently the income effect should be smaller in absolute value. Thus, the net cross-price effect on RQOSt from changes on RPFSt is positive. Similar results were found by Tilley (1979), and Ward and Tilley (1980). Other orange juices retail demand in the USA (RQOSt) is quite sensitive to income changes and quite insensitive to advertising expenditures. Seasonality seems to have no practical effects on the consumption of other orange juices as measured by the Zlt through zst estimated coefficients and respective standard errors. Lagged wholesale prices in Florida (WPOF1t_ 1 ) and the proportion of other orange juices sold in cardboard containers (RpOSt) are quite important determinants of the retail price of other orange juices (Table 21). A 10 cent increase on WPOFlt-l is likely to increase current retail prices (RPOSt) by more than 8.3 cents per gallon. All coefficients for the variables included in the estimates of other orange juices' wholesale movement in Florida (Table 22) are large relative to their respective standard errors. Seasonal variable (Zlt to ZSt} estimates indicate that other orange juice movement is practically uniform during the season. Only during the bimonth of April and May (Z3t) does the movement of other orange juice increase by nearly half of the increments in the other bimonths. Empirical estimates for the other orange juice wholesale price equation (Table 23) indicate that the

PAGE 118

104 lagged orange production forecast in the USA (PdSt_ 1 ), freeze effects (Dt), and time-related changes in prices (Tt) are important factors determining price. The effect of freeze seems to be very relevant since its occurrence is likely to increase WPOFlt by approximately 31 cents per gallon in real terms. Other orange juice stocks in the USA (WSkOSt) do not have important effects on Florida's wholesale prices (WPOFlt), as indicated by the coefficient value relative to its standard error. Structural-form elasticity and flexibility estimates for retail and wholesale other orange juices are presented in Appendix G. Most of the explanatory variables of consumption of other orange juices in the USA (RQOSt) have elastic responses. Consumption of other orange juices is expected to increase by 3.4629, 1.5593 and 1.1466 percent for each 1 percent increase in income, in price of orange flavored synthetics and drinks,and in retail FCOJ prices, respectively. Retail and other orange juices own-price elasticity is estimated to be -2.0587, indicating that consumption of other orange juices exhibits elastic responses to changes in price (RPOSt). Lagged other orange juices wholesale price (WPOFlt_ 1 ) transmission elasticity (.5561) indicates that RPOSt increase by more than .5 percent for each 1 percent increase in Florida's wholesale prices. The flexibility estimated for the retail proportion of other orange juices in cardboard containers (-.0521) suggest a quite small reduction on RPOSt for each 1 percent increase in the proportion.

PAGE 119

105 The own-price elasticity (WPOFlt) for Florida's movement of other orange juices is estimated to be -.2284. Other orange juices retail transmission elasticity (.5352) indicates that a 1 percent increase in the consumption of this product in the USA (RQOSt) is likely to cause an increase of nearly .5 percent in Florida's wholesale movement of the product. Among the estimated flexibilities for Florida's whole sale prices (WPOFlt) only the flexibility relative to lagged orange production forecast (-.7400) suggests that each 1 percent increase in previous crop forecast is likely to result in a reduction of current Florida's wholesale price of other orange juices by approximately .7 percent. Adjustments in wholesale orange juice prices to changes in crop forecasts by Florida processors are different, depending on the product. The previous discussion indicated that Florida's FCOJ prices are relatively more affected by changes in current orange production forecasts. The results presented for other orange juices, however, indicated that Florida's wholesale prices for this product respond more slowly to changes in the crop forecast. The Canadian Market Canadian structural-form equations for FCOJ exports from Brazil and the USA were estimated in a form almost identical to the conceptual structure developed in Chapter III. Only two variables were deleted from the formulation.

PAGE 120

106 Canadian import prices of FCOJ from other countries (MPFCOct) were excluded from the quantity equations, given extremely high levels of multicolinearity with FCOJ prices from the USA and Brazil. In the USA export price equation to Canada (XPFCSt) import price from other countries (MPFSOct) was also excluded, given its high level of multicolinearity with import prices from Brazil (MPFSBt). Most of the parameters have estimated coefficients with signs consistent with expectations. However, in the USA FCOJ export equation to Canada, the unexpected signs of the estimatd coefficients for Brazil's export prices (XPFCBt) and current Florida's advertising expenditures (ACt) do not appear to create any serious problem for the model, since the estimate relative to XPFCBt is not large relative to its standard error, and the cumulative impact of advertising (current and lagged) is still consistent with expectations. Brazilian FCOJ Exports to Canada Parameter estimates for most of the variables included in the structural-form equations of Brazil's FCOJ exports to Canada are large relative to their respective standard errors (Tables 24 and 25). Brazilian FCOJ exports to Canada (Table 24) were found to be negatively related to Brazil's price (XPFCBt) and positively related to the USA FCOJ and other orange juices export prices to Canada (XPFCSt and XPOCSt). This last relationship shows that, for each 10 cents per gallon increase in the prices of FCOJ and other

PAGE 121

107 Table 24. Structural-form estimates of Brazil FCOJ export demand equation to Canada (XQFCBt' Equation 17) Explanatory 1 variables Intercept Endogenous variables XPFCBt XPFCS~ Predetermined variables + MPFCOct XPOCS~ IC+ t AC+ t + ACt-1 Zlt Z2t Z3t Z4t z5t Coefficient -1390.2600 -405.7944 412.5790 505.7004 134.4298 6.4238 3.0683 -32.2467 7.6037 -7.4476 -62.2866 -8.6692 Standard error 375.9153 253.2894 292.5199 176.0427 58.9665 13.6043 15.1706 67.6270 83.9818 89.0928 63.8325 59.2104 1 variable superscript indicates the a priori expected sign.

PAGE 122

108 Table 25. Structural-form estimates of Brazil FCOJ export price equation to Canada (XPFCBt, Equation 18) Explanatory 1 variables Intercept Endogenous variables XPFCS~ Predetermined variables PdSp~_ 6 RxBCt RxBSt Coefficient -.0124 .7089 -.2872 -.0347 Standard error .4093 .0738 .5156 .0498 1 variable superscript indicates the a priori expected sign.

PAGE 123

109 orange juices exports from the USA to Canada, Brazilian FCOJ exports to the same market will increase by approximately 41 and 51 gallons per thousand persons, respectively. Thus, the Brazilian product seems to be quite sensitive to changes in price policies of the USA products in Canada. The Tilley and Lee (1981) study on Canadian FCOJ imports also emphasized that the USA exerts a strong competitive impact on Brazil's ability to export to Canada. Income changes in Canada were found to affect Brazil's FCOJ exports to the Canadians. However, Florida's advertising programs in Canada appear to have no practical effects on export demand of the Brazilian product, since advertising parameters estimates (ACt and ACt_ 1 ) are quite small relative to their corresponding standard errors. This indicates that Brazilian exports to Canada do not benefit from Florida's FCOJ promotional programs in that market. Seasonal coefficients estimates (Zlt to Z5t) indicate that there is no important seasonal variation. The Brazilian export price of FCOJ to Canada (Table 25) is determined by the USA price to the same market (XPFCSt). For each 10 ~ent change in the USA price per gallon, the price of FCOJ from Brazil increases approximately 7 cents. The estimated coefficients is almost ten times its standard error. These results conform with previous hypotheses that the USA prices have an extremely important role in the derivation of Brazil's FCOJ export price to Canada. Other variables in equation 18 (PdSpt-G and RxBSt) do not appear

PAGE 124

110 to have significant effects on the price determination. Both estimated coefficients are small relative to their estimated standard errors, even though their signs conform with a priori expectation. As anticipated, given the government policy, the exchange rate variable relating cruzeiros per USA dollar (RxBSt) had an estimated coefficient with the right sign, but it was small relative to its standard error. Estimates of structural-form elasticities and flexibilities (Appendix G) summarize the previous discussion of Brazilian FCOJ export structures in Canada. FCOJ exports were found to be responsive to own-price, to the price of alternative products and to income. Each 1 percent change in XPFCB, is associated with more than 1 percent change in the opposite direction in Brazil's FCOJ exports to Canada (-1.1884). Likewise changes in alternative product prices (XPFCSt and XPOCSt) cause a change in Brazilian FCOJ exports to Canada in the same direction by almost 3 percent in the case of the USA FCOJ prices (2.8003) or 4 percent in the case of the USA prices of other orange juices (3.9404). The income elasticity (3.8719) indicates that Brazilian exports of FCOJ to Canada will increase by approximately 4 percent for each 1 percent increase in income per capita in Canada. Florida's advertising elasticities (.1368 and .0659) indicate that Canadian demand for FCOJ from Brazil is relatively insensitive to current and lagged promotional expenditures on Florida's products in Canada.

PAGE 125

111 The export price transmission elasticity (1.6429) relative to FCOJ from the USA gives another measure of the importance of this variable in the derivation of Brazil's export price to Canada. For each 1 percent increase in XPFCSt, the XPFCBt is likely to increase by more than 1.6 percent. The estimated flexibility for lagged orange production in the state of Sao Paulo (-.0475) supports the previous conclusion that crop changes have no major effects on future export prices to Canada. USA FCOJ Exports to Canada Tables 26 and 27 present the structural-form estimates for the USA FCOJ Canadian market equations. The export demand equation (Table 26) has some estimated coefficients that are not greater than their respective standard errors and/or signs that do not conform with a priori expectation. Even though exports are negatively related to export price (XPFCSt) as conceptually formulated the estimated coefficient is small relative to its standard error. Tilley and Lee (1981) also found a negative relation. Products expected to behave as substitutes for FCOJ from the USA in this market appear to have no significant effect on the demand for the USA product. The Brazilian FCOJ price (XPFCBt) coefficient is negative and quite small relative to its standard error. A positive relation exists between USA FCOJ export and XPOCSt, suggesting some degree of substitution even though the coefficient is small relative

PAGE 126

112 Table 26. Structural-form estimates of the USA FCOJ export demand equation to Canada (XQFCSt, Equation 19) Explanatory 1 variables Intercept Endogenous variables XPFCSt XPFCB~ Predetermined variables + MPFCOct XPOCS~ IC+ t AC+ t + ACt-1 Zlt Z2t Z3t Z4t z5t Coefficient -183.3615 -80.0033 -16.3897 65.0389 83.1731 -6.2466 8.9006 -15.6414 8.2555 -36.4766 -15.6500 -8.3631 Standard error 139.4842 108.5401 93.9835 65.3210 21.8796 5.0479 5.6291 25.0932 31.1616 33.0581 23.6852 21.9701 1 variable superscript indicates the a priori expected sign.

PAGE 127

113 Table 27. Structural-form estimates of the USA FCOJ export price equation to Canada (XPFCSt' Equation 20) Explanatory 1 variables Intercept Endogenous variables WPFFl~ MPFSB~ Predetermined variables + MPFSOct RxSCt Coefficient .9912 .4962 .4534 -.0058 Standard error .1297 .0713 .0869 .0011 1 variable superscript indicates the a priori expected sign.

PAGE 128

114 to its standard error. Most of USA exports to Canada are in consumer size packages, while practically all FCOJ exports from Brazil are in 55 gallon drums of 65 brix concentrate that require further reprocessing. The different characteristics of FCOJ from these two sources may explain the results of this study. The USA product is quite differentianted so that its market is relatively independent of potential substitutes. However, FCOJ from Brazil is quite sensitive to changes in the price of the USA products (FCOJ and other orange juices) as discussed previously. Brazil exerts little, if any, competitive impact on the USA ability to export FCOJ to Canada. The USA FCOJ exports to Canada are quite sensitive to income per capita; i.e., one dollar increase in income is likely to increase imports from the USA by more than 83 gallons per thousand persons. Florida's orange juice advertising efforts in Canada do not seem to have been successful. The coefficient for current advertising (ACt) is negative but smaller than the positive coefficient for prior period advertising. Tilley and Lee (1981) found advertising results that were similar. The USA product does not exhibit a seasonal pattern. All variables included in the final estimates of the USA FCOJ export price equation have coefficient that conform with a priori expectations and are large relative to their respective standard errors. The Brazilian product imported into the USA appears to have an important role in deriving

PAGE 129

115 XPFCSt. Its impact on price is almost the same as the impact of Florida's wholesale prices. Each 1 cent change in one of these prices will change the USA FCOJ export price to Canada (XPFCSt) by practically .5 cent. This finding is quite important since it gives support to the hypothesis that Brazilian FCOJ imported by the USA is a major factor in the export program. Because duty drawback is available on exports in the USA, imports can be used to average down FCOJ exports prices and/or as a way to make FCOJ imports profitable. The estimates of the coefficient for the USA exchange rate relative to the Canadian dollar indicate that, for each cent increase in that rate the export price is likely to decrease by .6 cent. In the data period used in this study the exchange rate has decreased, especially in the recent three years (Table D.3. Appendix D). Thus, the expected result was an increase in the export price to Canada and consequently a decrease in FCOJ exports, which is consistent with observed exports. This result supports the role of exchange rate in the derivation of export prices stressed by Schuh (1974) and Kost (1976), not only qualitatively, but also in terms of magnitude. In the case of FCOJ exports to Canada, variation in the rate of exchange between the two currencies has had substantial effects on USA export prices to that market. Except in the case of the income elasticity, all other structural-form el~sticity estimatesfor the USA FCOJ export

PAGE 130

116 structure in Canada are inelastic (Appendix G). Even though relatively less responsive than in the case of the Brazilian product, USA FCOJ exports to Canada are expected to increase more than 2 percent for each 1 percent increase in income. The own-price elasticity relative to FCOJ exports to Canada (-.4616) indicates that quantity sold in Canada is less responsive than Florida's wholesale domestic movement (own-price elasticity equal to -.6618), and of Brazilian FCOJ export price to Canada (-1.1884). In the price equation all price transmission elasticities are quite iMportant. For each 1 percent increase in Florida's wholesale price (WPFFlt) the USA export price to Canada (XPFCSt) is expected to increase by almost .4 percent (.4465). Likewise change in the USA import price of FCOJ from Brazil is expected to increase XPFCSt by nearly .2 percent (.1857). On a relative basis, the impact of Brazilian FCOJ price to the USA in the determination of XPFCSt is still almost half of the impact of Florida's wholesale prices. The estimate of exchange rate transmission elasticity (-.5287) indicates that 1 percent increase in the rate of the USA dollar per Canadian dollar is associated to a price decline for FCOJ exports from the USA to Canada by more than .5 percent. Own-price elasticity estimate in this study referent to USA FCOJ export to Canada (-.4616) is not far away from that of Tilley and Lee (-.64), considering the differences in both studies. However, cross-price elasticities relative

PAGE 131

117 to price of FCOJ from Brazil, are quite different. These different results are likely to be associated with the ways that the USA FCOJ exports to Canada are treated. While Tilley and Lee differentiated between FCOJ retail size and bulk shipments, in the present study they are not treated differently. The European Markets Structural-form equations for FCOJ exports from Brazil and the USA to the two groups of European countries (the EEC7 and the European non-EEC3) were estimated as suggested by their respective conceptual formulation, except for two minor changes in the price equations. High levels of multicolinearity were found within the exchange rate indices (RxBEt, RxBNt' RxSEt' and RxSNt). Since the USA dollar is the basic currency in FCOJ trade, the exchange rates of cruzeiros per USA dollar were used in the Brazilian equations. Import prices of FCOJ to the USA from other countries (MPFSOct) were deleted from the USA export price equations for the European markets since this variable was colinear with import prices from Brazil (MPFSBt). With few exceptions, most of the coefficient estimates have signs that conform to a priori expectations. Substitution relationships are difficult to verify in either the USA or Brazilian export demand relationships. These results seem to reinforce the idea that differentiation of FCOJ from the USA has resulted in export flows to European markets that are

PAGE 132

118 quite independent of price changes by potential competitors. The other unexpected sign is the negative effect of changing values of the indices of the USA exchange rate to the EEC7 country currencies on the USA FCOJ export price to these countries. Since the index is an inverse function of individual country exchange rates, changes in the index should have an effect on export prices in an opposite direction from that of changing individual exchange rates. Individually, an exchange rate increase (devaluation) is expected to decrease export prices. So the index of exchange rates should be positively related to export price if all exchange rates included in the index are simultaneously devalued. These and other specific results relative to the estimates of the Brazilian and the USA FCOJ export equations for these two markets, will be the major topics in the following discussion. Brazilian FCOJ Exports to the EEC7 Brazilian exports to the EEC7 countries are found to be negatively related to own-price (XPFEBt) and positively associated with time-related changes (Tt) on exports (mainly income) and to Florida's product advertising expenditures in those countries (AEt) (Table 28). These results show that XQFEBt is sensitive to changes on its own-price and, that, over time, there is evidence to indicate an increasing trend in Brazil's exports of FCOJ to this market, even though in absolute terms it is small. Three-Party Program efforts intended to promote Florida's orange products in this region (AEt) seem to increase Brazilian FCOJ sales as

PAGE 133

119 Table 28. Structural-form estimates of Brazil FCOJ export demand equation to the EEC7 countries (XQFEBt' Equation 21) Explanatory 1 Coefficient Standard variables error Intercept 87.1555 47.1473 Endogenous variables XPFEBt -147.8751 80.7974 XPFES~ -9.2872 34.8037 Predetermined variables T+ t 3.1557 .9424 AE+ t 17.9493 11. 9529 Zlt -4.8740 16.1997 Z2t -51. 7059 16.5576 Z3t -81. 4971 16.3508 Z4t -68.0360 16.2112 z5t -23.2469 16.2125 l bl Varia e superscript indicates the a priori expected sign.

PAGE 134

120 well. Estimates for the seasonal variables (Z2t to Z5t) indicate that, in relation to the October-November bimonth, Brazil's FCOJ exports to the EEC7 countries decrease substantially until the April-May bimonth and then continue to decrease by small quantities until recovering to the base period level of exports in October-November. Practically all estimates of the export price equation coefficients (Table 29) for FCOJ from Brazil (XPFEBt) are large relative to their respective standard errors. The USA price to the same market (XPFESt) is an important determinant of Brazil's FCOJ export price to the EEC7. The coefficient for XPFESt is almost seven times greater than its standard error. Its numerical value indicates that a 10 cent increase in the USA price translates into almost a 3.5 cent increase Brazilian FCOJ price to this market. Devaluation of the cruzeiro relative to the USA dollar by one unit is likely to reduce XPFEBt by 13 cents per gallon. Thus, Brazilian FCOJ exports to the EEC7 countries would be expected to increase considerably with devaluations of the cruzeiro relative to the USA dollar. Most of the previous results are reexpressed by elasticity estimates (Appendix G). Each 1 percent increase in current Florida's advertising expenditures (AEt) through the Three-Party Program represents more than .5 percent increase in Brazil's exports of FCOJ to the EEC7 market (.5292). Likewise, change in the USA export price of FCOJ to this market (XPFESt) is expected to increase Brazilian export price of FCOJ to the EEC7 countries by almost .6

PAGE 135

121 Table 29. Structural-form estimates of Brazil FCOJ export price equation to the EEC7 countries (XPFEBt, Equation 22) Explanatory 1 variables Intercept Endogenous variables XPFESt Predetermined variables PdSp~-G + RxBEt RxBSt Coefficient 1.2336 .3473 -.5505 -.1252 Standard error .4033 .0507 .5404 .0512 1 variable superscript indicates the a priori expected sign.

PAGE 136

122 percent (.5610). Exchange rate transmission elasticity (-2.0659) is quite high, suggesting a strong response of the Brazilian export price (XPFEBt) to changes in the exchange rate of cruzeiros per USA dollar. USA FCOJ Exports to the EEC7 Tables 30 and 31 show the estimates of the structural form parameters of the USA FCOJ export structure in the EEC7 countries. Most of the major variables have estimated coefficients that are large relative to their standard errors. In the export demand equation (Table 30) the coefficient for the price of FCOJ (XPFESt) indicates that FCOJ exports to this market from the USA are quite sensitive to changes in XPFESt. For each 10 cent per gallon increase in XPFESt, FCOJ exports decrease by more than 2.4 gallons per thousand persons. The estimated coefficient for Brazilian FCOJ price (XPFEBt) does not support the hypothesis of substitution for the USA product. The estimated relation has a negative sign but is small relative to its standard error. This result means that the USA exports of FCOJ to the EEC7 countries are independent of changes in the Brazilian FCOJ prices in the same market. A similar conclusion was obtained in the case of Brazil's FCOJ exports to the EEC7 countries relative to changes in the USA prices of FCOJ. These results are consistent with the fact that the Brazilian product is exported in bulk drums of 65 brix concentrate that requires reprocessing while FCOJ from the USA is already in consumer packages that are ready for

PAGE 137

123 Table 30. Structural-form estimates of the USA FCOJ export demand equation to the EEC7 countries (XQFESt' Equation 23) Explanatory 1 variables Intercept Endogenous variables XPFESt XPFEB~ Predetermined variables T+ t AE+ t Zlt Z2t Z3t Z4t z5t Coefficient 11. 3701 -24.0565 -1. 9782 .3514 1.8695 -.0536 3.8950 6.6364 7.0104 1.6055 Standard error 7.563 5.5861 12.9682 .1513 1.9185 2.6001 2.6575 2.6244 2.6019 2.6022 1 variables superscript indicates the a priori expected sign.

PAGE 138

124 Table 31. Structural-form estimates of the USA FCOJ export price equation to the EEC7 countries (XPFESt, Equation 24) Explanatory 1 variables Intercept Endogenous variables WPFFl~ MPFSB~ Predetermined variables Coefficient 1.2518 .2540 .7380 -.0109 Standard error .4838 .2293 .2791 .0040 1 variable superscript indicates the a priori expected sign.

