Citation
Analysis of U.S.-Mexico sugar trade : impacts of the North American Free Trade Agreement (NAFTA) and projections for the future

Material Information

Title:
Analysis of U.S.-Mexico sugar trade : impacts of the North American Free Trade Agreement (NAFTA) and projections for the future
Creator:
Sano, Daisuke
Publication Date:
Language:
English
Physical Description:
xi, 136 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Imports ( jstor )
Market prices ( jstor )
Prices ( jstor )
Sugar cane ( jstor )
Sugar industry ( jstor )
Sugars ( jstor )
Supply ( jstor )
Supply and demand ( jstor )
Sweeteners ( jstor )
Table sugars ( jstor )
Dissertations, Academic -- Food and Resource Economics -- UF
Food and Resource Economics thesis, Ph. D
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2004.
Bibliography:
Includes bibliographical references.
General Note:
Printout.
General Note:
Vita.
Statement of Responsibility:
by Daisuke Sano.

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:
022829416 ( ALEPH )
880636988 ( OCLC )

Downloads

This item has the following downloads:


Full Text












ANALYSIS OF U.S.-MEXICO SUGAR TRADE: IMPACTS OF THE NORTH
AMERICAN FREE TRADE AGREEMENT (NAFTA) AND PROJECTIONS FOR THE FUTURE

















By

DAISUKE SANO


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

UNIVERSITY OF FLORIDA


2004

































Copyright 2004

by

Daisuke Sano



































To my parents and sister
















ACKNOWLEDGMENTS

I would like to express my gratitude to Dr. Thomas H. Spreen, chair of my

supervisory committee, for providing me with resources throughout my program of studies. My dissertation would not have been possible without his guidance, unflagging enthusiasm and helpful ideas.

I also would like to express my gratitude to Dr. Lisa A. House, cochair of my

supervisory committee, for her generous and patient support during my course of studies. Under her supervision, my research has been completed in an efficient and successful manner.

Special appreciation is extended to Dr. Chris 0. Andrew for his encouragement throughout the program and for broadening my horizons. His inspiration has been a compass after the direction of my career had been shifted. I am also deeply grateful to Dr. Luis R. Garcia for providing me with insights and detailed data which were indispensable for this study; to Dr. Terry L. McCoy for his contributing to the depth of my study; and to Dr. Kenneth L. Buhr for his sharing suggestions for the breadth of my study. Kind support from Dr. Jeffrey R. Burkhardt and staff in the department for their various services are also appreciated. I cannot fail to thank John S. Lander, who opened my eyes to studying overseas, for his unconditional support and wisdom. Friendship from my colleagues in the program is a cherished treasure.

Last but not least, the financial support received from the Food and Resource Economics Department throughout my program has been greatly appreciated.


iv
















TABLE OF CONTENTS

page

ACKNOW LEDGM ENTS ................................................................................................. iv

LIST OF TABLES........................................................................................................ vii

LIST OF FIGURES ....................................................................................................... ix

A B S T R A C T ...................................................................................................................... xii

CHAPTER

1 INTRODUCTION ................................................................................................... I

Background............................................................................................................ I
P rob lem S tatem en t........................................................................................................5
Researchable Questions ............................................................................................ 5
O bjectiv e s .....................................................................................................................6
Organization of the Study ........................................................................................ 6

2 SUGAR INDUSTRIES AND SWEETENER MARKETS IN THE UNITED
STATES AND M EXICO ..........................................................................................8

The M exican Sugar Industry and Sweetener M arket .............................................. 9
M exico's Sugarcane Production....................................................................... 9
M exico's Sugar Production ............................................................................ 10
The M exican Sugar Industry and Government Involvement .......................... 12
M exico's Sugar Consumption ..........................................................................14
M exico as a Sugar Exporter ............................................................................ 15
Development and Adoption of High Fructose Corn Syrup .....................................17
The U.S. Sugar Industry and Sweetener M arket .....................................................19