PAGE 139

125 immediate consumption. The fact that the Brazilian product has a quite constant high share in the market (Table 9, Chapter II) and is valued substantially below the USA product (Table 8, Chapter II) may explain the small sensitivity of Brazil's FCOJ exports to changes in the USA FCOJ prices. The USA FCOJ exports to the EEC7 countries are also growing over time as reflected by the coefficient for Tt' the time trend. The Three-Party Program in promoting Florida's product (AEt) seems to have no impact on the USA FCOJ exports to this market. Specific studies on the effectiveness of the Three-Party Program (Lee, 1977,1978; and Lee et al. 1978) have reached quite favorable conclusions for the program. Seasonal variable estimates for February-March through June-July (Z2t through Z4t) indicate that exports to the EEC7 market increase significantly from the base period (October-November). The peak of the USA exports to the EEC7 coincide with the smallest rate of Brazilian exports of FCOJ to the same market. These results suggest some seasonal complementarity between the two sources in supplying FCOJ to the EEC7 countries which may explain the lack of substitution between the two products. Estimates of the price equation parameters (Table 31) indicate that the USA import price of FCOJ from Brazil (MPFSBt) plays a major role in the derivation of the export price of FCOJ from the USA to the EEC7 market. For each 10 cent per gallon increase in MPFSBt' XPFESt is

PAGE 140

126 expected to increase XPFESt by 7.4 cents. An increase in Florida's wholesale price (WPFFlt) is estimated to result in an increase of only 2.5 cents. Similar estimates on the Canadian market (Table 27) give these variables almost the same impact on the USA export prices to Canada, with Florida's wholesale price having relatively more impact. Together, these results suggest that the Brazilian prices to the USA market are relatively more important in deriving the USA export price to the EEC7 countries than to Canada, with Florida's wholesale price having relatively more impact. This inference is quite consistent, given that it should be much easier to maintain the EEC7 export price lower relative to the USA domestic price than to offer similar discounts in Canada. Thus, a two-price system (domestic prices higher than export prices), using imports from Brazil to average down the export price, is more likely to be associated with USA exports to the EEC7 countries than exports to Canada. As already mentioned in the introductory discussion of this sub-section, the negative sign of the USA exchange rate index (RxSEt) was not expected. Given the number of curren cies (seven) involved in the derivation of the index, and that the USA dollar has not had a systematic pattern of behavior 7 7 A devaluation of the USA dollar to the deutsche mark (West Germany) does not necessarily imply a devaluation of the dollar relative to the franc in France or to the pound in the United Kingdom.

PAGE 141

127 toward all of them during the period of time considered in this study, then the index estimate turns out to be an average measure of devaluations and increased valuations at the same time. Thus, the results should not necessarily conform with the conceptual expectation. Elasticity estimates derived from the structural-form equations (Appendix G) reinforce most of the previous inferences regarding the USA FCOJ export structure to the EEC7 market. Own-price elasticity results indicated that compared with the USA domestic and the Canadian market, exports to the EEC7 countries tend to be relatively much more responsive to price changes. A 1 percent cut in FCOJ price is likely to increase Florida's domestic movement by approximately .7 percent and exports to Canada by nearly .5 percent. However, the same changes in the export prices to the EEC7 market are expected to increase exports to this region by almost 2 percent. Similar relations among these elasticities were found by Ward (1976a) relative to the FCOJ from Florida. These results then support Ward's conclusion that discounting the exporting price to the EEC7 market is desirable. Elasticity estimate of current expenditures of the Three-Party Program indicates that 1 percent increase in AEt will increase the USA FCOJ exports to the EEC7 countries by approximately .5 percent. The umbrella effect on Brazilian exports of FCOJ to the same market from this advertising effort to promote Florida's product is slightly more than the

PAGE 142

128 corresponding effect on FCOJ exports from the USA as measured by the respective advertising elasticities (.5292 and .5029). Price transmission elasticity reinforces the idea that the USA import price from Brazil (MPFSBt) plays a major role in the derivation of the USA export price to the EEC? countries (XPFESt). The impact on XPFESt for each 1 percent change in MPFSBt is relatively higher (.4362) than the impact on XPFESt for the same change in Florida's wholesale price (.3298). However, the effect on XPFESt from changing the values of the dollar relative to the EEC? country currencies (RxESt), as measured by the exchange rate transmission elasticity (-1.3934), is more than sufficient to offset the combined effect of changing MPFSBt and WPFFlt. These results may explain the recent downward trend of the USA FCOJ exports to the EEC? market. The strengthening of the dollar relative to major currencies in Europe has offset any effort by the same agents to increase exports to this region through price reductions and/or promotional programs. Brazilian FCOJ Exports to the European Non-EEC3 Structural-form equation estimates relative to Brazil's FCOJ exports to the European non-EEC3 countries are shown in Tables 32 and 33. Most of the estimated coefficients for export demand (Table 32) are not substantially greater than their associated standard errors. Based on these results it appears that Brazilian FCOJ exporting to this market is largely a seasonal phenomenon as identified by the seasonal

PAGE 143

129 Table 32. Structural-form estimates of Brazil FCOJ export demand equation to the European non-EEC3 countries (XQFNBt' Equation 25) Explanatory 1 variables Intercept Endogenous variables XPFNBt XPFNS~ Predetermined variables T+ t AN+ t Zlt Z2t Z3t Z4t zst Coefficient 510.9708 -182.6938 -26.6495 .5505 -5.9842 133.7020 58.9341 12.6402 53.3214 224.2376 Standard error 531.7365 281.7113 199.7314 8.3220 6.9020 62.0089 62.5844 63.0043 62.6468 64.0905 1 variable superscript indicates the a priori expected sign.

PAGE 144

130 Table 33. Structural-form estimates of Brazil FCOJ export price equation to the European non-EEC3 countries (XPFNBt, Equation 26) Explanatory 1 variables Intercept Endogenous variables XPFNS: Predetermined variables PdSp~_ 6 + RxBNt RxBSt Coefficient .7131 .4177 .4009 -.0829 Standard error .4231 .0951 .5638 .0517 1 variable superscript indicates the a priori expected sign.

PAGE 145

131 variables (Zlt to ZSt). Most of the estimates have signs consistent with expectations. FCOJ exports are negatively related to own-price (XPFNBt} and, again, the results do not support the assumption that the USA product is a potential substitute to FCOJ from Brazil. Time-related changes, as measured by the coefficient of Tt' seem to have no significant effect on XQFNBt. The effect of Three-Party Program promotion of Florida's FCOJ on Brazil's exports to the European non-EEC3 countries conforms with expectation. Seasonal variability estimates seem to indicate that most of the Brazilian exports of FCOJ to these countries are seasonally related, especially during the bimonths of August-September through December-January when Brazil's sales to this market increase considerably. These results may suggest that Brazil's exports of FCOJ to these countries are also complementary to other sources. The price equation estimates (Table 33) are quite similar to effects of the same variables in the derivation of Brazil's FCOJ export price to the Canadian and the EEC? markets. The results indicate that a quite high proportion of variations in Brazil's FCOJ export price to this market (XPFNBt) can be explained by changes in the USA export price of FCOJ to the same market (XPFNSt} and by the exchange rate of the cruzeiro against the USA dollar (RxBSt) For each 10 cents per gallon that is added to XPFNSt' the Brazilian product price (XPFNBt) is expected to increase by 4.2 cents per gallon.

PAGE 146

132 Elasticity and flexibility estimates relative to structural-form equations (Appendix G) reinforce the previous related discussion, even though most of the export demand estimates have estimated coefficients smaller than the respective standard errors. However, two of the price equation estimates are relatively important to the analysis. The USA FCOJ price transmission elasticities relative to the derivation of Brazil's FCOJ export price indicate that the estimate for this market (.7969 percent) is the second largest response of Brazilian export prices to variation of the respective export price from the USA. As indicated by the estimated exchange rate transmission elasticity (-1.3911), for each 1 percent devaluation of the Brazilian cruzeiro relative to the USA dollar more than a 1 percent reduction in XPFNBt is expected. USA FCOJ Exports to the European Non-EEC3 Tables 34 and 35 display the structural-form estimates for the USA FCOJ export equations to the European non-EEC3 countries. In the export demand equation (Table 34), export quantity of FCOJ to this region from the USA (XQFNSt) is quite sensitive to variation in its own-price (XPFNSt). For each 10 cent per gallon reduction in XPFNSt, USA exports to this market are expected to increase by more 23.2 gallons per thousand persons. Estimate of the effect of Brazilian FCOJ price to the same market (XPFNBt) indicates that the Brazilian product behaves like a substitute for FCOJ from

PAGE 147

133 Table 34. Structural-form estimates of the USA FCOJ export demand equation to the European non-EEC countries (XQFNSt, Equation 27) Explanatory 1 variables Intercept Endogenous variables XPFNSt XPFNB: Predetermined variables T+ t AN+ t Zlt Z2t Z3t Z4t z5t Coefficient 111.9906 -232.2003 67.3028 1.4686 2.5567 38.0835 4.0872 16.4419 -1. 3898 -.3866 Standard error 133.7687 50.2464 70.8700 2.0935 1. 7363 15.5995 15.7443 15.8500 15.7600 16.1232 1 variable superscript indicates the a priori expected sign.

PAGE 148

134 Table 35. Structural-form estimates of the USA FCOJ export price equation to the European non-EEC3 countries (XPFNSt, Equation 28) Explanatory 1 Coefficient Standard variables error Intercept -.3826 .3273 Endogenous variables WPFFl~ .3890 .1452 MPFSB~ .2225 .1763 Predetermined variables + MPFSOct + RxSNt .0080 .0030 1 variable superscript indicates the a priori expected sign.

PAGE 149

135 USA even though the parameter is small relative to its standard error. The time variable (Tt) coefficient estimate does not give strong support that XQFNSt is increasing over time, which is consistent with the characteristics of the market previously discussed. Even though the estimated parameter is not large relative to its standard errors, the Three-Party Program parameter estimate indicates that each dollar increase in the program expenditures is likely to increase FCOJ exports from the USA by more than two and a half gallons per thousand persons. All price equation estimates (Table 35) are consistent with a priori expectations and the nature of this market concerning orange juice consumption patterns. Estimates of the USA import price of FCOJ from Brazil (MPFSBt) do not strongly indicate that the Brazilian product plays the same role in the derivation of FCOJ export prices from the USA as it does for other markets. To these countries, the major role is played by Florida's wholesale price (WPFFlt). The estimated coefficient of the exchange rate index (RxSNt) is almost three times its standard error (.0030), even though it is relatively small in absolute terms (.0080). Each one unit reduction in the index, which is likely to be associated with devaluation of the dollar relative to all individual currencies, is expected to reduce the XPFNSt by less than 1 cent per gallon. Structural-form elasticity estimates (Appendix G) reinforce the previous inferences on the USA FCOJ exports

PAGE 150

136 to the European non-EEC3 countries. Export quantities were found to be quite responsive to changes in the own-price of FCOJ from the USA (XPFNSt), as measured by the own-price elasticity of demand for exports (-2.4662). Promotion of Florida's FCOJ through the Three-Party Program was found to be relatively effective, since each 1 percent increase in the program expenditures is expected to be associated with more than 1 percent increase in FCOJ exports from the USA to this market. In relative terms this is the highest response of FCOJ sales to advertising efforts. This better response to advertising could be associated with the weight placed on orange juice quality in these countries and the ability of the USA product to fulfill these requirements since most of the FCOJ exports from the USA are in final consumer containers where quality is more easily controlled. Price transmission elasticities are very consistent with previous discussions relative to the implications of the price equation estimated coefficients. For each 1 percent increase in the price per gallon of Florida's wholesale FCOJ prices (WPFFlt) or of the USA import price of FCOJ from Brazil (MPFSBt), XPFNSt is expected to increase by .4 and .1 percent, respectively. These results show that the Brazilian product does not have a significant role in the derivation of the XPFNSt as compared with the role of Florida's FCOJ price. The exchange rate price transmission elasticity estimate (.8779) is quite large. The relative impact on XPFNSt of changing exchange rates (.8779) is more

PAGE 151

137 than twice the effect of variations in the Florida wholesale price (.4349. Estimates of the Derived Reduced-Form Equations This section contains a discussion of the derivation of the reduced-form estimates of the model, and of major implications of these estimates for the orange juice trade; also included are the results of model validation tests which give some insight into the performance of the estimated model. The reduced-form estimates are a restricted reduced-form version of the system of simultaneous equations (Equation 31, Chapter III). The matrix of the derived reduced-form estimates of predetermined variables (II) was obtained by solving the following equation (37) where Band rare matrices of the structural-form coefficients of endogenous and predetermined variables estimated by the 2SLS procedure. Equation 37 is simply the estimated version of Equation 33 in Chapter III. In order to carry out the matrix operation in equation 37, nonlinear relations in the structural formulation of the model (Chapter III), defined by the identities for orange juice stocks (Equations 5, 6, 15, and 16) and for retail proportion of other orange juices in cardboard containers (Equation 12), were linearized as follows. Population means were used in the population ratios of Equat1ons 6

PAGE 152

138 and 16 in the definitions of wholesale quantity per thousand persons of FCOJ and of other orange juices in stock in the USA (WQkFSt and WQkOSt). The other nonlinear relations involving ratios of major varialbes of the model (Equations 5, 12 and 15) were linearly approximated by first-order terms of the Taylor's Series expansion about the variable means (Womack & Matthews, 1972). As summarized by them, the 8 other procedures used to handle this problem require initial values for all predetermined variables making them complex and quite expensive for large models. Taylor's Series expansions have been shown to be very convenient (Womack and Matthews). Thus, the identities defined by Equations 5, 12 and 15, in Chapter III, were approximated by the following linear expressions 9 evaluated about the means of the respective variables. WSkFSt = 2.63843 + .00185 WQkFSt .00489 (WQFFlt (38) RpOSt = .49691 + .00738 RQObSt .00367 RQOSt (39) WSkOSt = 1.14481 + .00678 WQkOSt .0076 WQOFlt (40) With these transformations the reduced-form estimates were derived through the matrix operation in Equation 37. 8 Newtonian numerical analytic techniques and Gauss-Siedel numerical method. 9 The linear approximation of an x/y ratio by the first-order terms of the Taylor's Series expansion, evaluated about the means of the variables, can be simplified as follows: A x/y = x/y + (1/y)x (x/y 2 )y where, x and y are mean values of the respective variables.

PAGE 153

139 Table 36 displays the derived reduced-form estimates which show the effect of predetermined variables on the endogenous variables in the model. These estimates are called long-run multipliers (Kennedy, 1979, p. 107) or simply multipliers (Goldberger 1964, p. 369), since a particular reduced-form coefficient indicates the total effect on the corresponding endogenous variable of a one unit change in its respective predetermined variable, assuming that other exogenous variables in the equation are fixed. It is a total effect in the sense that it accounts for the interdependent relationships among other current endogenous variables. The discussion in this section covers the impact of a selected group of multipliers on some endogenous variables. The first sub-group includes lagged wholesale prices of orange juices in Florida (WPFFlt-l and WPOF1t_ 1 ), export price of other orange juices to Canada from the USA (XPOCSt), retail price of frozen orange-flavored synthetics and drinks in the USA (RPShSt), and income in the USA and in Canada (ISt and ICt). The second sub-group includes exchange rates, orange production, and freeze. Except for freeze, discussion of which is based on the magnitude of the respective estimates in Table 36, the multiplier impact relative to all other predetermined variables is discussed using the derived reduced-form elasticity and flexibility estimates (Table 37) evaluated about the means of the respective variables. The results relative to the last sub-group of predetermined variables will be discussed as separate topics in this sub-section.

PAGE 154

Table 36. Estimates of the derived reduced-form equations 1 Endogenous Predetermined Variables Intercept Variables WPFFlt 1 WOkFSt_ 1 WSkFSt_ 1 WPOFlt 1 WQkOSt_ 1 MOFSOct XOFOcSt xaocst XPOCSt WQsFFlt WQsOFlt ROFSt 110 9390 28.5133 0022 2 3459 -82.4137 0023 0022 -.0020 .0002 .0151 .0022 0023 RPFSt 2 3035 2657 .0000 -.0219 -.0134 0000 0000 0000 .0000 0001 .0000 0000 WOFFlt -615.2930 18 4420 0074 7 9742 -195.7170 -. 0287 0074 -. 0068 0031 0512 0074 -.0287 WPFFlt 2 2562 0023 .0000 0267 0163 .0000 .0000 .0000 .0000 .0002 0000 0000 MOFSBt -159.0500 32.8080 -.0033 -3.5814 -73.5118 0056 -.0033 0031 -.0006 .0230 -.0033 0056 MPFSBt .4203 .4642 0000 -.0062 -.0038 0000 .0000 .0000 0000 0000 .0000 .0000 ROOSt 299 9130 24.1034 -.0018 -1.9831 103 3150 0023 .0018 .0017 -.0002 0127 -.0018 .0023 RPOSt .7858 .0204 0000 -. 0017 .7460 0000 0000 0000 0000 .0000 0000 .0000 WOOFlt -139.3410 13.9491 -.0011 1 1477 -59.7905 0048 0011 0010 .0005 0074 0011 .0048 WPOFl 1 2.2660 0042 .0000 0003 .0182 .0001 0000 .0000 0000 0000 0000 .0001 XQFCB 1 1097.7700 26 4321 -.0019 -2.0039 -1.2269 -. 0002 -.0019 .0017 0000 505 7130 -.0019 -.0002 XPFCB 1 1.6190 .1500 0000 .0114 0070 0000 0000 0000 0000 .0001 0000 .0000 1--' ,i::,. XOFCSt -394.0080 19.3879 0014 1.4698 .8999 .0002 0014 0013 0000 65 0295 0014 0002 0 XPFCSt 2.3013 .2116 0000 -.0160 -. 0098 0000 .0000 0000 0000 .0001 .0000 .0000 XOFEBt 224.7440 20.8117 .0006 6887 4217 0001 0006 -. 0006 0000 -.0044 0006 0001 XPFEB 1 1 9751 1192 0000 -.0039 -.0024 .0000 .0000 0000 0000 .0000 0000 .0000 XOFESt 36 0853 --8.0199 0002 .2654 .1625 0000 0002 -.0002 0000 -.0017 .0 002 0000 XPFES 1 2 1351 3432 0000 -. 0114 -.0070 0000 .0000 0000 0000 0000 0000 0000 XOFNBt 320.0910 -10.7258 0011 1 2098 7407 .0001 0011 0010 0000 0078 0011 0001 XPFNB 1 9589 0436 .0000 -.0049 0030 .0000 0000 0000 0000 0000 0000 0000 XQFNSt 39 8620 21.2605 0022 2 3981 1.4683 .0003 0022 -. 0021 0000 0154 0022 0003 XPFNS 1 5886 .1042 0000 -. 0118 -.0072 0000 .0000 0000 0000 .0001 0000 .0000 WOkFS 1 528.2810 -40 1080 9887 -12.1463 121 8440 0343 9887 9125 -. 0036 -6 8357 9887 0343 SpB 1 1 8358 -.4620 0000 -.0204 -. 0125 0000 0000 0000 0000 0001 .0000 .0000 WOkOS 1 139 3410 13 9491 0011 1 1477 59 7905 9952 0011 0010 -. 1058 -.0074 0011 9952 RpOS 1 1 5976 -. 0885 0000 .0073 3792 0000 0000 .0000 0000 0000 .0000 0000 WSkFSt 6 6245 1~4 .0018 0615 1 1825 .0002 0018 -. 0017 0000 -. 0124 0018 0002 WSkOS 1 3.1708 .2 028 0000 0167 8694 .0067 0000 0000 -. 0001 0001 0000 0067 1 1 n order to derive these estimates the identity equations ( ratios) for WSkFSt (Equation 5), RpOSt ( Equation (continued) 12) and WSkOSt (Equation 15) were linearly approximated by Taylor's Series expansion evaluated about the means of the respective variables.

PAGE 155

Table 36 (continued). Endogenous Predetermined Variables Variables RPShSt ROObSt 1s 1 IC 1 AS 1 ASt-1 ACt ACt-1 AE 1 AN 1 PdSt PdSt_ 1 ROFS 1 115 1170 .1 635 53 3599 -. 0193 1.3682 8584 0014 0021 -. 0038 0004 292 7250 231.8290 RPFS 1 0171 0000 .0101 .0002 0001 .0001 0000 0000 0000 .0000 -2.7701 -1.5656 WOFFl 1 248 7860 .3883 147 7760 -.0655 1.1648 .9833 .0049 -.0070 0128 0015 1076 9800 -350.3990 WPFF1 1 0208 .0000 0123 0002 .0001 0001 .0000 .0000 0000 .0000 -3.3778 1.9090 MOFSB t 90 6325 .1458 57 9126 0294 1992 .2487 0022 .0031 0057 0007 -467.2940 -70.5299 MPFSB 1 0049 0000 0029 .0000 0000 0000 .0000 .0000 0000 .0000 -.7877 -.4452 ROOS 1 127 2950 .2050 81 4618 0163 .2730 3460 -.0012 .0017 0032 0004 -256.9260 -64 3643 RPOS 1 1079 -. 0015 0691 0000 .0002 0003 .0000 0000 .0000 .0000 .1 722 -.6888 WOOFl 1 73 6680 1186 47 1436 .0094 1580 .2002 .0007 .0010 .0018 0002 -156 5450 71.8977 WPOFl 1 0224 .0000 0143 .0000 0000 .0001 .0000 0000 0000 0000 .2703 -4 3943 XOFCBt 1 5650 .0024 .9218 134 4460 .0078 0064 6.4224 3 0701 .0032 .0004 -253.9700 -143 5350 I-' ti::. XPFCBt .0089 0000 0053 .0001 0000 .0000 0000 0000 0000 .0000 -1.4413 -.8146 I-' XOFCSt -1 1479 -.0018 -.6761 83 1610 -. 0057 -.0047 -6.2457 8.8993 -.0024 -.0003 186.2860 105.2820 XPFCS 1 0125 0000 .0074 .0001 .0001 .0000 .0000 .0000 .0000 0000 -2.0332 -1.1491 XQFEB 1 5379 .0008 -.3168 -.0057 0027 -.0022 .0004 -.0006 17 9482 -.0001 87 2841 49.3300 XPFEB 1 0031 .0000 0018 0000 .0000 .0000 0000 .0000 .0000 0000 -.4999 2825 XOFES 1 -.2073 -.0003 -.1221 -. 0022 .0010 -.0008 .0002 -.0002 1.8691 0000 33.6352 19.0095 XPF'ESt 0089 0000 0052 0000 .0000 .0000 .0000 0000 0000 0000 -1.4393 -.8134 XOFNB 1 -.9449 -. 0015 -.5565 .0099 -.0047 -.0039 0007 .0011 -.0019 -5.9844 153 3311 86 6577 XPFNB 1 0038 .0000 0023 .0000 0000 0000 .0000 .0000 0000 .0000 6220 .3516 XOFNS 1 -1.8729 -.0029 -1.1031 .0197 0093 -.0077 0015 -.0021 0038 2.5562 303 9320 171.7720 XPFNSt .0092 .0000 .0054 0001 .0000 .0000 .0000 .0000 .0000 .0000 -1.4892 .8417 WOkFS 1 -157.6920 .2417 -89 5921 -8.7416 -.9633 .7327 6565 -.9355 -1.7059 -.1997 -1619.1200 237 5640 SpBt 0160 .0000 .0094 0002 .0001 .0001 .0000 0000 0000 .0000 -2 6900 -1.4638 WOkOSt -73.6680 -.1186 -47.1436 -. 0094 -.1580 -.2002 .0007 -.0010 -.0018 -.0002 156.5450 -7113977 RpOSt -. 4672 .0066 .2990 -.0001 -.0010 .0013 0000 0000 0000 .0000 9429 .2362 WSkFS 1 -1. 5083 -. 0023 -. 8884 -.0159 .0075 .0062 .0012 .0017 -. 0031 0004 -8.2618 2.1529 WSkOS 1 -1.0711 -.0017 -. 6855 -.0001 -. 0023 -.0029 .0000 0000 .0000 .0000 2.2762 -1.0454 (continued)

PAGE 156

Table 36 (continued) Endogenous Predetermined Variables Variables PdSpt-6 RxBSt RxSCt RxSEt RxSNt Dt Tt 21t 22 1 23t 24t 25t ROFS 1 1365 .0031 .0001 -. 0005 .0003 -29.6390 .5163 17 1773 14.0193 9 4302 20 6636 11.6209 RPFS 1 -.0013 0000 0000 0000 0000 2336 -.0041 0045 .0036 0016 0008 0025 WOFFlt .4640 0104 0004 .0017 0010 -19.2111 .3113 66.9426 45 9514 -26.6951 4.5090 30.9170 WPFFlt -.0016 .0000 0000 .0000 0000 .2848 -.0049 0055 .0044 -.0019 .0009 .0030 MOFSB 1 82 5074 1.0163 .0002 .0008 0004 24 9513 -.4288 29.4274 -.2683 -18 6441 -15 2616 -6.1674 MPFSB 1 -. 7701 -. 0095 .0000 .0000 0000 .0664 -.0012 .0013 .0010 -.0004 .0002 0007 ROOS 1 -.1154 -.0026 0001 .0004 0002 15 6287 2695 -1 6045 -2.3224 7 1170 -1.5098 3.7593 RPOS 1 0001 .0000 0000 0000 0000 .0587 -.0010 0007 -. 0013 -.0057 0007 0037 WOOFl 1 -.0668 -.0015 0001 .0003 0001 1 2273 -.0176 14 2919 13.0192 2 5718 12.0962 14.2487 WPOF1 1 .0000 0000 .0000 .0000 0000 .3167 I-' -.0056 0043 .0040 .0008 .0037 .0043 N XOFCB 1 72 8357 13.5404 7244 .0004 -. 0002 21.4147 -.3719 -31.8328 7.9346 -7.5916 -62.2168 -8 4421 XPFCB 1 -.5353 0378 0041 .0000 .0000 .1215 -.0021 .0023 .0019 .0008 .0004 .0013 XOFCS 1 36 7671 .9653 .5313 -. 0003 .0002 -15.7076 2728 -15 9450 8.0127 -36.3710 -15.7012 -8.5297 XPFCSt -.3499 -.0043 0058 .0000 .0000 .1714 -.0030 .0033 .0026 -.0012 .0006 .0018 XOFEB 1 115.8940 18 9400 0000 .6609 .0001 -7 3598 3.2835 -5.0163 -61.8196 -81.4476 -68 0600 -23.3249 XPFEB 1 -. 7480 -.1276 0000 -.0038 .0000 .0421 .0007 .0008 0007 -.0003 .0001 .0004 XOFES 1 12.2012 -.0835 .0000 .2547 0000 -2.8361 .4007 -.1084 3.8512 6.6555 7.0012 1 5754 XPFES 1 -. 5687 -.0070 .0000 -.0109 .0000 .1214 -.0021 .0023 .0019 -.0008 .0004 .0013 XOFNB 1 -55.5387 15.3645 -. 0001 .0003 -.8235 -12.9289 .7750 133.4520 58 7343 12.7271 52 2793 224 1000 XPFNB 1 3291 -.0838 0000 0000 .0033 .0525 -.0009 .0010 0008 -.0004 .0002 .0006 XOFNS 1 62.0730 -6.1449 0001 -.0005 -1.6324 -25.6276 1 9136 37.5881 3 6912 16 6142 -1 4733 -.6584 XPFNS 1 -.1719 -.0021 .0000 .0000 .0080 .1256 0022 .0024 0019 0008 .0004 .0013 WOkFS 1 61.9708 1.3867 0559 -.2324 .1275 50.4741 -1.2900 -38.6896 -50.9173 4.4623 -24.4465 -36 5797 SpB 1 .7685 0095 .0000 .0000 0000 .2184 -.0038 .0042 0034 -.0015 .0007 .0023 WOkOS 1 0668 .0015 -.0001 -.0003 0001 -1.2273 .0176 -14.2919 -13.0192 -2.5718 -12 0962 -14.2487 Rp0S 1 .0004 .0000 .0000 .0000 .0000 -.0574 0010 .0059 0085 .0261 .0055 0138 WSkFS 1 .1124 .0025 -.0001 -.0004 .0002 .1873 -.0039 -.3989 -.3189 1388 -.0673 -.2189 WSkOS 1 .0010 0000 0000 .0000 .0000 -.0178 .0003 -.2078 -.1893 -.0374 -.1759 -.2072

PAGE 157

Table 37. Derived reduced-form elasticity and flexibility estimates for selected predetermined variables evaluated at the means of the respective variables. 1 Endogenous Elasticity (percent) V ar iables WPFFlt-l WPOFltl XPOCSt RPShSt 1st IC ASt ASt -1 ACt ACt 1 t RQFSt -. 0786 3097 -. 0001 4784 7688 0003 .0351 .0223 0000 .0000 RPFSt 1729 -.0119 0001 0168 0344 .0006 0006 .0006 .0000 0000 WOFFlt 0341 .4932 0001 6934 1.4278 -. 0006 .0200 .0172 .0000 0000 WPFFlt .0023 0223 0003 0314 .0644 0009 .0009 0009 0000 .0000 MQFSBt 9320 2 8475 .0008 3.8827 8 6015 .0039 .0526 0667 .0002 0003 MPFSBt 1 0212 0114 0000 0163 .0334 .0000 .0000 0000 .0000 .0000 ROOSt 1776 -1 0381 .0001 1.4146 3 1384 0006 0187 0241 0000 .0000 RPOSt .Q100 .4976 0000 0796 .1767 0000 0009 .0014 0000 0000 WOOFlt .0943 -. 5514 0001 .7514 1.6670 .0003 .0099 .0128 .0000 .0000 I-' WPOFlt 0031 -. 0183 0000 0249 .0557 .0000 0000 0007 .0000 0000 w XQFCBt .1617 -.0102 3.9405 .0144 0295 3 8724 .0004 .0004 .1368 0660 XPFCBt .3133 -. 0199 0003 0280 .0519 0010 0000 .0000 .0000 0000 XOFCSt .1008 0064 4308 0090 -.0184 2.0362 0003 0002 -.1131 1625 XPFCSt .1907 .0120 .0001 .0170 .0349 0004 0008 0000 0000 0000 XOFEBt -. 2139 0059 0001 0083 .0170 -.0003 0003 0002 .0000 0000 XPFEBt .2504 -. 0069 0000 .0098 .0198 .0000 .0000 .0000 0000 0000 XOFESt -.7522 .0208 0002 0293 0599 .0010 .0009 .0007 .0001 0001 XPFESt 4463 0124 0000 .0175 .0354 .0000 0000 0000 .0000 0000 XOFNBt -.0423 .0040 .0000 -.0056 .0115 -.0002 -.0002 -.0001 .0000 0000 XPFNBt 0931 .0087 0000 .0122 .0257 0000 .0000 0000 0000 0000 XOFNSt .2528 .0238 -.0002 0336 -.0686 .0011 .0010 -.0009 .0001 -. 0001 XPFNSt 1167 -.0110 0001 0155 .0316 .0000 .0000 0000 0000 .0000 WQkFSt 0281 1164 -.0061 -.1666 .3281 -.0288 -.0063 0048 0016 0023 SpBt 8494 0313 .0002 .0444 0903 .0017 .0017 0017 .0000 0000 WOkOSt .0824 .4816 -.0001 -.6563 -1.4562 -.0003 -.0087 0112 0000 .0000 RpOSt -. 1863 1.0885 0000 -1.4832 3 2910 0010 0196 -. 0258 0000 .0000 WSkFSt -. 0607 .5954 -.0058 .8400 -1.7153 0277 -.0258 0216 .0015 0022 WSkOSt -. 1744 1.0196 0001 1 3892 -3 0825 0004 -.0184 -.0236 .0000 .0000 1 S ee Appendix F for means of the variables used in the model. (continued)

PAGE 158

Table 37 {continued). Endogenous Elasticity (percent) Flexibility (percent) Variables AEt ANt RxBSt RxSCt RxSEt RxSNt PdSt PdSt_ 1 PdSpt-6 ROFSt 0000 0000 0001 0000 -. 0001 .0001 1836 1451 0000 RPFSt 0000 0000 0000 0000 0000 .0000 4102 -.2314 -.0001 WOFFlt 0001 0001 0002 0001 -. 0003 0002 4530 -. 1471 .0001 WPFFlt .0000 0000 0000 0000 0000 0000 -. 7699 -.4343 .0001 MOFSBt 0005 0008 .2268 .0006 0022 -.0011 -3 0212 -.4552 .1856 MPFSBt .0000 .0000 -. 1642 .0000 0000 .0000 -.3944 -. 2225 -.1341 ROOSt .0001 0001 -. 0002 0001 0003 -.0001 -. 4309 1078 -.0001 RPOSt 0000 0000 .0000 0000 0000 0000 -.0192 -.0766 0000 WOOFlt 0000 0001 -.0001 .0001 .0002 0001 -. 2410 .1105 0000 WPOFlt 0000 0000 0000 0000 0000 .0000 0454 -.7363 .0000 XOFCBt .0001 0001 6507 4482 .0002 -. 0001 -.3535 1995 0353 XPFCBt 0000 .0000 -.6202 -.8662 .0000 .0000 .6851 3865 -.0885 I-' XOFCSt 0000 0001 0394 4473 -.0002 0002 .3187 .1990 0242 ,i:,. .i,. XPFCSt .0000 .0000 .0304 5287 .0000 .0000 .4170 .2353 0250 XOFEBt .5292 0000 1.5296 0000 6679 .0001 .2042 .1152 0943 XPFEBt 0000 0000 -2 1055 .0000 -.7847 0000 -. 2390 -.1348 -.1244 XOFESt 5028 0000 -.0615 0000 2.3485 .0000 .7179 .4050 .0906 XPFESt .0000 0000 -.0715 .0000 -1.3934 0000 4259 -.2 403 .0585 XOFNBt .0000 -. 9574 4762 0000 0001 .3184 1377 0777 -.0173 XPFNBt 0000 .0000 -1.4062 0000 0000 6909 3023 .1706 .0556 XQFNSt 0001 1.2326 -. 4806 -.0001 -.0006 -1 9026 .7349 .4640 .0584 XPFNSt 0000 .0000 -.0185 .0000 .0000 8779 -.3794 1908 -.0152 WQkFSt -.0034 -. 0057 0076 0040 -.0160 0088 -.2581 0378 .0034 SpBt 0000 0000 .1372 .0000 .0000 .0000 -1.0836 6114 .1894 WQkOSt 0000 .0000 0001 -.0001 -.0002 .0001 .2410 -.1105 0000 RpOSt 0000 .0000 0000 0000 .0000 0000 .4518 1130 0001 WSkFSt -.0033 .0045 .0073 -.0037 -.0145 0072 .6944 .1806 0033 WSkOSt 0000 .0000 0000 .0000 .0000 .0000 .4455 -.2043 .0001

PAGE 159

145 The most important effects of lagged wholesale price of FCOJ in Florida (WPFFlt-l) are on the USA imports of FCOJ from Brazil, the USA exports, especially to the EEC7 countries, and Brazilian exports to Canada and to the EEC7 countries. Low previous period prices (WPFF1t_ 1 ) results in a more than proportional decrease in the current USA import price of FCOJ from Brazil (MPFSBt). However, the impact on Florida's current wholesale price (WPFFlt and WPOFlt) is practically zero. The net effect is an increase in the price spread (SpBt) and more imports from Brazil are brought into the USA. The extended effects on FCOJ exports to Europe and Canada from the USA and Brazil are directly related to the effects on the USA imports from Brazil. USA FCOJ export prices are affected by the current Florida wholesale price (WPFFlt) and the import price from Brazil (MPFSBt). Since WPFFlt is relatively unaffected, the impact on export prices of WPFFlt-l is through the reduction on MPFSBt. As this price declines, USA export prices also decline, as indicated by the transmission elasticities in Table 37 (column one). Given lower export prices due to cheaper imports, USA exports to the EEC7 countries and to Canada increase as expected. However, USA exports to the European non-EEC3 countries declined, even though the respective export price also declined. This means that the substitution effect of the lower Brazilian price offsets the effect of the lower USA price. In addition, it is possible that some imports within this period are contractually related to prior

PAGE 160

146 period prices. Brazil's FCOJ exports to the European and Canadian markets are expected to be affected by low previous wholesale prices in Florida in two different ways. The first effect is a reduction of the Brazilian prices to these markets, since the corresponding prices from the USA declined and Brazilian prices are positively related to USA prices. Thus, more exports of FCOJ from Brazil are expected to Europe and Canada. The second effect is that lower prices of FCOJ from the USA in those countries should reduce Brazilian FCOJ exports to the same markets, since some degree of substitution is expected to exist between the two FCOJ sources. The elasticity estimates in Table 37 express the net effect of these opposite forces. In Canada, the net result is a reduction in Brazilian exports (-.1617 percent) implying that the substitution effect is greater than the price effect. In Europe the net effect is positive suggesting that the volume of Brazil's exports to the EEC7 and the European non-EEC3 countries increase. The magnitude of these effects are quite different. To the EEC7 countries, a 1 percent decrease in WPFFlt-l will increase exports from Brazil by nearly .2 percent. However, to the European non-EEC3 countries, the effect on Brazil's FCOJ exports is very small (-.0423 percent). Reduced-form elasticity estimates relative to the lagged wholesale price of other orange juices in Florida (Table 37, column two) suggests major impacts on the USA imports of FCOJ from Brazil (-2.8475 percent) and on the

PAGE 161

147 retail domestic market for other orange juices. One percent reduction of WPOFlt-l is likely to increase retail consumption of other orange juices in the USA by more than 1 percent and increase FCOJ imports from Brazil by almost 3 percent. These results suggest that imports in the USA domestic market have increased as a result of other orange juices' demand growth (chilled and canned single strength). In the Canadian market, estimates of the effects of the USA export price of other orange juices (Table 37, column three) relative to FCOJ exports from Brazil and the USA (XQFCBt and XQFCSt) indicate that the Brazilian product exhibits an elastic (3.9405 percent) response while the USA product exhibits a relatively inelastic (.4308 percent) response. If XPOCSt increases by 1 percent, exports of FCOJ from Brazil increase by almost 4 percent while the USA exports increase by less than .5 percent. These results support the idea that differentiation of the USA FCOJ has resulted in a very efficient protective barrier to competition from other sources. Derived reduced-form coefficient estimates of retail price of frozen orange-flavored synthetics and drinks in the USA (RPShSt) seem to indicate strong substitute relationships in the USA demand for FCOJ imports from Brazil (MQFSBt) and in the retail domestic market for other orange juices (RQOSt). Both cross-price elasticities (Table 37, column four) are elastic, but imports of FCOJ from Brazil are relatively more responsive to changes in

PAGE 162

148 RPShSt (3.8827 percent) than retail consumption of other orange juices (1.4146 percent). Some effects are also translated to the USA domestic market for FCOJ (RQFSt and WQFFlt). This analysis suggests that imports brought into the USA are related to increases in consumption of orange juices other than FCOJ. Imports of FCOJ from Brazil (MQFSBt) and other orange juices (RQOSt and WQOFlt) are positively related to income. Reduced-form estimates of income elasticities (Table 37, column five) indicate that for each 1 percent increase in income per capita in the USA (ISt) the results of a chain reaction would be an increase of almost 9 percent in FCOJ imports from Brazil (8.6015 percent), of more than 3 percent in retail consumption of other orange juices (3.1384 percent), and of almost 2 percent in the corresponding wholesale sales in Florida (1.6670 percent). FCOJ domestic sales are likely to increase by practically .8 percent at the retail level (.7688 percent) and by more than 1 percent at the wholesale level in Florida (1.4278 percent). In Canada, income effects are relatively strong for both sources of supply. As the estimates of income elasticities (Table 37, column six) indicate, a 1 percent increase in per capita income in Canada is expected to increase Brazilian FCOJ exports (XQFCBt) by almost 4 percent and USA exports (XQFCSt) by more than 2 percent. This relatively favorable response for the Brazilian product may suggest that Canadian consumers that use orange juice of

PAGE 163

149 higher quality (assumed to be the USA product) do not increase consumption at the same rate as the other consumers or that new consumers are more likely to be using the Brazilian product. Impact of Exchange Rate The impact on endogenous variables of changing the rate of the Brazilian cruzeiro relative to the USA dollar, and the rates of the USA dollar relative to the Canadian dollar and to European currencies, expressed in terms of 1975 prices, are presented in Table 37 (columns thirteen to sixteen), as reduced-form elasticities. As expected the impacts of changing exchange rates are stronger on prices than on export quantities. Larger structural-form estimate of the rate of USA dollar per Canadian dollar (RxSCt) relative to the associated standard error (Table 27), makes the respective reduced-form elasticity estimates (Table 37, column fourteen) quite relevant. One percent devaluation of the USA dollar relative to Canadian currency is likely to decrease USA FCOJ export prices to Canada (XPFCSt) by more than .5 percent and consequently exports of FCOJ increase almost by the same percentage. The extended effect reduced Brazil's export price to Canada (XPFCBt) by almost .9 percent and export quantity {XQFCBt) by more than .4 percent. This last result is the net effect of the impact of lower export prices of Brazilian FCOJ which is directly related to the USA price (XPFCSt), thus reducing exports of FCOJ from Brazil. The

PAGE 164

150 implications for FCOJ exports to the Canadian market are important, since devaluation of the USA dollar against the Canadian dollar is likely to increase exports of the USA product and to decrease sales of the Brazilian FCOJ. Increased valuations of the dollar, which occurred during the time period of this study, is likely to improve Brazil's exports and to reduce USA exports to this market. Estimates of the indices used to capture the effects of variation in the rate of the dollar relative to European currencies on USA exports of FCOJ are different for the two groups of countries. Decreasing the indices (devaluation of dollar relative to all currencies for a given group of countries) is expected to reduce prices and consequently to increase exports. Increasing the indices is expected to have the opposite effect which would result in higher export prices and consequently less FCOJ exported. Reduced-form elasticity estimates (Table 37, columns fifteen and sixteen) conform with the expectations only in the case of the European non-EEC3 countries. If the index decreases by 1 percent the USA export price of FCOJ (XQFNSt) is likely to decrease by almost .9 percent and exports are expected to increase by practically 2 percent. The extended effects on exports from Brazil are also consistent with the expectations, but the elasticities are relatively small. The estimates relative to the EEC? countries are only partially consistent with that rationale. Based on the elasticity estimates, a reduction of the index by 1 percent