3 CONCEPTUAL AND THEORETICAL FRAMEWORK.....................................32

Conceptual Framework.......................................................................................... 32
Theoretical Framework.......................................................................................... 35
Sweetener M arket Analysis............................................................................ 35
Sweetener demand................................................................................... 35
Sweetener supply...................................................................................... 36
U.S.-M exico Bilateral Sugar Trade System .................................................. 37


v











4 EM PIRICAL M ODELS AND DATA SOURCE................................................... 47

Empirical M odels.................................................................................................... 47
U.S. Sweetener Demand M odel ..................................................................... 47
M exican Sweetener Demand M odel .............................................................. 48
U.S. Sweetener Supply M odel........................................................................ 50
M exican Sweetener Supply M odel................................................................. 52
U.S.-M exico Bilateral Sugar Trade M odel..................................................... 53
Simplifying assumptions .......................................................................... 54
M odel calibration ..................................................................................... 56
Simulated Scenarios ....................................................................................... 58
Game Theory Analysis ................................................................................... 61
Sources of Data...................................................................................................... 63

5 EMPIRICAL RESULTS AND INTERPRETATION............................................75

Demand and Supply Analyses .............................................................................. 75
Bilateral Sugar Trade Analysis............................................................................... 77
Game Theory Analysis .......................................................................................... 85

6 CONCLUSIONS AND IMPLICATIONS FOR POLICY .......................................115

Conclusions and Implications for Policy ..................................................................115
Impact of Changes in Trade Regime .................................................................1 16
M exico's Export Potential.................................................................................117
The Impact of Changes in M exican M arket Situation.......................................117
The Impact of Changes in U.S. Sugar Policy ....................................................118
Alternative Sugar Policy by the United States ..................................................120
Limitation of the Study and Suggestions for Future Research .................................122

APPENDIX

A MAJOR EVENTS IN THE SUGAR INDUSTRY HISTORY IN MEXICO AND
T H E U .S . ..................................................................................................................12 3

B CORN STATISTICS................................................................................................125

C DERIVATION OF INVERSE LINEAR EQUATIONS ..........................................128

BIOGRAPHICAL SKETCH ...........................................................................................136


vi
















LIST OF TABLES


Table page

2-1 Cane Sugar Production in Selected Countries, 1997-2000 Average...................25

2-2 World Sugar Consumption in 2000.....................................................................27

2-3 Quota and Tariff Schedule Imposed on Mexican Sugar Exported to the U.S..........27

4-1 Assumptions for Baseline (Status quo) Scenario ................................................66

4-2 Assumptions for Mexican Sweetener Market Situations ....................................67

4-3 Assum ptions for U .S. Sugar Policies .................................................................. 69

4-4 Listing of Examined Scenarios ................................................................ ....70

4-5 Strategies for the Sugar Trading Game ................................................................71

4-6 Data Sources for U.S. Demand ...........................................................................72

4-7 Data Sources for Mexican Demand ....................................................................72

4-8 D ata Sources for U .S. Supply.............................................................................. 73

4-9 D ata Sources for M exico Supply ......................................................................... 73

4-10 D ata Sources for M iscellaneous ......................................................................... 74

5-1 Summary of the U.S. Supply-Demand Analysis................................................ 91

5-2 Summary of the Mexican Supply-Demand Analysis ...........................................92

5-3 Impact of Changes in Mexican Sweetener Market on Pay-offs to the Industries and
N ation's W elfare [B illion U S$] .............................................................................10 1

5-4 Impact of Changes in U.S. Price Stabilization Policy on Pay-offs to the Industries
and Nation's Welfare [Billion US$] - Flexible Quota Allocations -.....................102

5-5 Impact of Changes in U.S. Quota Allocation Policy on Pay-offs to the Industries
and Nation's Welfare [Billion US$] - Minimum Quota Allocations -..................103


vii









5-6 Pay-off Matrix for the Trade Policy Game [Billion US$].....................................107