PAGE 165

151 will increase the USA export price by more than 1 percent and will reduce exports by more than 2 percent. Extended effects on Brazilian exports to this market have the same directional impact on prices and quantities exported. The relationships between export prices and export quantities are consistent in both cases. However, reduction of the index in this market is associated with higher export prices. This apparent inconsistency is related to the fact that the rates of the dollar relative to a group of foreign currencies do not necessarily change in the same direction, even though the derived index of these exchange rates may suggest a net devaluation or increased valuation. In these cases the effects on FCOJ exports (price and quantity} will not be related to the direction of the index itself but on the relative impact of variation in individual exchange rates on exports to the respective countries. Thus, it is possible that a given value of the index may not indicate that all individual exchange rates used in its derivation have changed in the same direction, and that export price and quantity may change differently as in the case of same directional change of all exchange rates. Impact of Orange Production The effects of variation in orange production in the USA and in Sao Paulo (Brazil} on trade of orange juice are analyzed through the estimates of the flexibilities in Table 37 (columns eighteen to nineteen}. It should be

PAGE 166

152 recalled that crop estimate variables refer to the average bimonthly forecasts of annual production in the case of the USA and to yearly estimates of production in the state of Sao Paulo in the case of Brazil. Current and lagged reduced-form estimates of flexi bilities for orange production in the USA (PdSt and PdSt_ 1 ) have important implications for the orange juice trade. The impacts on USA FCOJ imports are quite strong. For example, if the current crop estimate (PdSt) decreases, import prices from Brazil (MPFSBt) increase since Florida's wholesale price for FCOJ increases and the Brazilian price is a direct fucntion of Florida's price (WPFFlt). However, the FOCJ import price from Brazil increases relatively less than the wholesale price in Florida. Thus, 1 percent decrease in PdSt is expected to increase the price spread (WPFFlt MPFSBt) by more than 1 percent and imports from Brazil by more than 3 percent. It is important to emphasize the highly elastic response of the USA imports from Brazil to changes in current estimates of ornage production (PdSt). In Canada, Brazil's FCOJ exports {XQFCBt) also increase even though export prices increase. Brazil ultimately exports more because the substitution effect associated with a higher export price from the USA is more than sufficient to offset the price effect associated with a higher export price of the Brazilian FCOJ. In Europe the two products seem to be quite independent of each other and as price increases exported quantity is

PAGE 167

153 likely to decrease. In the EEC7 market for FCOJ exports from Brazil and the USA the result of reduction of current orange production forecasts in the USA is higher prices and consequently smaller quantities exported from both suppliers. There is an exception in the case of Brazil's exports to the European non-EEC3 countries. As current crop estimates indicate, a decrease in orange production in the USA leads to a decrease of both export price and quantity of FCOJ from Brazil. The reduced-form flexibility estimates relative to variations in yearly orange production in Sio Paulo (Brazil) indicate that their impacts on the orange juice trade are small and perhaps not very meaningful, given that some of the structural-form estimates involved in the derivation are small relative to their associated standard errors (Tables 19, 25, 29 and 33). Impact of Freeze The analysis relative to the impact of freeze in the USA, or more specifically in Florida (Dt), as formulated in the structural-form of the model, will be based on the derived reduced-form estimates presented in Table 36. The occurrence of freezes in Florida is likely to add more than 28 cents per gallon to the wholesale price of Florida's FCOJ (WPFFlt) and practically 32 cents per gallon to the wholesale price of Florida's other orange juice prices (WPOFlt). Even though the effects on prices are relatively higher on other orange juices, the consequent

PAGE 168

154 impacts on wholesale orange juice movements are quite different. As prices increase Florida's FCOJ movement decreases by almost 20 gallons per thousand persons and other orange juice movement increases by nearly one gallon per thousand persons. Associated with these changes there is a substantial increase in FCOJ stocks and a small reduction on other juices' stocks, which is practically equal to the change in WQOFlt. In order to rationalize these results it is necessary to look at the extended effects on the USA imports from Brazil, exports to other markets and domestic consumption. Import price of FCOJ from Brazil is expected to have a small increase of nearly 6 cents per gallon that, associated with the variation in Florida's wholesale price (WPFFlt), results in increase in the spread (SpBt) of almost 22 cents per gallon. As the spread widens, imports of FCOJ from Brazil increase by approximately 25 gallons per thousand persons. At retail level in the USA, higher FCOJ prices decrease consumption substantially. However, other orange juice prices increase only slightly and consumption increases by more than 15 gallons per thousand persons. A strong substitution effect toward chilled and canned single strength is quite evident. The impact on reduction of the USA exports of FCOJ is relatively high in the European non-EEC3 countries, and smaller in Canada and the EEC7 countries. The occurrence of a freeze in Florida also affects Brazilian exports of FCOJ to these markets. In European markets higher prices of

PAGE 169

155 the Brazilian product result in less exports for the EEC7 and the European non-EEC3 countries. Higher prices for the USA product have no significant substitution effect toward the Brazilian product. In Canada, however, the estimates suggest that the effects of high prices on Brazilian FCOJ are more than offset by the substitution effect caused by high prices of the USA FCOJ to that market. These results again suggest that some substitution of Brazilian FCOJ for the USA FCOJ in Canada occurs. In the European markets the results indicate that these products are quite independent of each other. Tests on Performance of the Estimated Model In this sub-section some insights on how well the estimated model forecasts conform to observed data are presented and discussed based on the derived reduced-form. Estimated and observed values of endogenous variables are compared through the results of formal tests as well as through a set of graphs which give quantitative and visual indications on the "goodness of fit" of the estimated model. As described in Chapter III, ratios of the root mean square error to the respective observed mean value of endogenous variables for six different periods within the data range were computed (Table 38) to give some quantitative idea of the performance of individual equations of the estimated model. It is interesting to observe that more than 80 percent (136 out of 168) of the computed ratios are less than 10 percent of the respective observed mean value.

PAGE 170

Table 38. Ratios of the root mean square error to observed mean value for comparing actual and derived reduced-form estimated values of endogenous variables for specific periods within the data range used in the study Jun-Jul Dec-Jan Dec-Jan Dec-Jan Dec-Jan Dec-Jan Endogenous 1972 to 1973-74 to 1974-75 to 1975-76 to 1976-77 to 1977-78 to variables Oct-Nov Oct-Nov Oct-Nov Oct-Nov Oct-Nov Oct-Nov 1973 1974 1975 1976 1977 1978 RQFSt .0141 .0175 .0081 .0100 .0271 .0177 RPFSt .0156 .0279 .0108 .0089 .0197 .0225 WQFFlt .3047 .0330 .0274 .0438 .0651 .0219 WPFFlt .2016 .0430 .0256 .0261 .0474 .0134 I-' (J1 MQFSBt .6146 .4156 .3731 .4571 .6963 .2047 MPFSBt .0561 .0594 .4082 .0681 .0855 .0363 RQOSt .0522 .0457 .0210 .0422 .0341 .0130 RPOSt .0070 .0080 .0055 .0140 .0156 .0126 WQOFlt .0276 .0218 .0069 .0229 .0237 .0145 WPOFlt .0190 .0418 .0100 .2072 .0235 .0186 XQFCBt .1927 .4804 .1332 .2436 .1811 .0792 XPFCBt .0223 .0394 .0393 .0356 .1301 .0361 XQFCSt .0352 .0399 .0705 .0629 .0661 .0289 XPFCSt .0119 .0202 .0113 .0131 .0362 .0181 (continued)

PAGE 171

Table 38 (continued) Jun-Jul Dec-Jan Dec-Jan Dec-Jan Dec-Jan Dec-Jan Endogenous 1972 to 1973-74 to 1974-75 to 1975-76 to 1976-77 to 1977-78 to variables Oct-Nov Oct-Nov Oct-Nov Oct-Nov Oct-Nov Oct-Nov 1973 1974 1975 1976 1977 1978 XQFEBt .0881 .2538 .0877 .0819 .1117 .1035 XPFEBt .0291 .0586 .0545 .0895 .1068 .0289 XQFESt .1216 .3707 .3760 .1052 .1991 .1780 XPFESt .0641 .0886 .0800 .0830 .0954 .0464 XQFNBt .2122 .1337 .1144 .1861 .1619 .1060 I-' U1 -i XPFNBt .0247 .0329 .0511 .0935 .0968 .0290 XQFNSt .1001 .2568 .0897 .0841 .0981 .1871 XPFNSt .0300 .0746 .0427 .0226 .0514 .0275 WQkFSt .0115 0096 .0102 .0126 .0429 .0205 WSkFSt .0433 .0340 .0396 .0604 .0762 .0425 WQkOSt .0322 .0421 .0347 .0482 .0171 .0420 WSkO~ t .0601 .0421 .0345 .0668 .0323 .0597 SpBt .0635 .1032 .0610 .0690 .0989 .0294 RpOSt .0474 .0345 .0204 .0455 .0462 .0303

PAGE 172

158 Most of the ratios that are greater than 10 percent are from equations for the USA FCOJ imports from Brazil and exports to the EEC7 countries, and Brazilian exports of FCOJ to Canada and to the European non-EEC3 countries. The final test on performance of the estimated model is given by the set of graphs (Figures 1 to 28) in Appendix II, where actual and reduced-form estimated values of endogenous variables are plotted over time. Together with previous tests and with statistical results of the structural-form estimates,this set of graphs gives some indication of the "goodness of fit" of the model.

PAGE 173

CHAPTER V SUMMARY AND CONCLUSIONS This chapter includes a brief review of major points addressed in the preceding chapters, concluding remarks associated with the study objectives, and suggestions with respect to improvement of future studies. Swnmary In recent years, trade of orange juices became an important economic issue for a large number of countries. On one side, for the leading producer and exporting countries like Brazil and the USA, orange juice and particularly frozen concentrated orange juice (FCOJ) exports have been significant as sources of foreign revenues as well as important to the economic survival of many internal industries, especially in Brazil. On the other side, for the importing (or producer-importer-exporter) countries like the EEC7 (seven countries of the European Economic Community), the European non-EEC3 (three European, nonEEC countries), Canada and the USA, FCOJ imports have also been quite important in fulfilling demand for orange juice. For the USA, as a producer, the foreign trade sector also adds substantial economic activity that may not have occurred otherwise. 159

PAGE 174

160 The major trade models addressing the orange juice issues with an international perspective (Ward, 1976a; Moretti, 1978; Tilley & Lee, 1981) have concentrated on one sector or on one country or region. In the present study, the economic model for orange juices encompasses the international linkage among the two leading producing countries and leading importing regions. Special emphasis is placed on the role of the foreign trade segment on the domestic orange juice industry. The objectives of this study are: 1. to identify major factors affecting trade of orange juice; 2. to evaluate the extend to which changes in domestic and foreign demands for the USA orange juice affect the USA FCOJ exports and imports; 3. to evaluate whether foreign and domestic promotion programs for the USA (Florida) orange juice provide an "umbrella" for market expansion of similar products from other producing-exporting countries; and 4. to determine the impact of pricing strategies and price changes on exports and imports of orange juice among producing-exporting and importing countries. A simultaneous equation econometric model was developed for major countries involved in orange juice trade. The analytical scenario consists of two orange juice products (frozen concentrated and other orange juices), two major producing and exporting countries (the USA and

PAGE 175

161 Brazil) and four major importing countries or groups of countries (Canada, the USA, the EEC7 and the European non-EEC3 countries). The EEC7 countries are West Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark and France. The European non-EEC3 countries are Sweden, Norway and Finland. The conceptual framework consists of a system of simultaneous demand and price equations for FCOJ and other orange juices for the USA retail, wholesale (Florida) and import (FCOJ from Brazil) markets, and Brazilian and USA FCOJ export markets in Canada, the EEC? and the European non-EEC3. Price equations stem from the simultaneous nature of the model and measure price transmission effects within different levels of a given market and across markets. Brazilian and USA orange juice supplies are assumed to be predetermined, given the nature of orange production and the time unit of the study (bimonth). Only the USA domestic retail market is included in the model, given its magnitude, as well as its linkages with export and import markets. The Brazilian domestic market is excluded because it is small and systematic data are not available. In the other markets (Canadian and European) trade of orange juice was simplified to include only the FCOJ export sector to these markets from Brazil and from the USA. The domestic orange juice producing sector in Canada and Europe is small or non-existent and it is reasonable to assume that all domestic demand forces

PAGE 176

162 can be captured by international trade demand and price equations. Brazil and the USA are by far the major suppliers of orange juice to these markets, and FCOJ is the most important traded product. The conceptual formulation of the economic model resulted in a system of simultaneous equations with 28 endogenous variables (22 behavioral and 6 identities) and 41 predetermined (6 lagged endogenous and 35 exogenous) variables. Two-stage least squares (2SLS) was used to estimate the over-identified structural-form equations of the system. Major Conclusions Two groups of factors are relevant to trade of orange juices (FCOJ and other orange juices) within different markets and across markets. One set of factors affects orange juice trade indirectly through prices in the system. The other set includes factors that have a major impact on quantity of orange juice traded. In the first group of factors, the analyses of this study are quite consistent in stressing the significant roles of Florida's wholesale prices, of orange production forecasts in the USA, of the USA import and export prices, of exchange rates and of Florida's freeze. Besides their important role in the derivation of the respective retail prices of orange juices in the USA, Florida's wholesale prices, especially of FCOJ, were found to be relevant in the simultaneous determination

PAGE 177

163 of the USA import price of Brazilian FCOJ and consequently in determining the price spread as well as in the derivation of the USA export prices of FCOJ. Current and previous wholesale prices of FCOJ in Florida are of major importance for Brazilian exporters to the USA. The current FCOJ wholesale price in Florida and import price from Brazil define the price spread which is an important variable in the simultaneous determination of the USA FCOJ imports from Brazil. These same prices are also of major importance in the simultaneous determination of the USA FCOJ export prices to Europe and Canada. The direct effects of orange production forecasts in the USA and Florida's freeze on orange juice prices are through the formulation of Florida wholesale prices. The extended effects on other prices and to traded quantities are estimated with the derived reduced-form estimates. These orange forecasts significantly affect Florida wholesale prices (especially of FCOJ) and the USA import price of FCOJ from Brazil. Since the impact on the import price was found to be relatively less than on Florida's price, then the price spread changed by the respective differential resulted in a significant impact on the USA imports of FCOJ from Brazil. A 1 percent reduction of current USA orange production is likely to cause an increase of .8 percent in Florida's FCOJ wholesale price of nearly .4 percent in the import price of FCOJ from Brazil, of more than 1 percent on the price spread and consequently of more than 3 percent on

PAGE 178

164 FCOJ imports by the USA from Brazil. Thus, changes in USA imports from Brazil are relatively elastic with respect to the price spread, given some reduction in the current USA orange production forecast. Higher wholesale prices of FCOJ in Florida and of prices of Brazilian FCOJ imports are likely to result in higher export prices for FCOJ from the USA and from Brazil. Higher export prices are likely at first to reduce exports of FCOJ to major consuming regions. The net effect on quantity of each supplier to each importing market will depend on the degree of substitution between the two products in each market. Derived reduced-form flexibilities indicate that, in the USA and Canadian markets, Brazil's FCOJ exports had a net increase, even though the corresponding export price also increased. These results suggested a relatively larger substitution effect on Brazil's exports to Canada and a relatively larger price effect on the USA exports to the same market. Brazilian and USA exports to Europe were found to be relatively independent of each other. The occurrence of a freeze in Florida is estimated to add approximately 30 cents per gallon to Florida's wholesale price of FCOJ and of other orange juices. Extended effect on USA import prices of FCOJ from Brazil was estimated at nearly 6 cents per gallon, which results in a substantial increase in the spread and consequently more FCOJ imports from Brazil and the USA. The extended effects to other export markets of FCOJ from Brazil and from the USA are translated as higher prices and the net effect on quantities

PAGE 179

165 indicate that, again, only Brazil's FCOJ net exports to Canada increased. All other relations indicate that price effects are sufficiently large to offset substitution effects, and exports of FCOJ from the USA and Brazil are reduced in Europe USA exports to Canada are reduced and Brazil's FCOJ exports to Canada increase. Reduced-form estimates of exchange rate impacts indicate that devaluations of the Brazilian cruzeiro relative to the USA dollar are likely to result in more exports of Brazilian FCOJ to European markets, especially to the EEC7 countries, through the primary effects of reducing export prices. If the rate of cruzeiros per dollar increases by 1 percent, the Brazilian FCOJ export price to the EEC7 market would decrease by more than 2 percent and consequently exports would increase by nearly 1.5 percent. Exchange rate estimates for the USA dollar relative to the Canadian dollar are important in explaining variation in USA exports of FCOJ to Canada. The analyses indicated that, as the USA dollar increases in value relative to the Canadian dollar by 1 percent, the USA export price increases by nearly .5 percent and exports decrease by approximately .4 percent. The extended effects also increase the Brazilian FCOJ export price to Canada by almost .9 percent but the larger substitution effect, given high USA prices, cause Brazilian exports to Canada to increase by almost .4 percent. In the European markets the estimates of the indices of exchange rates relative to the USA dollar are important to

PAGE 180

166 USA export prices and quantities. However, given a problem of consistency of the index (devaluation relative to a given currency does not necessarily mean devaluation to any other currency in the index), only the estimats for the European non-EEC3 countries agree with unilateral changes relative to the dollar. As the index decreases, which is associated with devaluation of the dollar, the USA FCOJ export price also decreases and exports increase. One percent decrease in the index is related to almost .9 percent decrease in export price and nearly 2 percent increase in USA FCOJ exports to the European non-EEC3 countries. The extension to Brazil's FCOJ exports to the same market results in higher prices and in a reduction in exports of a relatively insignificant magnitude. The results for the EEC7 countries indicate that, as the index decreases, the USA export price to this market increases by 1.13934 percent) andexportsdecrease by 2.3485 percent. Since the possibility exists that individual exchange rates in the index change in different directions, there was not a unique way to relate these changes to devaluation or increased valuation of the dollar. Even though smaller in magnitude, the extended effects upon Brazil's exports of FCOJ to the EEC7 countries were similar to the effects on USA exports. Major factors affecting the quantity of orange juice traded are own-prices, prices of substitute products, income, and seasonal variations. Most of the structural form estimates relative to own-price variables are larger

PAGE 181

167 than the associated standard errors so that the corresponding numerical values can be viewed with some confidence. Structural-form elasticity estimates indicate that FCOJ demand is elastic for the USA imports from Brazil with respect to the price spread (1.6607 percent), for Brazilian exports to Canada with respect to own-price (-1.1884 percent), the USA exports to the EEC7 (-1.7351 percent), and to the European non-EEC3 countries (-2.4662 percent) with respect ot own-price. Brazilian own-price elasticities for FCOJ exports to the European markets were found to be inelastic in both cases, -.7273 percent to the EEC7 and -.3374 percent to the European non-EEC3. The USA own-price elasticity of FCOJ exports to Canada was also found to be inelastic (-.4616 percent). Estimates of income elasticities are 8.6015 percent for USA imports of FCOJ from Brazil, 3.1384 percent for retail consumption of other orange juices in the USA, and 3.8724 and 2.0362 for Brazilian and USA FCOJ exports to Canada. Most estimates of advertising coefficients were small relative to their respective standard errors. In the European markets the results, in general, indicate positive impact on USA FCOJ exports, especially to : the European non-EEC3 countries. In this market the estimates support the effectiveness of the program and the extended effect on Brazil's exports of FCOJ to this market is negative. In the EEC7 market the program results are both positive and of same approximated relative magnitude for the USA and Brazilian

PAGE 182

168 exports of FCOJ. Even through there is no clear.,..cut evidence to fully support the statement that foreign and domestic programs of the USA (Florida) orange juice products provide an "umbrella" for market expansion for the same products from other suppliers (Brazil), the analysis does not conclusively reject that hypothesis. Analyses of seasonal variation suggest that in the USA market, demand for orange juices increases during relatively colder months (October to March) and decreases during those months that are relatively warmer (April to September). This pattern is also quite similar to the one for imports of FCOJ from Brazil. In Canada, Brazil and the USA FCOJ exports follow the same seasonal pattern. In the EEC7 market these two suppliers of FCOJ appear to have opposite seasonal tendencies. October-November bimonths appear to be the peak for Brazilian exports and the lowest for USA exports. April-May bimonths are the lowest USA exports. April-May bimonths are the lowest for Brazil and the peak for the USA. Suggestions for Future Studies Improvements in the actual data base should be oriented to correct deficiencies in the Brazilian trade with the USA (USA imports or Brazil exports}, in Brazil's exports to the European countries, especially to the European non-EEC3 countries (_Sweden, Norway and Finland), and in the USA trade with Canada (USA exports or Canadian

PAGE 183

169 imports). Data on per capita income in Europe, instead of a time proxy as used in this study, would be useful. Expansion of the model in future studies should give endogenous treatment to the USA exports of other orange juices to Canada and should include Israeli exports to Europe. Some variables did not respond as expected to the specification used in this model such as advertising, exchange rate index and stock variables. Alternative formulations (other than linear functional form) should be considered, as well as distributed lag structures in the case of advertising variables. The exchange rate indices developed to measure the effects on prices from variations of individual exchange rates for the European markets did not perform consistently to indicate devaluation or increased valuation of the currency considered. Alternative ways to measure these changes may also improve the overall performance of the model. Estimates of the lag structure, such as for Florida's wholesale prices of FCOJ and orange production forecasts in the USA, suggested that the respective dependent variables (prices in these examples) may have given better results if a longer lag structure is used. The final set of suggestions deals with further analyses based on the estimates of this study. For example, more accuracy of the reduced-form estimates is obtainable, as pointed out by Womack and Matthews (1972, p. 98), if evaluation at some point other than the mean values (say

PAGE 184

170 previous period values or last values of the data series) is used to linearize the ratios of the equations for FCOJ stock, retail proportion of other orange juices in cardboard containers, and other orange juices' stocks by Taylor's Series expansions.

PAGE 185

APPENDIX A EXPORTS OF ORANGE JUICE

PAGE 186

Table A.l. Leading orange juice exporting countries, 1969-1978 Year 'la USAb C d Spain e South Brazi Mexico Israel f e Arica ---------------1,000 gallons of single strength juice equivalent------------1969 29,528 28,805 381 23,779 14,228 12,322 1970 42,513 42,278 1,016 23,162 23,120 9,782 1971 98,238 46,199 2,287 42,458 23,882 8,765 1972 110,719 50,205 8,384 43.192 26,931 10,671 1-e' 1973 153,710 62,522 13,212 43,406 24,518 19,055 "1 I') 1974 137,782 64,106 13,974 49,078 23,120 11,433 1975 229,802 70,140 4,446 42,764 20,834 6,987 1976 266,568 85,418 10,798 71,705 45,224 8,257 1977 271,248 86,539 30,361 38,242 21,596 5,208 1978 423,865 58,037 29,853 42,859 24,518 5,717 Total 1,763,973 594,249 114,712 390,645 247,971 98,197 Percent 44.49 14.99 2.89 9.85 6.25 2.48 (c;:ontinued)

PAGE 187

Table A. l. Year 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 Total Percent (continued) f Greece Moroccoe e Italy 1 e Be ize Argentinae Other 1 countriese, Total --------------1,000 gallons of single strength juice equivalent--------------3,811 6,352 12,703 12,703 12,703 12,703 10,671 2 9,146 2 12,830 2 103,531 2.61 23,247 29,091 14,863 20,452 22-866 15,752 10,671 13,466 13,466 13,466 177,340 4.47 30,107 32,647 29,090 30,488 24,898 21,087 16,641 5,970 8,765 11,814 211,507 5.33 7,113 5,589 7,495 7,113 7,367 7,340 5,716 5,208 4,700 4,954 62,495 1.58 635 1,143 7,876 6,605 4,065 3,049 2,286 3,430 3,430 3,557 36,076 .91 17,530 17,149 16,641 15,244 15,752 24,644 11,941 17,022 14,100 14,608 164,631 4.15 197,584 231,301 304,146 342,707 404,074 383,968 434,931 513,737 506,801 646,078 3,965,327 100.00 1 Includes Algeria, Cyprus, Ghana, Jamaica, and Trinidad-Tobago. 2 Estimates. Sources: Computed from aBrazil, Servi90 de Estatistica Economica Financeira, Ministerio da Fazenda, 1970-1979; bunited States Department of Commerca, Bureau of the Census, 1969-1979a; CMexico, Secretaria de Industria y Comercio, 1972-1979; Israel, Central Bureau of Statistics, 1970-1980; eFood and Agriculture Organization of the United Nations, 1980; and fMorais and Medeiros, 1978.

PAGE 188

Table A.2. Brazilian exports of orange juice by country of destination, 1969-1979 Year West United The Belgium & Denmark France EEC Sweden Germany Kingdom Netherlands Luxembourg total -----------------1,000 gallons of single strength juice equivalent-------------1969 12,172 608 4,454 113 330 00 17,677 1,059 1970 24,200 739 5,212 366 241 00 30,758 1,322 1971 35,584 987 11,201 1,062 880 683 50,397 3,156 1972 39,870 920 18,550 2,017 899 783 63,039 8,345 1:-' 1973 70,238 579 33,110 1,679 1,059 884 107,549 13,200 -...J 1974 43,034 752 26,243 2,066 673 305 73,073 19,495 1975 56,041 5,358 54,472 2,721 1,371 254 120,217 20,491 1976 55,256 6,870 85,453 5,567 5,440 305 158,891 23,203 1977 39,873 9,798 67,614 2,726 7,354 51 127,416 24,593 1978 37,419 17,030 57,947 2,292 5,724 00 120,412 24,190 1979 50,068 19,510 73,962 4,936 6,988 00 155,464 22,338 Total 463,755 63,151 438,218 25,545 30,959 3,265 1,024,893 161,392 Percent 21.72 2.96 20.52 1.20 1.45 .15 48.00 7.56 (continued)

PAGE 189

Table A.2. (continued) Year Other European Norway Finland European 1 non-EEC Canada USA Israel Oth er 2 Total countries non-EEC total ----------------1,000 gallons of single strength juice equivalent----------------1969 1970 1971 1972 1973 1974 111 192 555 503 503 849 19 13 133 514 1,380 1,180 1~75 1,538 3,875 1 0' 76 1,931 7,622 1977 2,177 6,302 1978 2,722 10,135 1979 3,844 13,317 Total 14,925 44,490 Percent .70 2.08 1 230 193 127 597 2,051 1,623 11,529 9,971 10,516 5,133 8,853 50,823 2.38 1,419 1. 720 3,971 9 959 17,134 23,147 37,'133 42,727 43,588 42,180 48,35 2 271,630 12.72 5,940 5,448 13,242 13,547 11,170 9,459 3,587 1,277 29,720 22,428 13,952 29,360 699 2,505 318 1,551 3,393 1,397 26,828 26,180 11,329 20,117 21,183 18,426 26,281 59,461 7,167 43,021 187,389 11,829 36,634 92,805 13,128 211,687 487,342 71,742 9.91 22.82 3.36 206 805 590 195 512 1,346 7,815 5,224 7,335 19,034 25,019 68,081 3.19 29,528 42,513 98,238 110,719 153,710 137,782 229,802 266,568 271,248 423,865 371,402 2,135,375 100.00 2 Includes Austria, East Germany, Poland, Portugal, Spain, Switzerland and Yugoslavia. Includes Argentina, Australia, Bolivia, Chile, Japan, Mozambique, Netherlands Antilles, New Zealand, Paraguay, Senegal, South Africa and Venezuela. Source: Computed from Brazil, Servi90 do Estatfstica Econ6mica e Financeira, Ministerio de Fazenda,1970-1979. I-' ...J u,

PAGE 190

Table A.3. The United States of America exports of orange juice by country of destination, 1969-1979 West United The Belgium & Denmark France Other 1 EEC Year Germany Kingdom Netherlands Luxembourg EEC total ---------------------1,000 gallons of single strength juice equivalent-----------1969 2,162 1,347 1,246 463 264 1,278 188 6,948 1970 4,075 2,188 2,854 703 496 1,231 201 11,748 1971 3,735 2,105 2,502 661 192 1,278 130 10,603 1972 4,867 1,659 4,888 553 72 1,225 240 13,504 J.973 5,060 2,594 7,361 575 400 1,721 285 17,996 1974 3,421 2,085 4,397 516 442 1,534 132 12,527 1975 3,602 2,517 6,907 512 321 2,177 307 16,343 1976 8,104 3,651 10,681 1,699 661 2,478 300 27,574 1977 5,226 4,009 9,222 418 701 2,641 551 22,768 1978 2,587 1,666 3,074 391 442 2,453 342 10,955 1979 2,786 1,965 4,092 394 66 4,409 320 14,032 Total 45,625 25,786 57,224 6,885 4,057 22,425 2,996 164,998 Percent 6.91 3.91 8.67 1.04 .61 3.40 .45 24.99 GOnti.nued} f-' ....J O'\

PAGE 191

Table A.3. Year 1969 1970 1971 1972 :t.973 1974 1975 1976 1977 1978 1979. Total Percent (continued) Sweden Norway Finland --------------1,000 gallons 2,448 65 00 5,177 10,155 6,125 8,640 7,729 8,157 8,018 8,415 3,788 3,522 72,174 10.93 115 549 811 1,440 1,687 1,619 1,407 1,744 384 478 10,299 1. 56 23 105 513 513 865 351 348 145 106 00 2,969 .45 Other European 2 non-EEC of single 1,260 1,608 1,120 1,425 2,600 2,545 1,835 2,.08~ 2 I 7/.1 5 1,394 l,
PAGE 192

Table A.4. The United States of America exports of FCOJ and other orange juices by major regions and countries of destination, 1972-1979 FCOJ Other oranse juices Total EEC7 1 Euroepan 2 C d All ot~er 3 EEC7l European 2 C d All other 3 non-EEC3 ana a countries non-EEC3 ana a countries ----------------1,000 gallons of single strength juice equivalent--------------1972 8,027 6,996 18,037 1,531 5,243 453 5,903 4,021 50,205 1973 12,898 10,240 21,139 2,878 4,813 353 5,880 4,321 62,522 1974 7,794 9,905 25,665 5,791 4,601 376 5,933 4,041 64,106 1975 11,925 9,962 29,024 5,239 4,111 165 5,964 3,750 70,140 1976 21,678 9,567 31,508 6,343 5,596 206 6,345 4,175 85,418 1977 17,268 10,081 31,545 10,593 4,950 223 6,887 4,993 86,539 1978 7,165 4,152 24,924 6,187 3,448 126 6,664 5,371 58,037 1979 8,938 3,876 28,230 7,495 4,775 125 6,280 6,259 65,978 Total 95,686 64,779 210,072 46,057 37,537 2,027 49,856 36,931 542,945 Percent 17.63 11. 93 38.69 8.48 6.92 .37 9.18 6.80 100.00 1 rncludes West Germany, The Netherlands, the United and France. 2 rncludes Sweden, Norway and Finland. Kingdom, Belgium, Luxembourg, Denmark 3 rncludes Australia, Hong Kong, Japan, Malaysia and forty other countries. Source: Computed from United States Department of Commerce, Bureau of the Census, 1969-1979a. I-' -.I 00

PAGE 193

APPENDIX B IMPORTS OF ORANGE JUICE

PAGE 194

Table B.l. Leading orange juice importers, 1969-1978 West The United Belgium & a a Other EEC EEC Year a Netherlandsa Kingdoma a France Denmark t a,l total Germany Luxembourg coun ries ---------------1,000 gallons of single strength juice equivalent----------------1969 67,455 18,420 111,536 8,257 27,566 7,749 2,668 243,651 1970 89,813 21,850 104,295 8,638 36,332 10,290 7,749 278,967 1971 132,624 26,042 137,705 13,593 38,999 9,146 17,531 375,640 1972 161,587 30,869 141,897 16,133 43,319 11,433 16,006 421,244 f-' 1973 204,017 48,400 76,660 18,L120 68,598 16,133 8,384 440,612 00 0 1974 144,057 61,357 83,841 19 944 67,963 16,769 6,860 400,791 1975 185,470 62,755 75,840 28,964 56,657 18,293 2,922 430,901 1976 186,232 57,546 79,451 35,951 61,612 21,342 1,397 443,531 1977 159,047 95,657 78,380 38,110 58,436 23,755 1,270 454,655 1978 126,828 71,266 77,745 37f094 60,849 13,974 2,160 389,916 Total 1,457,130 494,162 967,350 225,104 520,331 148,884 66,947 3,879,908 Percent 24.10 8.18 16.00 3.72 8.61 2.46 1.11 64.18 (continued)

PAGE 195

Table B .1. (continued) Year Other Swedena Norwaya Finlanda EuropeaR 2 non-EEC European non-EEC Canadab USAc total Other 3 a, countries Total -------------------1,000 gallons of single strength Juice equivalent----------------1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 Total Percent 25,915 38,872 36,205 33,283 38,364 41,921 36,586 47,257 43,319 35,188 4,573 4,700 5,335 6,479 7,622 3,684 3,938 5,462 7,368 8,384 7,114 7,622 7,622 11,814 7,876 16,133 9,655 14,101 8,257 12,703 376,910 69,233 91,209 6.23 1.15 1.51 38,872 42,430 42,430 48,908 51,322 64,533 63,517 63,517 65,423 57,038 73,044 89,940 89,432 96,038 105,692 121,190 119,539 134,783 132,498 113,186 537,990 1,075,342 8.90 17.79 31,886 29,091 36,458 39,610 48,957 7,133 2,633 18,943 44,495 28,624 49,450 21,005 65,326 33,046 65,079 34,878 68,799 49,713 84,845 141,181 2,022 5.971 6,987 6,606 9,909 20,961 39,254 49,035 30,361 16,896 519,501 381,651 188,902 8.59 6.31 3.13 358,636 406,602 527,460 607,993 633,794 613,397 688,066 727,306 736,026 746,024 6,045,304 100.00 1 Includes Italy and Ireland. 2 Includes Austria, Bulgaria, East Germany, Iceland, Spain and Yugoslavia. 3 Includes Antigua, Australia, Barbados, Jamaica, South Africa, and Trinidad Tobago. Sources: Computed from aFood and Agriculture Organization of the United Nations, 1980; bstatistics Canada, 1969-1979; and cunited States Department of Commerce, Bureau of the Census, 1969-1979b.

PAGE 196

Table B.2. The United States of America imports of orange juice by country of origin, 1969-1979 Year 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 Total Percent Brazil Mexico Belize Other 1 countries Total ------------1,000 gallons of single strength juice equivalent------------3,802 1,308 15,413 34,534 13,327 14,115 28,214 31,091 33,749 129,891 148,703 454,147 84.41 697 577 1,230 5,726 5,848 5,160 3,315 1,473 13,790 9,860 7,376 55,052 10.23 1,711 746 1,415 1,196 1,349 364 00 00 182 00 00 6,963 1.29 923 2 885 3,309 8,100 1,366 1,517 2,314 1,992 1,430 332 21,900 4.07 7,133 2,633 18,943 44,495 28,624 21,005 33,046 34,878 49,713 141,181 156,411 538,062 100.00 1 Includes Argentina, Israel, Italy, Japan, Spain, South Africa and eleven other countries. Source: United States Department of Commerce, B ur eau of the Census,1969-1979b. I-' 00 N

PAGE 197

Table B.3. Canadian imports of orange juice by country of origin, 1969-1979 USA Brazil Mexico Belize Other 1 Total Year countries ----------1,000 gallons of single strength juice equivalent--------------1969 16,092 7,571 3,677 368 3,178 31,886 1970 19,560 5,189 3,565 237 540 29,091 1971 19,826 11,352 3,629 494 1,158 36,458 1972 22,299 12,336 2,321 1,174 1,480 39,610 1973 25,751 16,761 4,018 1,087 1,340 48,957 I-' co w 1974 32,577 7,479 5,358 2,076 1,960 49,450 1975 35,798 26,302 1,230 1,243 753 65,326 1976 39,394 22,160 1,673 1,835 17 65,079 1977 38,256 23,215 6,004 540 784 68,799 1978 35,270 45,869 3,406 00 300 84,845 1979 40,296 44,651 2,712 00 128 87,787 Total 325,119 222,885 38,592 9,054 11,638 607,288 Percent 53.54 36.70 6.35 1.49 1.92 100.00 1 Includes Argentina, Jamaica, Israel, South Africa and nineteen other countries. Source: Computed from Statistics Canada, 1969-1979.

PAGE 198

TableB.3.1. Canadian imports of FCOJ and other orange juices by countries of origin; 1972-1979 Year FCOJ Other orange juices Total USA Brazil All ot~er 1 Total USA All other 2 Total countries countries ----------------1,000 gallons of single strength juice equivalent-------------1972 17,136 12,296 4,135 33,567 5,162 881 6,043 39,610 1973 20,103 16,732 5,934 42,769 5,649 539 6,188 48,957 1974 23,339 7,417 7,913 38,669 9,238 1,543 10,781 49,450 1975 25,235 25,835 2,825 53,895 10,562 869 11,431 65,326 1976 30,673 21,919 3,316 55,908 8,720 451 9,170 65,079 1977 28,144 22,415 6,512 57,071 10,112 1,616 11,728 68,790 1978 25,479 45,585 3,581 74,645 9,790 410 10,200 84,845 1979 31,498 44,273 2,776 78,547 8,798 442 9,240 87,787 Total 201,607 196,472 36,992 435,071 68,031 6,754 74,782 509,853 Percent 29.54 38.53 7.26 85.33 13.34 1. 33 14.67 100.00 1 Includes Argentina, Belize, Israel, Jamaica, Mexico and nineteen other countries. 2 Includes same countries as in footnote 1 and small amount of juice recorded as from Brazil. Source: Computed from Statistics Canada,1969-1979. I--' co .i:::,.

PAGE 199

Table B.4. EEC imports of orange juice from Brazil and the United States of America, 1969-1979 EEC EEC7 1 Other EEC 2 EEC Brazil USA Total Brazil USA Total Brazil USA Total ----------------1,000 gallons of single strength juice equivalent--------------1969 1970 1971 1972 1973 17,677 30,758 50,397 63,039 107,549 6,760 11,547 10,473 13,264 17,711 24,437 42,305 60,870 76,303 125,260 1974 73,073 12,395 85,468 ~~75 120,217 16,036 136,253 1976 158,891 27,274 186,165 1977 127,416 22,217 149,633 1978 120,412 10,613 131,025 1979 155,464 13,712 169,176 Total 1,024,893 162,002 1,186,895 Percent 86.13 13.62 99.75 00 00 00 00 00 00 00 00 00 00 00 00 00 188 201 130 240 285 188 201 130 240 285 17,677 30,758 50,397 63,039 107,549 6,948 11,748 10,603 13,504 17,996 24,625 42,506 61,000 76,543 125,545 132 132 73,073 12,527 85,600 307 307 120,217 16,343 136,560 300 300 158,891 27,574 186,465 551 551 127,416 22,768 150,184 342 342 120,412 10,955 131,367 320 320 155,464 14,032 169,496 2,996 2,996 1,024,893 164,998 1,189,891 .25 .25 86.13 13.87 100.00 1 EEC7 as defined in this study includes West Germany, the United Kingdom, The Netherlands, Belgium, Luxembourg, Denmark and France. 2 rncludes Italy and Ireland. Source: Computed from Tables A. 2. and A. 3 ., Appendix A. f-' (X) u,

PAGE 200

Table B.5. European non-EEC imports of orange juice from Brazil and the United States of America, 1969-1979 Year 1969 1970 } 971 1972 1973 1974 1975 1976 1977 1973 1979 Total Percent 1 2 European non-EEC3 Other European non-EEC Brazil USA Total Brazil USA Total ----------------1,000 gallons of single strength juice 1,189 1,527 3,844 9,362 15,083 21,524 25,904 32,756 33,072 37,047 39,499 220,807 58.50 2,513 5,315 10,809 7,449 10,593 10,281 10,127 9,773 10,304 4,278 4,000 85,442 22.64 3,702 6,842 14,653 16,811 25,676 31,805 36,031 42,529 43,376 41,325 43,499 306,249 81.14 230 193 127 597 2,051 1,623 11,529 9,971 10,516 5,133 8,853 50,823 13.17 1,260 1,608 1,420 1,425 2,600 2,545 1,835 2,084 2,745 1,394 1,433 20,349 5.39 1,490 1,801 1,547 2,022 4,651 4,168 13,364 12,055 13,261 6,527 10,286 71,172 18.86 European non-EEC Brazil USA Total equivalent----------------1,419 1,720 3,971 9,959 17,134 23,147 34,433 42,727 43,588 42,180 48,352 271,630 71.97 3,773 6,923 12,229 8,874 13,193 12,826 11,962 11,857 13,049 5,672 5,433 105,791 28.03 5,192 8,643 16,200 18,833 30,327 35,973 49,395 54,584 56,637 47,852 53,785 377,421 100.00 1 ~uropean non-EEC3 as defined in this study includes Sweden, Norway and Finland. 2 rncludes ~ ustria, Bulgaria, East Germany, Iceland, Poland, Spain, Switzerland and Yugoslavia. Source: Computed from Tables A.2. and A.3.,Appendix A.

PAGE 201

APPENDIX C FLORIDA'S ORANGE JUICE

PAGE 202

Table C.l. Florida's frozen concentrated orange juice supply and demand, 1972-1979 Year Beginning Domestic Imports Total Domestic Exports Total Ending stocks supply supply demand demand stocks ----------------1,000 gallons of single strength juice equivalent---------------1972 98,155 564,541 22,939 685,635 530,468 29,126 559,594 126,041 1973 126,041 729,582 12,323 867,946 604,249 39,724 643,973 223,973 1974 223,973 696,370 10,216 930,559 664,446 41,621 706,067 224,492 1975 224,492 766,090 19,721 1,010,303 700,014 45,560 745,574 264,729 1976 264,729 758,537 26,194 1,049,460 790,932 50,476 841,408 208,052 I-' co co 1977 208,052 687,881 52,397 948,330 769,278 47,386 816,664 131,666 1978 131,666 697,740 123,221 952,627 747,395 37,083 784,478 168,149 1979 168,149 770,929 99,933 1,039,011 800,311 45,787 846,098 192,913 Total 1,445,257 5,671,670 366,944 7,483,871 5,607,093 336,763 5,943,856 1,540,015 Percent 19.31 75.79 4.90 100.00 74.92 4.50 79.42 20.58 Source: Computed from Florida Canners Association, 1971-1980.

PAGE 203

Table C.2. Florida's other orange juice (canned and chilled) supply and demand, 1972-1979 Year Beginning Domestic Total Domestic Exports Total Ending stocks supply supply demand demand stocks ---------------1,000 gallons of single strength Juice equivalent------------1972 12,878 154,344 167,222 145,124 6,765 151,889 15,333 1973 15,333 176,885 192,218 167,190 6,676 173,866 18,352 1974 18,352 174,081 192,433 169,285 7,449 176,734 15,699 1975 15,699 201,497 217,196 189,896 7,819 197,715 19,481 1976 19,481 210,078 229,559 203,346 8,243 211,589 17,970 I-' 00 '..0 1977 17,970 210,043 228,013 207,506 8,279 215,865 12,148 1978 12,148 234,731 246,879 226,217 7,937 234,154 12,725 1979 12,725 263,675 276,400 246,127 8,552 254,679 21,721 Total 124,586 1,625,334 1,749,920 1,554,771 61,720 1,616,491 133,429 Percent 7.12 92.88 100.00 88.85 3.53 92.38 7.62 Source: Computed from Florida Canners Associatio~ 1971-1980.