5-7 Indexed Pay-off Matrix for the Trade Policy Game without Coalitions [Baseline =
10 0 ].......................................................................................... .. -. .... ....- ---- ..........10 8

5-8 Indexed Pay-off Matrix for the Trade Policy Game Played by Two Coalitions of
Countries without the U.S. HFCS industry [Baseline=100] ..................................109

5-9 Indexed Pay-off Matrix for the Trade Policy Game Played by the Industry
Coalition and the Government Coalition without the U.S. HFCS industry
[Baseline=100].................................................................110

5-10 Indexed Pay-off Matrix for the Trade Policy Game Played by the Grand Coalition
without the U.S. HFCS industry [Baseline=100]...................................................1 I1

5-11 Indexed Pay-off Matrix for the Trade Policy Game Played by Two Coalitions of
Countries with the U.S. HFCS industry [Baseline=100] .......................................112

5-12 Indexed Pay-off Matrix for the Trade Policy Game Played by the Industry
Coalition and the Government Coalition with the U.S. HFCS industry
[Baseline=100]............................................................113

5-13 Indexed Pay-off Matrix for the Trade Policy Game Played by the Grand Coalition
with the U.S. HFCS industry [Baseline=100].......................................................114

C-1 Coefficients for Inverse Linear Functions -U.S.-...................................................129

C-2 Coefficients for Inverse Linear Functions -M exico- .............................................130


viii
















LIST OF FIGURES


Figure page

2-1 Distribution of Individual Farm Size of Mexican Sugarcane Growers.........24 2-2 Map of Mexico's Sugar Producing and Processing States...................................24

2-3 Annual Rainfall and Irrigation Rate in Sugarcane Fields ....................................25

2-4 Mexican Sugar Production, 1988-2002................................................................26

2-5 Main Use of Sugarcane Derivatives in Mexico................................................... 26

2-6 U.S. Refined Sugar and HFCS Use per Capita .................................................. 28

2-7 U .S. H FC S Production ......................................................................................... 28

2-8 U .S. H FC S Supply .............................................................................................. 29

2-9 U .S. H FC S E xport ................................................................................................. 29

2-10 Transition of U.S. HFICS Export ......................................................................... 30

2-11 Consumption of Sugar and HFCS per Capita in Mexico.....................................30

2-12 Transition of Prices of Sugar and HFCS ..............................................................31

3-1 Conceptual Framework for the Analysis in this Study ...................................... 45

3-2 Two-country Trade Model (a) with Quota System and (b) without Quota System. 46

4-1 Im age of M odel C alibration ................................................................................ 65

4-2 Forecasted Indirect Sugar and HFCS Consumption in Mexico under Alternative
S c e n ario s ..................................................................................................................6 8

5-1 U.S. Sugar Import Forecast (Scenario I "Baseline")...........................................93

5-2 U.S. Sugar Import Forecast (Scenario 2 "P-S-F") ...............................................93

5-3 U.S. Sugar Import Forecast (Scenario 3 "A-S-F").............................................. 94


ix









5-4 U.S. Sugar Import Forecast (Scenario 4 "PA-S-F")............................................94

5-5 U.S. Sugar Import Forecast (Scenario 13 "PA-S-M") .........................................95

5-6 U.S. Sugar Import Forecast (Scenario 8 "PA-B-F") ...........................................95

5-7 U.S. Sugar Import Forecast (Scenario 12 "PA-C-F") .........................................96

5-8 U.S. Sugar Import Forecast (Scenario 16 "T-S-F") .............................................96

5-9 Forecasted Equilibrium Sugar Prices in the U.S. and Mexican Markets (Scenario 4
"PA -S-F")..................................................................................... .......-97

5-10 Forecasted Equilibrium Sugar Prices in the U.S. and Mexican Markets (Scenario 8
"PA-B-F") ................................................................................97

5-11 Forecasted Equilibrium Sugar Prices in the U.S. and Mexican Markets (Scenario 12
"PA-C-F") .............................................................................. -...98