PAGE 204

Table C.3. Florida's total orange juice supply and demand, 1972-1979 Year Beginning Domestic Imports Total Domestic Exports Total Ending stocks supply supply demand demand stocks ----------------1,000 gallons of single strength Juice equivalent---------------1972 111,033 718,885 22,939 852,857 675,592 35,891 711,483 141,374 1973 141,374 906,467 12,323 1,060,164 771,439 46,400 817,839 242,325 1974 242,325 870,451 10,216 1,122,992 833,731 49,070 882,801 240,191 1975 240,191 967,587 19,721 1,227,499 889,910 53,379 943,289 284,210 1976 284,210 968,615 26,194 l,279;0J.9 994,278 58,719 1,052,997 226,022 I-' 0 1977 226,022 897,924 52,397 1,176,343 976,864 55,665 1,032,529 143,814 !978 143,814 932,471 123,221 1,199,506 973,612 45,020 1,018,632 180,874 1979 180,874 1,034,604 99,933 1,315,411 1,046,438 54,339 1,100,777 214,634 Total 1,569,843 7,297,004 366,944 9,233,791 7,161,864 398,483 7,560,347 1,673,444 Percent 17.00 79.03 3.97 100.00 77.56 4.32 81.88 18.12 Source: Computed from Tables C.l. and C.2.

PAGE 205

APPENDIX D REAL EXCHANGE RATE VALUES AND INDICES

PAGE 206

192 Real Exchange Rate Values and Indices Exchange rate is assumed to be an exogenous variable explaining variations in the import-export orange juice demand relations. This appendix describes the method of calculation for bimonthly indices of real exchange rate values of the USA dollar and the Brazilian cruzeiro against two groups of major European country currencies (EEC7 and European non-EEC3). The procedure is the same one used by United States Board of Governors of the Federal Reserve System described in the Federal Reserve Bulletin, volume 64, August 1978, page 700. Real exchange rate values were obtained by deflating the corresponding nominal values by available price indices (1975=100); i.e., the wholesale price indices in the case of the USA and Brazil, and the consumer price indices in the case of the other countries. Export and import data, as well as the consumer price indices for Canada, EEC7 and European non-EEC3 countries, are those reported by the International Monetary Fund, 1970-1980. Nominal exchange rate values and wholesale price indices for the USA and Brazil are from the United States Board of Governors of the Federal Reserve System (1972-1980) (the USA exchange rates), United States Department of Commerce, Bureau of Economic Analysis (1975-1980) (the USA wholesale price index), the Funda9io Getdlio Vargas (1972-1980) (Brazil exchange rate (cruzeiros per USA dollar), and Brazil wholesale price index).

PAGE 207

193 A real exchange rate value index is defined, in general, as Rit is the rate value n 100 II R~i where, 1 1.t 1.= base period (February-March of the .th l. country currency of 1973) exchange at time t, and w. is l. the weight of the .th country. The USA exchange rate values l. are expressed in cents per unit of foreign currency. Brazil's exchange rate values are expressed in cruzeiros per unit of foreign currency and, except the rate of cruzeiros per USA dollar which is reported by Funda9ao Getulio Vargas, all other relations were derived from the series of the USA exchange rate values and Brazil's exchange rate values against the USA dollar. Weights and base period exchange rate values and indices of exchange rates in the case of the two groups of European countries are shown in Tables D.2. through D.8. for Brazil and the U.S.A.

PAGE 208

Table D.l. Real exchange rate values for the base period (February-March of 1973) and weight by country Country-currency name EEC7 countries West Germany-deutsche mark The Netherlands-guilder The United Kingdom-pound Belgium/Luxembourg-franc Denmark-krone France-franc European non-EEC3 countries Sweden-krona Norway-krone Finland-markka 40.62320 36.94900 225.68700 2.65551 16.24170 22.44340 24.94670 17.86910 23.84450 Base period 2 Brazil 3.122420 2.840110 17.348000 .204114 1.248390 1. 725090 1.917570 1. 373480 1.832810 1 cents per unit of foreign currency. 2 cruzeiros per unit of foreign currency. are average trade value shares (the USA dollar value of exports and imports over value trade) of each country in the group for five years (1972-1976). Weight 3 1.00000 .32256 .13402 .18619 .11780 .03644 .20299 1. 00000 .54226 .25873 .19901 3 h Weig ts total f-' i.O .i:::.

PAGE 209

Table D.2. Bimonthly real exchange rate values of the Brazilian cruzeiro against the USA dollar, 1972-1979 Years Dec./Jan. Feb./Mar. Apr./May June/July Aug./Sept. Oct./Nov. ---------------------------cruzeiros per USA dollar--------------------------1971/72 7.83404 7.81361 7.78804 7.66094 7.63494 1972/73 7.80436 7.68680 7.75189 7.82969 7.99783 7.70315 1973/74 7.77196 7.86933 7.41221 7.65313 8.19957 8.34040 1974/75 8.25952 8.13836 8.25035 8.17524 8.05437 8.02680 1975/76 8.01980 7.83889 8.05101 7.97060 7.60091 7.61769 1976/77 7.69496 7.63197 7.48677 7.64978 7.82986 7.79201 1977/78 7.85064 7.77026 7.70497 7.67583 7.66363 7.63256 1978/79 7.92505 8.03409 Sources: Computed from Funda9ao Getulio Vargas,1972-1980 (nominal exchange rate values and wholesale price index in Brazil); and United States Department of Commerce, Bureau of Economic Analysis, 1975-1980 (wholesale price index in the USA). 1--' I.D Ul

PAGE 210

Table D.3. Bimonthly real exchange rate values of the USA dollar against the Canadian dollar, 1972-1979 Years Dec./Jan. Feb./Mar. Apr./May June/July Aug./Sept. Oct./Nov. ----------------------------cents per Canadian dollar--------------------------1971/72 110.560 111.870 112.643 113.487 113.453 1972/73 110.293 107.670 105.845 104.889 102.074 105.171 1973/74 103.585 103.389 104.207 102.890 97.991 97.353 1974/75 98.041 98.796 97.109 97.617 97.760 99.591 197 5/76 101.078 103.749 104.413 105.242 105.196 105.716 1976/77 101.829 99.402 97.971 98.593 98.265 96.243 1977/78 96.633 94.410 92.943 94.362 92.009 89.655 1978/79 88.515 87.705 Source: Computed from United States Board of Governors of the Federal Reserve System, 1972-1980 (nominal exchange rates); United States Department of Commerce, Bureau of Economic Analysis, 1975-1980 (wholesale price index in the USA); and International Monetary Fund, 1970-1980 (consumer price index in Canada). t-' \0 O'I

PAGE 211

Table D.4. Bimonthly real exchange rate values of the Brazilian cruzP-iro against the Canadian dollar, 1972-1979 Years Dec./Jan. Feb./Mar. Apr./May June/July Aug./Sept. Oct./Nov. ------------------------cruzeiros per Canadian dollar------------------------1971/72 8.66123 8.74111 8.77264 8.69420 8.66198 1972/73 8.60768 8.27663 8.20443 8.21140 8.16269 8.10111 1973/74 8.05012 8.13592 7.72380 7.86957 8.03483 8.11953 1974/75 8.09774 8.04035 8.01179 7.98042 7.87394 7.99405 1975/76 8.10607 8.13260 8.40620 8.38847 7.99571 8.05287 1976/77 7.83560 7.58719 7.33482 7.54236 7.69405 7.49939 1977/78 7.58639 7.33605 7.16132 7.24307 7.05113 6.84284 1978/79 7.01443 7.04579 Source: Derived from Tables D.2. (RxBSt) and D.3. (RxSCt). I-' -..J

PAGE 212

TabJ.e D.5. Years 1971/72 1972/73 1973/74 1974/75 1975/76 1976/77 1977/78 1978/79 Bimonthly real exchange rate value index (February-March of 1973=100 of the Brazilian cruzeiro against the EEc71 country currencies, 1972-1979 Dec./Jan. Feb./Mar. Apr./May June/July Aug./Sept. Oct./Nov. -----------------------------------percent-----------------------------------101.731 101.452 102.360 104.543 104.518 104.201 100.000 97.888 90.050 91 976 93.962 103.716 102.944 104.148 102.851 102.248 98.463 93.455 89.192 87.335 89.467 97.967 97.919 98.087 98.997 97.736 100.989 104.041 102.049 99.280 100.482 101.610 97.270 94.637 93.321 88.792 88.127 90.312 90.456 87.095 84.695 81.952 81.691 1 Includes West Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark and France. Source: Derived from United States Board of Governors of the Federal Reserve System, 1972-1980 (the USA nominal exchange rate values); United States Department of Commerce, Bureau of Economic Analysis,1975-1980 (wholesale price index in the USA); International Monetary Fund (consumer price index and population of European countries used to derive weighted average consumer price index for the EEC7), Table D.2. (RxBSt) and Table D.l. (base period values and weight). I-' CX)

PAGE 213

Table D.6. Bimonthly real exchange rate value index (February-March of 1973=100) of the Brazilian cruzeiro against the European non-EEC31 country currencies, 19721979 Years Dec./Jan. Feb./Mar. Apr./May June/July Aug./Sept. Oct./Nov. -----------------------------------Percent------------------------------------1971/72 102.764 101.938 101.240 102.743 103.485 1972/73 102.468 100.000 99.614 92.223 93.924 96.186 1973/74 104.125 102.662 104.775 103.925 102.489 98.081 1974/75 92.534 88.629 87.258 89.078 97.416 96.879 1975/76 95.741 96.133 93.287 94.765 97.274 93.552 1976/77 91.246 92.353 96.530 92.370 92.963 95.216 1977/78 91.915 92.456 95.324 95.688 92.567 90.813 1978/79 88.818 89.117 1 rncludes Sweden, Norway and Finland. Sources: Derived from United States Board of Governors of the Federal Reserve System, 1972-1980 (the USA nominal exchange rate values); United States Department of Commerce, Bureau of Economic Analysis, 1975-1980 (wholesale price index in the USA); and International Monetary Fund,1970-1980 (consumer price index and population of European countries used to derive a weighted average consumer price index for the European non-EEC3); Table D.2. (RxBSt); and Table D.l. (base period values and weights). f-' 1.0 1.0

PAGE 214

Table D.7. Bimonthly real exchange rate value index (February-March of 1973=100) of the USA dollar against the EEC7 1 country currencies, 1972-1979 Years Dec./Jan. Feb./Mar. Apr./May June/July Aug./Sept. Oct./Nov. -----------------------------------Percent-----------------------------------1971/72 103.682 103.130 103.716 104.196 103.817 1972/73 105.799 100.000 98.723 91.686 95.688 94.180 1973/74 104.841 105.378 100.426 102.364 109.074 106.847 1974/75 100.417 94.435 93.742 95.146 102.658 102.255 1975/76 102.338 100.962 102.373 104.727 102.881 101.138 1976/77 99.389 99.769 98.970 96.814 96.402 94.600 1977/78 90.686 89.086 90.528 90.331 86.836 84.105 1978/79 84.499 85.387 1 Includes West Germany, The Netherlands, the United Kingdom, Belgium, Luxembourg, Denmark and France. Sources: Derived from United States Board of Governors of the Federal Reserve System 19721980, (the USA nominal exchange rate values); United States Department of Commerce, Bureau of Economic Analysis, 1975-1980 (wholesale price index in the USA); and International Monetary Fund (consumer price index and population of European countries used to derive a weighted average consumer price index for the EEC?); and Table D.l. (base period values and weights). N 0 0

PAGE 215

Table D.8. Bimonthly real exchange rate value index (February-March of 1973=100) of the USA dollar against the European non-EEC3 1 country currencies, 1972-1979 Years Dec./Jan. Feb./Mar. Apr./May June/July Aug./Sept. Oct./Nov. -----------------------------------Percent-----------------------------------1971/72 104.735 103.625 102.577 102.401 102.790 1972/73 104.039 100.000 100.461 93.904 97.718 96.396 1973/74 105.252 105.088 101.034 103.424 109.329 106.432 1974/75 99.427 93.838 93.733 95.101 102.597 102.139 1975/76 102.074 100.645 101.865 104.019 102.145 100.204 1976/77 98.498 98.875 98.430 96.282 96.217 94.803 1977/78 91.021 89.543 91.051 90.874 92.292 90.178 1978/79 91.577 93.149 1 Includes Sweden, Norway and Finland. Source: Derived from United States Board of Governors of th~ Federal Reserve System 19721980, (the USA nominal exchange rate values); United States Department of Commerce, Bureau of Economic Analysi~ 1975-1980 (wholesale price index in the USA); and International Monetary Fund, 1970-1980 (consumer price index and population of European countries used to construct a weighted average consumer price index for the European non-EEC3); and Table D.l. (base period values and weights). N 0

PAGE 216

APPENDIX E VARIABLE DEFINITIONS AND DATA SOURCES

PAGE 217

203 Variable Definitions and Data Sources Endogenous variables are listed according to their equational ordering formulation in the model. Predetermined variables are listed according to the number of capital letters (five through one) and type of transaction or type of variable (import, export, retail, wholesale exchange rate, advertising, disposable income, and orange production). Unless otherwise specified, non-monetary variables are expressed in single strength or reconstituted juice equivalent gallons per thousand persons. Monetary variables are expressed in the USA dollar in terms of 1975 prices. Prices are expressed in dollars per single strength or reconstituted juice equivalent gallon. Exchange rates are expressed in units of the exporting country currency (Brazilian cruzeiros or cents) per unit of foreign currency, or as an index. Advertising variables are expressed in dollars per thousand persons. Disposable income variables are expressed in thousand dollars per capita. The variable subscript t stands for time period (December-January bimonth of 1971/72 to February-March bimonth of 1979). Endogenous variables RQFSt = retail quantity of FCOJ in the USA. Source: computed from Florida Department of Citrus and A. C. Nielsen Company, 1972-1980.

PAGE 218

204 RPFSt = retail price of FCOJ in the USA. Source: computed from Florida Department of Citrus and A. C. Nielsen Company, 1972-1980 (nominal values), and United States Department of Commerce, Bureau of Economic Analysis, 1975-1980 (consumer price index). WQFFlt = wholesale quantity of FCOJ in Florida. Source: computed from Florida Canners Association, 1971-1980. WPFFlt = Wholesale price of FCOJ in Florida. Source: computed from Florida Deaprtment of Citrus, Economic Research Department, 1972-1980 (nominal values), and United States Department of Commerce, Bureau of Economic Analysis, 1975-1980 (producer price index). WSkFSt = wholesale stock of FCOJ in the USA, expressed in bimonths of Florida supplies. Source: derived (see Equation 5 in Chapter III). WQkFSt = wholesale quantity in stock of FCOJ in the USA. Source: derived (see Equation 6 in Chapter III). MQFSBt = import quantity of FCOJ by the USA from Brazil Source: computed from United States Department of Commerce, Bureau of the Census, 1969-1979b. SpBt = price spread between Florida's FCOJ wholesale price (WPFFlt) and the USA FCOJ import price from Brazil (MPFSBt). Source: derived (see Equation 8 in Chapter III). MPFSBt = import price of FCOJ by the USA from Brazil. Source: computed from United States Department of Commerce, Bureau of the Census, 1969-1979b (no~inal

PAGE 219

205 values), and United States Department of Commerce, Bureau of Economic Analysis, 1975-1980 (producer price index) RQOSt = retail quantity of other orange juices in the USA. Source: same as RQFSt. RPOSt = retail price of other orange juices in the USA. Source: same as RPFSt. RpOSt = retail proportion of other orange juices marketed in cardboard containers in the USA. Source: derived (see Equation 12 of Chapter III). WQOFlt = wholesale quantity of other juices in Florida. Source: same as WQFFlt. WPOFlt = wholesale price of other orange juices in Florida. Source: same as WPFFlt. WSkOSt = wholesale stock of other orange juices in the USA, expressed in bimonths of Florida supplies. Source: derived (see Equation 15 in Chapter III). WQkOSt = wholesale quantity in stock of other orange juices in the USA. Source: derived (see Equation 16 in Chapter III). XQFCBt = export quantity of FCOJ to Canada from Brazil. Source: computed from Brazil, Carteira de Comercio Exterior Banco do Brasil S.A. (CACEX), 1972-1979. XPFCBt = export price of FCOJ to Canada from Brazil. Source: computed from Brazil, Carteira de Comercio Exter ior do Banco do Brasil S.A. (CACEX), 1972-1979 (nominal values), and International Monetary Fund, 1970-1980 (consumer price index).

PAGE 220

206 XQFCSt = export quantity of FCOJ to Canada from the USA. Source: computed from United States Department of Commerce, Bureau of the Census, 1969-1979a. XPFCSt = export price of FCOJ to Canada from the USA. Source: computed from United States Department of Commerce, Bureau of the Census, 1969-1979a (nominal values}, and International Monetary Fund, 1970-1980 (consumer price index). XQFEBt = export quantity of FCOJ to the EEC? countries from Brazil. Source: same as XQFCBt. XPFEBt = export price of FCOJ to the EEC7 countries from Brazil. Source: same as XPFCBt. XQFESt = export quantity of FCOJ to the EEC? countries from Brazil. Source: same as XPFCBt. XQFESt = export quantity of FCOJ to the EEC7 countries from the USA. Source: same as XQFCSt. XPFESt = export price of FCOJ to the EEC? countries from the USA. Source: same as XPFCSt. XQFNBt = export quantity of FCOJ to the European non-EEC3 countries from Brazil. Source: same as XQFCBt. XPFNBt = export price of FCOJ to the European non-EEC3 countries from Brazil. Source: same as XPFCBt. XQFNSt = export quantity of FCOJ to the European non-EEC3 countries from the USA. Source: same as XQFCSt. XPFNSt = export price of FCOJ to the European non-EEC3 countries from the USA. Source: same as XPFCBt.

PAGE 221

207 Predetermined variables MQFSOct = import quantity of FCOJ by the USA from other countries. Source: computed from United States Department of Commerce, Bureau of the Census, 1969-1979b. MPFSOct = import price of FCOJ by the USA from other countries. Source: same as MPFSBt. MPFCOct = import price of FCOJ by Canada from other countries. Source: computed from Statistics Canada, 1969-1979 (nominal values) and International Monetary Fund, 1970-1980 (consumer price index). XQFOcSt = export quantity of FCOJ to other countries from the USA. Source: same as XQFCSt. XQOCSt = export quantity of other orange juices to Canada from the USA. Source: same as XQFCSt. XPOCSt = export price of other orange juices to Canada from the USA. Source: same as XPFCSt. RPShSt = retail price of frozen orange-flavored synthetics and drinks in the USA. Source: same as RPFSt. RQObSt = retail quantity of other orange juices in cardboard containers in the USA. Source: same as RQFSt. WPFFlt-l = WPFFlt lag by one period. WPOFlt-l = WPOFlt lag by one period. WSkFSt-l = WSkFSt lag by one period. WSkOSt-l = WSkOSt lag by one period. WQkFSt-l = WQkFSt lag by one period. WQkOSt-l = WQkOSt lag by one period.

PAGE 222

208 WQsFFl = wholesale quantity supplied of FCOJ in Florida. Source: same as WQFFlt. WQsOFlt = wholesale quantity supplied of other orange juices in Florida. Source: same as WQFFlt. RxBSt = exchange rate of the Brazilian cruzeiro against the USA dollar. Source: computed from Funda9ao Getulio Vargas, 1972-1980 (nominal exchange rate values and wholesale price index in Brazil), and United States Department of Commerce, Bureau of Economic Analysis, 1975-1980 (wholesale price index in the USA). RxBCt = exchange rate of the Brazilian cruzeiro against the Canadian dollar. Source: derived, given the values of RxBSt and RxSCt. RxBEt' RxBNt' RxSEt' RxSNt = exchange rate value indices (February-March of 1973=100) of the Brazilian cruzeiro {B) and the USA dollar (S) against currencies of the EEC7 (E) and the European non-EEC3 (N) countries, respectively (for more detail see Appendix D). Source: derived from United States Board of Governors of the Federal Reserve System, 1972-1980 (nominal exchange rate values), United States Department of Commerce, Bureau of Economic Analysis, 1975-1980 (wholesale price index in the USA), International Monetary Fund, 1970-1980 (consumer price indices and population of European countries used to construct weighted

PAGE 223

RxSCt = 209 average consumer price indices for EEC7 and European non-EEC3 groups of countries), and computed values of RxBSt. exchange rate of the USA dollar against the Canadian dollar. Source: computed from United States Board of Governors of the Federal Reserve System, 19721980 (nominal exchange rate values), United States Department of Commerce, Bureau of Economic Analysis, 1975-1980 (wholesale price index in Canada). ACt, ACt-l = current and lagged advertising expenditures of the Florida Department of Citrus in Canada. Source: computed from Florida Department of Citrus, 1980 (nominal values), and International Monetary Fund, 1970-1980 (consumer price index in Canada). AEt, ANt = annual advertising expenditures of the Three Party Program in EEC7 and the European non-EEC3 countries, respectively. Source: computed from Florida Department of Citrus, 1980 (nominal values), and International Monetary Fund, 1970-1980 (consumer price indices and population of European countries used to construct weighted average consumer price indices for EEC7 and European non-EEC3 countries). ASt, ASt-l = current and lagged advertising expenditures (generic and brand) of the Florida Department of Citrus in the USA. Source: computed from Florida Department of Citrus, 1980 (nominal values), and United States Department of Commerce, Bureau of Economic Analysis, 1975-1980 (consumer price index).