5-12 Forecasted Equilibrium Sugar Prices in the U.S. and Mexican Markets (Scenario 16
"T -S -F") ......................................................................---..... --------................98

5-13 Forecasted Sugar Demand and Supply for both the United States and Mexico
(Scenario 1 "Baseline")................................................................................... 99

5-14 Forecasted Sugar Demand and Supply for both the United States and Mexico
(Scenario 4 "P-S-F") ................................................................................... ...-99

5-15 Forecasted Equilibrium Sugar Prices in the United States and Mexico (Scenario 14
"P A -B -M ") .............................................................................................................10 0

5-16 Forecasted Equilibrium Sugar Prices in the United States and Mexico (Scenario 15
"P A -C -M ") ................................................................................................ .........100

5-17 Absolute Effects of Production Improvement and IFCS Adoption on Pay-off to the
M exican Sugar Industry .........................................................................................104

5-18 Absolute Effects of Production Improvement and HFCS Adoption on Pay-off to
M exico 's W elfare ...................................................................................................105

5-19 Relative Gain and Loss of Pay-offs to the Mexican Sugar Industry and Welfare
Caused by Production Improvement and HFCS Adoption ....................................106

B-i Recent Corn Production and Consumption for Selected Countries .......................126

B-2 Food and Industrial Corn Use in the U.S., 1980-2002...........................................126

B-3 Corn Price (No.2 Yellow) in Chicago Market, 1981-1998 ....................................127


x










B-4 Exports of Products Made from Corn in 2002.......................................................127


xi















CHAPTER 1
INTRODUCTION

This chapter introduces the research problem for U.S.-Mexico sugar trade.

Background, the researchable questions, and objectives are provided, followed by the organization of the dissertation.

Background

Trade issues surrounding the world sugar market are often seen as classic examples in agricultural economics, yet the market still provides us with important questions today. In the case of the U.S.-Mexico sugar trade, the main issues boil down to two aspects: the provisions of North American Free Trade Agreement (NAFTA) and the role of high fructose corn syrup (HFCS), a substitute for sugar, in sweetener markets. In the following section, a summary of this area, focusing on these two aspects, is provided.

NAFTA was implemented in 1994, creating a freer trade environment among

Mexico, the United States, and Canada by eliminating tariffs. Among other regional trade agreements involving North America, Latin and Caribbean countries, it is the least ambitious on paper of the major trade agreements, but it has been the most successful adhering to the negotiated schedule in lowering tariffs (McCoy, 2002). In terms of agricultural trade between Mexico and the United States, many tariffs were eliminated immediately while others being phased out over periods of 5 to 15 years (USDA, 2001c). In addition to a transition period of up to 15 years for certain products, NAFTA has


1









special safeguards to protect import-sensitive crops, including sugar, which are defined under side agreements between Mexico and the United States (USDA, 2001c).

When trading sugar with the United States, the Mexican sugar industry faces two counteracting conditions under the NAFTA regime: increased access to the U.S. market which would facilitate sugar exports at favorable prices; and the pressure of increased imports of HFCS from the United States which have been gaining an increasing share of Mexico's sweetener market since 1994. Under the provisions of NAFTA, both an overquota tariff for Mexican sugar which enters into the United States and a tariff on exported HFCS which enters the Mexican market are regulated in such a way that both tariffs will be reduced to zero by 2008 and 2004, respectively. In addition to the rules of the tariffs, Mexican sugar is subject to U.S. import quota allocations. Mexico is allowed to access two kinds of quotas, depending on Mexico's domestic balance in the sweetener market: if Mexico's sugar production exceeds its sweetener consumption (the sum of sugar and HFCS consumption in two consecutive years -"net surplus sweetener producer status"), Mexico receives 25,000 MT of sugar import quota; and if not, Mexico receives 7,258 MT of quota. Additionally beginning in 2000, the sugar import quota expands from 25,000 MT to 250,000 Mt as long as Mexico satisfies the conditions of a net surplus sweetener producer. Mexico can export over the 7,258 MT quota without attaining net surplus sweetener, but any sugar exported in this scenario would be subject to taxation in the form of tariffs as mentioned above. In 2008 when all the restrictions, i.e. both tariffs and quotas, are lifted, Mexico will have free and unlimited access to the U.S. sugar market.