PAGE 224

210 ICt = quarterly disposable income per capita in Canada, seasonally adjusted at annual rates. Source: computed from Statistics Canada, 1979 (nominal values), United States Board of Governors of the Federal Reserve System, 1972-1980 (exchange rates), and International Monetary Fund, 1970-1980 (consumer price index and population in Canada). ISt = quarterly disposable income per capita in the USA, seasonally adjusted at annual rates. Source: computed from United States Department of Commerce, Bureau of Economic Analysis, 1975-1980 (nominal values and consumer price index) and United States Department of Commerce, Bureau of the Census, 1975-1980 (population). PdSpt' PdSpt-G = current and lagged annual estimates of orange production in the state of Sao Paulo (Brazil), expressed in billion boxes of 90 pounds (approximately 40.77 kilograms). Source: computed from United States Department of Agriculture, Foreign Agricultural Service, 1980a and 1980c. PdSt = estimates of orange production in the USA, expressed in billion boxes of 90 pounds (approximately 40.77 kilograms). Source: computed from Florida Crop and Livestorck Reporting Service, 1972-1980. PlCt, PlEt, PlNt, PlOct, PlSt = population in Canada, the EEC7, the European non-EEC3, other countries and the USA, expressed in million persons. Source: Inter national Monetary Fund, 1970-1980 (Canada and European

PAGE 225

211 countries} and United States Department of Commerce, Bureau of the Census, 1975-1980. Dt = dummy variable to account for the effects of the USA (Florida} freeze in 1977. Dt is equal to one for observations from February-March bimonth of 1977 on, and zero otherwise. Tt = linear time trend to account for time-related changes in Florida's wholesale price equations. or as a proxy variable for income per capita in the export demand equations for the European countries. Zit= five (i=l, ,5} dummy variables to account for seasonal effects on demand. Zlt is equal to one for December-January bimonths, and zero otherwise. Similarly, Z2t through zst is equal to one for February-March, April-May, June-July and August September bimonths, respectively, and zero otherwise.

PAGE 226

APPENDIX F MEANS, STANDARD DEVIATIONS AND COEFFICIENTS OF VARIATION FOR VARIABLES USED IN THE MODEL

PAGE 227

213 Means, standard deviations and coefficients of variation for variables used in the model Variables 1 2 Standard Coefficient of 3 Mean deviation variation (percent) Endogenous variables RQFSt 362.0988 37.2812 10.30 RPFSt 1.5336 .1945 12.68 WQFFlt 539.9482 86.1218 15.95 WPFFlt .9964 .1663 16.69 WSkFSt 2.7021 1.0110 37.42 WQkFSt 1424.6138 490.5815 34.44 MQFSBt 35.1260 47.7737 136.01 SpBt .5428 .1019 18.77 MPFSBt .4536 .1364 30.07 RQOSt 135,4163 32.3195 23.87 RPOSt 2.0397 .1625 7.97 RpOSt .4740 .0997 21.03 WQOFlt 147.5382 20.1639 13.67 WPOFlt 1. 3530 .1458 10.78 WSkOSt 1.1602 .4607 39.71 WQkOSt 168.9039 61.6674 36.51 XQFCBt 163.1442 109.1834 66.92 XPFCBt .4778 .1261 26.39 XQFCSt 191. 9097 43.7903 22.82 XPFCSt 1.1073 .1521 13.74 (continued)

PAGE 228

214 Means, ... (continued) Variables 1 Mean 2 Standard Coefficient of 3 deviation variation (percent) XQFEBt 97.0726 48.4249 49.89 XPFEBt .4751 .1077 22.67 XQFESt 10.6397 7.2904 68.52 XPFESt .7674 .2290 29.84 XQFNBt 252.9446 161. 5484 63.87 XPFNBt .4672 .0885 18.94 XQFNSt 83.9191 40.8927 48.73 XPFNSt .8913 .1305 14.64 Predetermined variables MQFSOct 7.1593 5.8655 81. 93 MPFSOct .5306 .1214 22.88 MPFCOct .5943 .1288 21.67 XQFOcSt 4.8153 2.7459 57.02 XQOCSt 46.1726 8.9247 19.33 XPOCSt 1. 2712 .1281 10.08 RPShSt 1. 5408 .0570 3.79 RQObSt 67.2895 28.8092 42.81 WPFFlt-l .9979 .1680 16.84 WPOFlt-l 1.3606 .1525 11.21 WSkFSt-l 2.6881 1.0126 37.67 WSkOSt-l 1.1670 .4598 39.40 WQkFSt-l 1409.7772 494.3198 35.06 (continued)

PAGE 229

215 Means, ... (continued) Variables 1 Mean 2 Standard Coefficient of 3 deviation variation (percent) WQkOSt-l 168. 5535 61.6873 36.60 WQsFFlt 555.7976 418.2683 75.26 WQsOFlt 153.5056 60.1723 45.07 RxBSt 7.8396 .2198 2.80 RxBCt 7.9101 .4940 6.25 RxSCt 100.9441 6.4257 6.37 RXBEt 96.2446 6.6291 6.89 RxSEt 98.1033 6.4797 6.61 RxBNt 95.9477 4.9062 5.11 RxSNt 97.8104 4.8607 4.97 ASt 9.2798 2.8285 30.48 ASt-1 9.4226 3.0982 32.88 ACt 3.4744 1.9475 56.05 ACt-1 3.5049 1.9786 56.45 AEt 2.8621 1.0075 35.20 ANt 40.4665 14.8154 36.61 ISt 5.2174 .1862 3.57 ICt 4.6990 .2971 6.32 PdSt .2271 .0187 8.22 PdSt-l .2267 .0193 8.51 PdSpt .0934 .0293 31.37 PdSpt_ 6 .0790 .0214 27.09 (continued)

PAGE 230

216 Means, ... (continued) Variables 1 Mean 2 Standard Coefficient of 3 deviation variation (percent) PlCt 22.7586 .5813 2.55 PlEt 197.5148 3.2573 1.65 PlNt 16.9168 .1336 .79 4 PlOct 197.5148 3.2573 1. 65 PlSt 214.0112 3.3803 1. 58 lEndogenous variables are listed according to their equational ordering formulation. Predetermined variables are listed according to the number of capital letters (five, four and two) and type of transaction or type of variables (import, export, retail, wholesale, exchange rate, adver tising, disposable income and orange production). 2 Most of the non-monetary variables are expressed in single strength or reconstituted juice equivalent gallons per thousand persons. WSkFSt, WSkOSt and respective lags are expressed in number of bimonths of Florida's supplies; RpOSt is a proportion of RQObSt to RQOSt; orange production variables are expressed in billion boxes of 90 pounds (approximately 40.77 kilograms); and population variables are expressed in million persons. Monetary variables are expressed in the USA dollar in terms of 1975 prices. Price variables are expressed in dollars per single strength or reconstituted juice equivalent gallon. Exchange rates are expressed in units of the exporting country currency (Brazilian cruzeiros or cents) per unit of foreign currency, or as an index. Advertising variables are also expressed in dollars per thousand persons. Disposable income variables are expressed in thousand dollars per capita. 3coefficient of variation=l00 (standard deviation)/mean. 4 Population of other countries was approximated by the EEC7 population.

PAGE 231

APPENDIX G STRUCTURAL-FORM ESTIMATES OF ELASTICITIES AND FLEXIBILITIES AT THE VARIABLE MEANS

PAGE 232

218 Structural-form estimates of elisticities and flexibilities evaluated at the variable means Dependent variable (Equation #1) Elasticity Flexibility ---------Percent----------RQFSt (Equation 1) Own-price (RPFSt) Cross-price (RPOSt) Cross-price (RPShSt) Income (ISt) Current advertising (ASt) Lagged advertising (ASt_ 1 ) RPFSt (Equation 2) Current FCOJ wholesale price transmission (WPFFlt) Lagged FCOJ wholesale price transmission (WPFF1t_ 1 ) WQFFlt (Equation 3) Own-price (WPFFlt) Cross-price (WPOFli) FCOJ retail transmission (RQFSt) Other orange juices retail transmission (RQOSt) WPFFlt (Equation 4) Current FCOJ stock (WSkFSt) Lagged FCOJ stock (WSkFSt-1) Current orange production estimates (PdSt) Lagged orange production estimates (PdSt-1) MQFSBt (Equation 7) Price spread (SpBt) Other orange juices wholesale transmission (WQOFlt) -.4180 -.6323 .5357 .8948 .0359 .0234 .5328 .2642 -.6618 .6141 .3901 .3621 1. 6607 5.0697 -.0374 -.0742 -.7959 -.4276 (continued)

PAGE 233

219 Structural-form, ... (continued) Dependent variable (Equation #1) MPFSBt (Equation 9) Current FCOJ wholesale price transmission (WPFFlt) Lagged FCOJ wholesale price transmission (WPFFlt-1) Exchange rate transmission (RxBSt) Lagged orange production estimates in Sao Paulo (PdSPt-6) RQOSt (Equation 10) Own-price (RPOSt) Cross-price (RPFSt) Cross-price (RPShSt) Income (ISt) Current advertising (ASt) Lagged advertising (ASt_ 1 ) RPOSt (Equation 11) Current other orange juices wholesale price transmission (WPOFlt) Lagged other orange juices wholesale price transmission (WPOFlt-1) Retail proportion of other orange juices in cardboard containers (RpOSt) WQOFlt (Equation 13) Own-price (WPFOlt) Other orange juices retail transmission (RQOSt) WPOFlt (Equation 14) Current other orange juices stocks (WSxOSt) Elasticity Flexibility ----------Percent---------.5123 1.0201 -.1642 -2.0587 1.1466 1.5593 3.4629 .0203 .0264 .0960 .5561 -.2284 .5352 1341 -.0521 -.0179 (continued)

PAGE 234

220 Structural-form, ... (continued) Dependent variable (Equation #) Current orange production estimates (PdSt) Lagged orange production estimates (PdSt-1) XQFCBt (Equation 17) Own-price export (XPFCBt) Cross-price export (XPFCSt) Cross-price export (XPOCSt) Income (ICt) Current advertising (ACt) Lagged advertising (ACt_ 1 ) XPFCBt (Equation 18) USA export price transmission (XPFCSt) Exchange rate transmission (RxBSt) Lagged orange production estimates in Sao Paulo (PdSPt-6) XQFCSt (Equation 19) Own-price export (XPFCSt) Cross-price export (XPFCBt) Cross-price export (XPOCSt) Income (ICt) Current Florida's advertising (ACt) Lagged Florida's advertising (ACt-1) XPFCSt (Equation 20) Florida's wholesale price transmission (WPFFlt) USA import price from Brazil transmission (MPFSBt) Exchange rate transmission (RxSCt) Elasticity Flexibility ----------Percent----------1.1884 2.8003 3.9404 3.8719 .1368 .0659 1. 6429 -.5693 -.4616 -.0409 .4303 2.0365 -.1131 .1626 .4465 .1857 -.5287 .0534 -.7400 -.0475 (continued)

PAGE 235

221 Structural-form, ... (continued) Dependent variable (Equation #) XQFEBt (Equation 21) Own-price export (XPFEBt) Cross-price export (XPFESt) Current Florida's advertising (AEt) XPFEBt (Equation 22) USA export price transmission (XPFESt) Exchange rate transmission (RxBSt) Lagged orange production estimates in Sao Paulo (PdSPt-6) XQFESt (Equation 23) Own-price export (XPFESt) Cross-price export (XPFEBt) Current Florida's advertising (AEt) XPFESt (Equation 24) Florida's wholesale price transmission (WPFFlt) USA import price from Brazil transmission (MPFSBt) Exchange rate transmission (RxSEt) XQFNBt (Equation 25) Own-price export (XPFNBt) Cross-price export (XPFNSt) Current Florida's advertising (ANt) Elasticity Flexibility ----------Percent----------.7237 -.0734 .5292 .5610 -2.0659 -1.7351 -.0883 .5029 .3298 .4362 -1.3934 -.3374 -.0939 -.9574 -.0915 (continued)

PAGE 236

222 Structural-form, ... (continued) Dependent variable (Equation #) XPFNBt (Equation 26) USA export price transmission (XPFNSt) Exchange rate transmission (RxBSt) Lagged orange production estimates in Sao Paulo (PdSPt-6) XQFNSt (Equation 27) Own-price export (XPFNSt) Cross-price export (XPFNBt) Current Florida's advertising (ANt) XPFNSt (Equation 28) Florida's wholesale price transmission (WPFFlt) USA import price from Brazil transmission (MPFSBt) Exchange rate transmission (RxSNt) Elasticity Flexibility ----------Percent---------.7969 -1. 3911 -2.4662 .3747 1.2377 .4349 .1132 .8779 .0678 1 see Appendix F for variable means.

PAGE 237

APPENDIX H ACTUAL AND REDUCED-FORM ESTIMATED VALUES OF ENDOGENOUS VARIABLES (Graphs)

PAGE 238

Gallons per 1,000 persons 500 460 420 380 340 300 260 Oct.-Nov. 1972 Oct -Nov 1973 224 Oct -Nov 1974 Oct.-Nov 1975 Oct -Nov. 1976 Oct -Nov 1977 Oct -Nov. 1978 Figure H 1 Actual and reduced-form estimated values of retail quantities of FCOJ in the USA (ROFSt) April-May/1972 to February-March/1979. Note: FCOJ quantities are expressed in gallons of s i ngle strength juice equivalent Dollars per gallon 2 4 2 2 2.0 1 8 1 6 1.4 1 2 1 0 -, Oct.-Nov 1972 Estimated ,,..,, / Cct -Nov. 1973 ,_ Oct.-Nov 1974 __ Oct.-Nov. 1975 Oct.-Nov. 1976 Oct -Nov. 1977 _,,-Oct.-Nov. 1978 Figure H 2. Actual and reduced-form estimated values of retail prices (1975=100) of FCOJ in the USA (RPFSt) April-May/1972 to February March/1979 .;

PAGE 239

225 Gallons per 1,000 persons 1000 900 800 700 600 500 400 300 Figure H.3. Dollars per gallon 1.4 1 2 1.0 0 8 0 6 Figure H.4. Oct.-Nov. 1972 Oct.-Nov. 1973 Oct.-Nov. 1974 Actual~ Oct.-Nov. 1975 Oct.-Nov. 1976 Oct.-Nov. 1977 Oct.-Nov. 1978 Actual and reduced-form estimated values of wholesale quantities of FCOJ in the Florida (WOFFlt}. April-May/1972 to February-March/1979 Note : FCOJ quantities are expressed in gallons of single strength juice equivalent. Oct .Nov. 1972 Estimated .,,/ Oct -Nov. 1973 ,_ ,_ Oct.-Nov. 1974 Oct.-Nov. 1975 Oct.-Nov. 1976 'Actual Oct.-Nov 1977 Oct .Nov. 1978 Actual and reduced-form estimated values of wholesale prices (1975=100) of FCOJ in Florida (WPFFlt), April-May/1972 to February-March/1979.

PAGE 240

226 Gallons per 1,000 persons 240 210 180 150 120 90 60 30 0 Figure H.5 Dollars per gallon 0 75 0.65 0.55 0.45 0.35 0 25 Figure H.6. Oct.-Nov 1972 Oct -Nov. 1973 Oct.-Nov. 1974 Oct.-Nov. 1975 Oct.-Nov. 1976 Oct.-Nov. 1977 I ,, ,, I \ J \ / Oct.-Nov. 1978 Actual and reduced-form estimated values of import quantities of FCOJ by the USA from Brazil (MQFSBtl. April-May/1972 to February-March/1979. Note: FCOJ quantities are expressed in gallons of single strength juice equivalent. Oct -Nov. 1972 ........ Actual Oct.-Nov. 1973 Oct -Nov. 1974 Estimated .,,.-,J Oct.-Nov. 1975 -'-\ \ Oct.-Nov 1976 Oct.-Nov. 1977 Oct.-Nov. 1978 Actual and reduced-form estimated values of import price (1975=100) of FCOJ by the USA from Brazil (MPFSBt), April-May/1972 to February-March/1979.

PAGE 241

Gallons per 1,000 persons 200 180 160 140 120 100 80 Oct.-Nov. 1972 I / I \ \ I \ V \ Oct -Nov. 1973 227 Oct.-Nov 1974 ,,,,,,. I / / / /' I/ \ Estimated ',, ,, Oct.-Nov. 1975 Oct.-Nov. 1976 Oct.-Nov. 1977 Oct.-Nov. 1978 Figure H.7. Actual and reduced-form estimated values of retail quantities of other orange juices in the USA (RQOSt), April-May/1972 to February-March/1979. Dollars per gallon 2 6 2 4 2.2 2.0 1 8 1.6 Note : Other orange juices are expressed in gallons of single strength juice equivalent. Oct -Nov. 1972 Oct.-Nov. 1973 Oct.-Nov. 1974 Oct.-Nov 1975 Actual Oct.-Nov. 1976 -~'.,,,, ,_ Oct -Nov. 1977 Oct -Nov 1978 Figure H.8. Actual and reduced-form estimated values of retail prices (1975=100) of other orange juices in the USA (RPOSt), April-May/1972 to February-March/1979.

PAGE 242

Gallons per 1 000 persons 200 180 160 140 120 100 Oct -Nov. 1972 Oct.-Nov. 1973 228 Estimated Oct.-Nov. 1974 Oct.-Nov. 1975 Actual Oct -Nov. 1976 Oct. Nov. 1977 Oct.-Nov. 1978 Figure H.9. Actual an@ reduced-form estimated values of wholesale quantities of other orange juices in Florida (WQOFlt)' April-May/1972 to February March/1979. Note : Other orange juices quantities are expressed in gallons of single strength juice equivalent Dollars per gallon 2 0 1.8 1.6 1.4 1.2 1 0 Oct.-Nov. 1972 / Estimated ,., / ... '-~ Oct.-Nov. 1973 \ '-Oct -Nov. 1974 Oct.-Nov 1975 Oct.-Nov. 1976 Oct.-Nov. 1977 Oct.-Nov. 1978 Figure H.10 Actual and reduced-form estimated values of wholesale prices (1975=100) of other orange juices in Florida (WPOFlt)' April-May/1972 to February-March/1979.

PAGE 243

Gallons per 1,000 450 400 350 300 250 200 150 100 50 0 persons Oct.-Nov. 1972 Estimated Oct.-Nov 1973 229 Oct.-Nov. 1974 f Oct.-Nov. 1975 Oct -Nov. 1976 Oct.-Nov. 1977 Oct.-Nov. 1978 Figure H.11. Ac t ual and reduced-form estimated values of export quantities of FCOJ to Canada from Brazil (XQFCBt). April-May/1972 to February-March/1979 Dollars per gallon 0.7 0.6 0.5 0.4 0.3 0.2 Note : FCOJ quantities are expressed in gallons of single strength juice equivalent Oct.-Nov 1972 Estimated Oct -Nov. 1973 Oct.-Nov. 1974 Oct.-Nov 1975 Actual~ ,-, I I I' I J / I Oct.-Nov. 1976 Oct.-Nov. 1977 Oct.-Nov. 1978 Figure H 12 Actual and reduced-form estimated values of export price (1975=100) of FCOJ to Canada from Braz i l (XPFCBtl, April-May/1972 to February-March/1979.

PAGE 244

Gallons per 1 ,000 persons 380 340 300 260 220 180 140 100 Oct -Nov. 1972 Oct.-Nov. 1973 230 Actual Oct.-Nov. 1974 Oct.-Nov. 1975 Oct.-Nov. 1976 Oct.-Nov. 1977 Oct.-Nov. 1978 Figure H 13. Actual and reduced-form estimated values of export quantities of FCOJ to Canada from the USA (XQFCSt), April-May/1972 to February-March/1979. Dollars per gallon 1 60 1.50 1.40 1.30 1 20 1.10 1.00 90 .80 Note: FCOJ quantities are expressed in gallons of single strength juice equivalent. Oct -Nov 1972 Oct.-Nov. 1973 Oct.-Nov. 1974 Oct .Nov. 1975 Oct.-Nov. 1976 Oct -Nov. 1977 Oct.-Nov. 1978 Figure H.14 Actual and reduced-form estimated values of export price (1975=100) of FCOJ to Canada from the USA (XPFCSt), April-May/1972 to February-March/1979.

PAGE 245

231 Gallons per 1,000 persons 225 200 175 ,.'\ "\ 150 \ / t I \ I 125 I \ I 100 \ I \ I 75 /'/ 50 Estimated 25 I 0 I Oct -Nov Oct. Nov Oct.-Nov. Oct. Nov. Oct .Nov Oct -Nov Oct -Nov. 1972 1973 1974 1975 1976 1977 1978 Figure H 15. Actual and reduced-form estimated values of export quantities of FCOJ to the EEC7 countries from Brazil (XQFEBt), April-May/1972 to February-March/1979 Dollars per gallon 65 60 55 50 .45 .40 35 30 25 Note : FCOJ quantities are expressed in gallons of single strength juice equivalent. Oct -Nov 1972 Oct .Nov 1973 Oct.-Nov. 1974 Oct .Nov. 1975 Oct.-Nov 1976 I I ..I / / Oct.-Nov. 1977 Oct.-Nov. 1978 Figu r e H 16 Actual and reduced-form estimated values of export price (1975=100) of FCOJ to the EEC7 countries from Brazil (XPFEBt)' April-May/1972 to February March/1979

PAGE 246

232 Gallons per 1,000 persons 30 25 20 '\ \ Estimated \ \ 15 '\ \. \ I \ I \ 10 I 5 0 Oct.-Nov. 1972 Oct .Nov 1973 Oct.-Nov. 1974 Oct.-Nov. 1975 Oct. Nov. 1976 Oct.-Nov. 1977 Oct.-Nov. 1978 Figure H 17. Dollars per gallon 1.3 1.1 0.9 0.7 0 5 0.3 Actual and reduced-form estimated values of export quantities of FCOJ to the EEC7 countries from the USA (XQFESt)' AprllMay/1972 to February-March/1979. Note : FCOJ quantit i es are expressed in gallons of single strength Juice equivalent Oct.-Nov. 1972 Oct .Nov. 1973 Oct -Nov. 1974 Oct.-Nov. 1975 Oct .Nov. 1976 Oct.-Nov 1977 Oct.-Nov. 1978 Figure H.18 Actual and reduced-form estimated values of export price (1975=100) of FCOJ to the EEC7 countries from the USA (XPFESt)' Aprll-May/1972 to February-March/1979.

PAGE 247

Gallons per 1,000 persons 700 600 500 400 300 200 100 0 I 233 Oct .Nov 1972 Oct.Nov. 1973 Oct.Nov. 1974 Oct.-Nov. 1975 Oct.-Nov. 1976 Oct. Nov. 1977 Oct.-Nov. 1978 Figure H.19. Dollars per gallon 70 65 60 55 50 .45 40 35 30 Actual and reduced-form estimated values of export quantities of FCOJ to the European non EEC3 countries from Brazil (XQFNBt)' Aprll-May/1972 to FebruaryMarch/1979. Note: FCOJ quantities are expressed In gallons of single strength juice equivalent. I ; ,----,,,,,_ Oct -Nov 1972 Oct.-Nov. 1973 Oct -Nov. 1974 Oct.-Nov. 1975 Oct.-Nov 1976 Oct .Nov. 1977 Oct.-Nov 1978 Figure H.20. Actual and reduced-form estimated values of export price (1975=100) of FCOJ to the European non-EEC3 countries from Brazil (XPFNBt)' Aprll-May/1972 to February-March/1979.

PAGE 248

Gallons per 1,000 persons 225 200 175 150 125 100 75 50 l \ / \.J Actual, 25 0 ,,,,,,, I I I I I Oct -Nov. 1972 Oct -Nov. 1973 234 I \,, I I I I I Oct Nov 1974 I I I I I Oct -Nov 1975 I I I Oct -Nov 1976 I I I I I Oct -Nov. 1977 I I I I I. Oct -Nov. 1978 Figure H.21. Actual and reduced-form estimated value of export quantities of FCOJ to the European non EEC3 countries from the USA (XQFNSt) April May/1972 to February March/1979 Dollars per gallon 1.3 1 2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 Note : FCOJ quantities are expressed in gallons of single strength juice equivalent. Oct -Nov. 1972 Oct. Nov. 1973 Oct.-Nov 1974 Oct -Nov 1975 Oct.-Nov 1976 Oct -Nov. 1977 Oct. Nov. 1978 Figure H.22. Actual and reduced-form estimated values of export price (1975=100) of FCOJ to the European non-EEC3 countries from the USA (XPFNSt), April May/1972 to February-March/1979

PAGE 249

Gallons per 1 000 persons 2250 2000 1750 1500 1250 1000 750 500 250 235 Oct. Nov 1972 Oct -Nov 1973 Oct.-Nov. 1974 Oct. Nov 1975 Oct.-Nov. 1976 Oct.-Nov. 1977 Oct.-Nov. 1978 Figure H.23. Number of blmonth s 4 5 4.0 3 5 3.0 2.5 2.0 1.5 1.0 0.5 Actual and reduced form estimated values of wholesale quantities in stocks of FCOJ In the USA (WQkFS ), Aprll-May/1972 to February-March/1979. Note : ~COJ quantities In gallons of single strength Juice equivalent. Est i mated--.... {' I I Oct.-Nov. 1972 Oct.-Nov. 1973 Oct -Nov. 1974 Oct. Nov. 1975 Oct -Nov. 1976 Oct.-Nov. 1977 Oct.-Nov. 1978 Figure H .2 4. Actual and reduced-form estimated values of wholesale stocks of FCOJ I n the USA (WSkFS t )' April-May/1972 to February-March/1979. Note: FCOJ stocks are expressed in number of blmonths of Florida's supplies.

PAGE 250

Gallons per 1,000 persons 270 240 210 180 150 120 90 60 30 A Estimated Oct.-Nov. 1972 Oct -Nov. 1973 236 Oct -Nov 1974 Oct.-Nov. 1975 Oct.-Nov. 1976 Oct.-Nov. 1977 ,. I I I Oct.-Nov 1978 Figure H 25 Actual and reduced-form estimated values of wholesale quantities in stocks of other orange Juices in the USA (WOkOSt)' April-May/1972 to February-March/1979. Note : Other orange juice quantities are expressed in gallons of single strength juice equivalent. Number of bimonths 2 25 2.00 1.75 1 50 1 25 1.00 .75 .50 25 Oct -Nov 1972 Oct -Nov 1973 Oct -Nov. 1974 Oct.-Nov. 1975 Oct.-Nov. 1976 Oct.-Nov 1977 Oct -Nov 1978 Figure H.26 Actual and reduced-from estimated values of wholesale stocks of other orange juices in the USA (WSkOSt), April May/1972 to February-March/1979. Note : Other orange juices stocks are expressed in number of bimonths of Florida's supplies.

PAGE 251

Dollars per gallon 80 .75 .70 .65 .60 .55 50 45 .40 .35 .30 Oct -Nov. 1972 237 Estimated I ,.... ""' _, I Oct.-Nov. 1973 Oct -Nov 1974 Oct.-Nov. 1975 I \ I \ I \ I \ I I \-J I I I" Oct.-Nov. 1976 Actual Oct.-Nov 1977 Oct.-Nov 1978 Figure H.27. Actual and reduced-form estimated values of the price (1975=100) spread between wholesale price of FCOJ in Florida (WPFFlt) and import price of FCOJ by the USA from Brazil (MPFSB t ), April-May/1972 to February-March/1979. Proportion .75 .70 .65 60 .55 50 .45 .40 35 30 25 Oct -Nov. 1972 Oct -Nov. 1973 Qct.-Nov. 1974 r" /' Estimated'\. / \,1 \ "" ,, ... \., Oct.-Nov 1975 ,-I Oct -Nov 1976 Oct -Nov 1977 ,, I I Oct.-Nov 1978 Figure H 28 Actual and reduced-form estimated values of retail proportion of other orange juices in cardboard containers i n the USA (RpOSt) April-May/1972 to February March/1979

PAGE 252

LIST OF REFERENCES Abrams, R. K.: International trade flows under flexible exchange rates. Econ. Rev., Federal Reserve Bank of Kansas City, March 1980, pp. 3-10. Alves, E. R. de A., and Pastore, A. C.: Import substitution and implicit taxation of agriculture in Brazil~ Amer. J. Agr. Econ. 60: 865-871, 1978. Brazil, Carteira de Comercio Exterior do Banco do Brazil S.A. (CACEX): Unpublished monthly data on Brazilian orange juice exports by country of destination. Rio de Janeiro, Brazil, 1972-1979. Brazil, Servi90 de Estatistica Econ8mica e Financeira, Ministerio da Fazenda: Comercio Exterior do Brazil Exporta9ao, various volumes, Funda9ao Instituto Brasileiro de Geografia e Estatistica, Rio de Janeiro, Brazil, 1970-1979. Bredahl, M. E., Meyers, W. H., and Collins, K. J.: The elasticity of foreign demand for U. S. agricultural products: The importance of the price transmission elasticity. Amer. J. Agr. Econ. 61: 58-63, 1979. Buckley, G. N.: Econometric analysis of the FCOJ wholesale retail subsector. M. S. thesis, University of Florida, Gainesville, Florida, 1977. Chern, W. S.: Econometric analysis of the consumer demand for orange juices in Canada. Economic Research Department, Florida Department of Citrus, ERD Report 73-2, University of Florida, Gainesville, Florida, 1973. Collins, K. J., Meyers, W. H., and Bredahl, M. E.: Multiple exchange rate changes and U. S. agricultural commodity prices. Amer. J. Agr. Econ. 62: 656-665, 1980. Florida Canners Association: Weekly statistical report on frozen concentrated orange juices, canned single strength orange juice and chilled orange juice, various reports. Winter Haven, Florida, 1971-1980. 238

PAGE 253

239 Florida Department of Citrus: Unpublished data of promotional expenditures of Florida orange juice. Lakeland, Florida, 1980. Florida Department of Citrus and A. C. Nielsen Company: Food store data on selected citrus and other fruit products, various volumes. Lakeland, 1972-1980. Florida Department of Citrus, Economic Research Department: Unpublished data on citrus price. Economic Research Department, Florida Department. of Citrus, University of Florida, Gainesville, Florida, 1972-1980. Florida Crop and Livestock Reporting Service: Florida agricultural statistics, citrus summary 1976. Orlando, Florida, December 1976. Florida Crop and Livestock Reporting Service: Florida agricultural statistics, citrus summary 1979. Orlando, Florida, December 1979. Florida Crop and Livestock Reporting Service: Monthly citrus forecast, various reports. Orlando, Florida, 1972-1980. Food and Agricultural Organization of the United Nations: Production Yearbook, various volumes. Rome, Italy, 1973-1979. Food and Agricultural Organization of the United Nations: Unpublished data on orange juice exports and imports 1967 to 1978. Rome, Italy, May 1980. Funda9ao Getulio Vargas: Conjuntura Econ8mica, various volumes. Rio de Janeiro, Brazil, 1972-1980. Goldberger, A. S.: Econometric Theory. New York: John Wiley and Sons, Inc., 1964. Goldman, S. M., and Uzawa, H.: A note on separability in demand analysis. Econometrica, 32: 387-398, 1964. Gould, J. P., and Ferguson, c. E.: Microeconomic Theory. Homewood, Ill.: Richard D. Irwin, Inc. 1980. Greenshields, B. L.: Changes in exchange rates: Impact on u. s. grain and soybean exports to Japan. United States Department of Agriculture, Economic Research Service, ERS Foreign 364, Washington, D. C., July 1974. Henderson, J. M., and Quandt, A Mathematical Approach. Hill Company, 1980. R. E.: Microeconomic Theory: 3rd ed. New York: McGraw

PAGE 254

240 International Monetary Fund: International financial statistics, various volumes. Washington, D.C., 1970-1980. Israel, Central Bureau of Statistics: Foreign trade statistics,monthly, various volumes. Jerusalem, Israel, 1970-1980. Johnson, P. R., Grennes, T., and Thursby, M.: Devaluation, foreign trade controls, and domestic wheat prices. Amer. J. Agr. Econ. 59: 619-627, 1977. Kennedy, P. : A Guide to Econometrics. New York: Cambridge, Mass. : The MIT Press, 1979. Kmenta, J.: Elements of Econometrics. New York: Macmillan Publishing Co., Inc., 1971. Kost, W. E.: Effects of an exchange rate change on agricultural trade. Agr. Econ. Res. 28: 99-106, 1976. Lee, Jong-Ying: A study of the impact of Three-Party Program on European demand for U.S. FCOJ. Economic Research Department, Florida Department of Citrus, CIR 77-2, University of Florida, Gainesville, Florida, June 1977. Lee, Jong-Ying: A study of the impact of Three-Party Program on European demand for U.S. FCOJ. Abstracts of Contributed Papers. Southern J. Agr. Econ. 10: 147, 1978. Lee, Jong-Ying: Florida Department of Citrus advertising research programs. Paper presented at NC117 Symposium on Advertising in the Food System, Airlee, Virginia, November 6, 1980. Lee, Jong-Ying, Myers, L. H., and Forsee, F.: Economic effectiveness of brand advertising programs for Florida orange juice in European markets. Economic Research Department, Florida Department of Citrus, ERD Report 79-1, University of Florida, Gainesville, Florida, August 1979. Lingens, H.: Production, marketing and trade in Europe: Consumption and marketing of citrus (processed citrus). In Grierson, W., ed.: 1977 Proceedings of the Inter national Society of Citriculture. Volume 2, Lake Alfred, Florida, December 1978, pp. 378-380. Maddala, G. S.: Econometrics. New York: McGraw-Hill, Inc., 1977.

PAGE 255

241 Malick, W. M.: A simultaneous equation model of the Florida retail orange juice marketing system. M. S. thesis, University of Florida, Gainesville, Florida, 1980. Melamed, Abraham. The citrus marketing board of Israel and the auction demand for weekly sales. Ph.D. dissertation, University of California, Berkeley, California, 1977. Mexico, Secretaria de Industria y Comercio. Anuario Estadistico del Comercio Exterior de los Estados Unidos Mexicanos, various volumes, Secretaria de Industria y Comercio, Ciudad de Mexico, 1972-1979. Morais, E. L. de, and Medeiros, M. de L. M. de: A participa9ao Brasileira no mercado internacional de laranja e de suco de laranja: Comportamento recente e perspectivas. Ministerio da Agricultura, (Secretaria Nacional de Planejamento Agricola) Brasilia, D. F., Junho 1978. Moretti, V. A.: Demand for Brazilian frozen concentrated orange juice exports. M. S. thesis, University of Florida, Gainesville, Florida, 1978. Myers, L. H.: The Brazilian citrus industry: Scope and outlook. Report to the Market and Economic Research Committee of the Florida Citrus Commission (FCC), Florida Department of Citrus, Lakeland, Florida, July 31, 1978a. Myers, L. H.: Marketing and consumption of processed citrus products in the United States, Canada and Latin America. In Grierson, W., ed.: 1977, Proceedings of the Inter national Society of Citriculture. Volume 2, Lake Alfred, Florida, December 1978b, pp. 369-374. Myers, L. H.: The Brazilian citrus industry: An overview. Report to the Market and Economic Research Committee of the Florida Citrus Commission (FCC), Florida Department of Citrus, Lakeland, Florida, December, 1978c. Myers, L. H., Tilley, D. S., Lee, Jong-Ying, Fairchild, G. F., and Brooks, T. L.: FCOJ import-export policies and programs: An overview. Economic Research Department, Florida Department of Citrus, Working Paper 79-5, University of Florida, Gainesville, Florida, February 21, 1979. Nguyen, K. Q.: Effects of prices, income and exchange rates on the European demand for frozen concentrated orange juice. M. S. thesis, University of Florida, Gainesville, Florida, 1977.

PAGE 256

242 Phlips, L.: Applied Consumption Analysis. Amsterdam: North Holland Publishing Co., 1974. Polopolous, L.: Foreign imports of orange concentrate and the Florida citrus industry. Unpublished report, Economic Research Department, Florida Department of Citrus, University of Florida, Gainesville, Florida, October 2, 1968. Prato, A. A.: An econometric study of consumer demand for fresh oranges and frozen concentrated orange juice. Ph.D. dissertation, University of California. Berkeley, California, 1969. Priscott, R.H.: Demand for citrus products in the European markets. M. S. thesis, University of Florida, Gainesville, Florida, 1969. Rausser, G. C.: A dynamic econometric model of the California-Arizona orange industry. Ph.D. dissertation, University of California, Davis, California, 1971. Richter, M. K.: Revealed preference theory. Econometrica 34: 635-645, 1966. Russell, R.R., and Wilkinson, M.: Microeconomics: A Synthesis of Modern and Neoclassical Theory. New York: John Wiley and Sons, 1979. Ryan, M. E., and Tontz, R. L.: A historical review of world trade policies and institutions. In McClure, G., ed.: Speaking of Trade: Its Effect on Agriculture. Agricultural Extension Service, University of Minnesota, Special Report No. 72, Nov. 1978, pp. 5-19. Schuh, G. E.: The exchange rate and U.S. agriculture. Arner. J. Agr. Econ. 56: 1-13, 1974. Silberberg, E.: Analysis. The Structure of Economics: A Mathematical New York: McGraw-Hill Book Company, 1978. Statistics Canada: Import by commodities and countries, various volumes. Ottawa, Canada, 1969-1979. Statistics Canada: National income and expenditure accounts fourth quarter and preliminary annual 1978. Ottawa, Canada, April, 1979. Taylor, E. A.: World citrus trade barriers. In Grierson, w., ed.: 1977 Proceedings of the International Society of Citriculture. Volume 2, Lake Alfred, Florida, December 1978, pp. 389-391.

PAGE 257

243 Tilley, D. S.: Importance of understanding consumption dynamics in market recovery periods. Southern J. Agr. Econ. 11: 41-46, 1979. Tilley, D. s., and Lee, Jong-Ying,: Import and retail level demand for orange juice in Canada. Can. J. Agr. Econ. July 1981, in press. Tomek, W. G. and Robinson, Prices. Ithaca, N.Y.: Press, 1981. K. L.: Agricultural Product 2nd ed. Cornell University United States Board of Governors of the Federal Reserve System: Federal Reserve Bulletin, various volumes. Washington, D.C., 1972-1980. United States Department of Agriculture, Foreign Agricultural Service: Brazil's orange juice industry. FAS-M-295, USDA-FAS, Washington, D.C., April 1980a. United States Department of Agriculture, Foreign Agricultural Service: Brazil: Annual citrus report. Unclassified agricultural attache report BP-0014, Sao Paulo, Brazil, June 9, 1980b. United States Department of Agriculture, Foreign Agricultural Service: Fresh and processed citrus fruit. FCF 2-80, USDA-FAS, Washington, D.C., July 1980c. United States Department of Commerce, Bureau of the Census: U.S. exports: Commodity by country. Report FT410. Washington, D.C., 1969-1979a. United States Department of Commerce, Bureau of the Census: u. s. imports: Commodity by country. Washington, D.C., 1969-1979b. United States Department of Commerce, Bureau of the Census: Current population reports: Population estimates and projections, various volumes (P-25 series), Washington, D.C., 1975-1980. United States Department of Commerce, Bureau of Economic Analysis: Survey of current business, various volumes. Washington, D.C., 1975-1980. Uzawa, H.: Preference and rational choice in the theory of consumption. In Arrow, K. J., Karlin, s., and Suppes, P., eds.: Mathematical Methods in the Social Sciences, 1959. Proceedings of the first Stanford Symposium, Stanford University Press, 1960, pp. 129-148.

PAGE 258

244 Veiga, A.: The impact of trade policy on Brazilian agriculture 1947-1967. Ph.D. dissertation, Purdue University, West Lafayette, Indiana, 1974. Vellianitis-Fidas, A.: The impact of devaluation on u. S. agricultural exports. Agr. Econ. Res. 28:107-116, 1976. Ward, R. W.: The economics of Florida's FCOJ imports and exports: An econometric study. Economic Research Department, Florida Department of Citrus, ERD Report 76-1, University of Florida, Gainesville, Florida, August 1976a. Ward, R. W.: A two-pricing system for exporting frozen concentrated orange juice. Southern J. Agr. Econ. 8: 131-135, 1976b. Ward, R. W., and Tilley, D. S.: Time varying parameters with random components: The orange juice industry. Southern J. Agr. Econ. 12: 5-13, 1980. Ward, R. W., and Niles, J. A.: An econometric analysis of FCOJ imports and exports. Florida State Horticultural Society J., Proceedings, pp. 61-65, November 1975. Wolff, J.: Medium-term prospects for world citrus production. In Grierson, w. ed.: 1977 Proceedings of the Inter national Society of Citriculture. Volume 2, Lake Alfred, Florida, December 1978, pp. 328-332. Womack, A. W., and Matthews, J.: Linear approximations of nonlinear relationships by the Taylor's Series expansion revisited. Agr. Econ. Res. 24: 93-101, 1972.

PAGE 259

BIOGRAPHICAL SKETCH Luiz Jose Maria Irias was born in Ponte Nova, Minas Gerais, Brazil, on January 9, 1944. He has a B.S. degree in agricultural engineering (1967) and an M.S. degree in agricultural economics (1970) from the Universidade Federal de Vi9osa, Minas Gerais, Brazil. He worked as an agricultural economics researcher for the Centro de Estudos Agricolas do Institute Brasileiro de Economia Agricola da Funda9ao Getulio Vargas, Rio de Janeiro (1969 to 1971) and for the Empresa Brasileira de Assist~ncia Tecnica e Extensao Rural (1971 to 1973). Since 1974 he has been working for the Empresa Brasileira de Pesquisa Agropecuaria, a federal government enterprise that carries out agricultural research in Brazil. In the fall of 1976, he started his graduate study at the University of Florida, Gainesville, USA, for the degree of Doctor of Philosophy with major in Food and Resource Economics. He is a member of the International Association of Agricultural Economists, American and Southern Agricultural Economics Associations, Sociedade Brasileira de Economistas Rurais and Associa9ao dos Engenheiros Agronomos do Distrito Federal. He is married to Raulina Grossi Irias and has two daughters, Ana Claudia and Ana Cristina. 245

PAGE 260

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Daniel S. Tiliey,airman Associate Professor of Food and Resource Economics I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Assistant Professor of Food and Resource Economics I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Max R. Langham Professor of Food and Resource Economics I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor Economics

PAGE 261

This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philo ophy. August, 1981 Dean culture Dean, Graduate School