In Mexico, the sugar industry has played an important role in the economy and the politics of the country. In spite of experiencing drastic economic and political changes,






3


including NAFTA, devaluation, privatization of the sugar cane processing industry in the 1990s, and several changes in the policy regime, sugar production has shown steady expansion over the past 10 years: Mexican sugar production expanded from 3.2 million MT in 1990 to 4.7 million MT in 2000 (COAAZUCAR, 2003a). A significant amount of surplus sugar destined to export has been generated since 1995, ranging from 200,000 MT in 1995 to over 1.1 million MT in 1998 (COAAZUCAR, 2003a). These records may appear favorable; however, Mexico stood to benefit little from NAFTA. From 1996 through 1999 Mexico successfully received a 25,000 MT import quota as a result of attaining net surplus producer status, yet it did not enjoy the expanded quota (250,000 MT) from 2000 through 2002 (USDA, 2003a), the amount equivalent to 20 percent of the U.S. minimum sugar import requirement under GATT, because Mexico's production fell short relative to its sweetener consumption. This indicates that Mexico missed the opportunity to export sugar under-quota even though it generated a significant surplus.

Combined with a slump in production that occurred in 1999 and 2000, the Mexican sugar industry underwent an economic crisis. In September 2001, the Mexican government expropriated 27 of 60 of Mexico's functioning sugar mills in order to maintain the industry (USDA, 2002b). Today, the circumstances surrounding the sugar industry remain unfavorable. At the industry level, many mills are financially vulnerable and suffer from low efficiency of production due to old technology or poor infrastructure. Foreign investment has not been successfully encouraged to provide capital for needed investments in new capital equipment. At the farm level, production efficiency is low due to fragmented farmland, which is a result of the Ejido system (Mexico's agrarian law) and social security program specifically tailored to sugar cane growers. Lack of credit






4


and old technology also contribute to low productivity. Although the price of sugar at the wholesale level has been privatized, the sugar price paid to growers is still controlled by a government agency, and hence farmers have little incentive to grow sugar cane other than to receive social benefits from the Mexican sugar program. At the national level, the Mexican government faces a dilemma between gaining competitiveness in the international market and maintaining social stability through offering employment and financial supports to the livelihood of a large number of growers and related workers. Overall, there has been little benefit to the Mexican sugar industry resulting from NAFTA.

The U.S. sugar market, where a large quantity of sugar is traded by a large number of sellers, has maintained commodity balance by assigning tariffs and import quotas to foreign sellers and maintaining domestic price support through the U.S. sugar program. As a result of GATT, the United States committed to accept a minimum import quota of 1.256 million MT of sugar in 1990; however, the U.S. sugar market has been maintained unchanged until today through successful lobbying efforts by the American Sugar Alliance (ASA), the sugar producers' primary alliance.

In the meantime, HFCS had been gaining its share in the U.S. sweetener market since the early 1970s when commercial production of HFCS became possible by the advancement of wet-milling technology. Today, more than 50 percent of caloric sweetener consumption in the United States is derived from corn syrup including HFCS (Congressional Research Service, Library of Congress, 1999). A similar phenomenon appears to be beginning in Mexico. The implementation of NAFTA resulted in opening the door for HFCS consumption in Mexico where nearly all caloric sweetener









consumption was derived from domestically-produced sugar before 1994. Reflecting this threatening trend of replacing domestic sugar consumption with HFCS, in 1996, the Mexican government imposed tariffs on HFCS claiming that U.S. companies were dumping HFCS at an unfair price and affecting the export volume and value of Mexican sugar. This action evolved into a trade dispute between the United States and Mexico and ended when the WTO panel ruled against Mexico's claim (Garcia Chaves et al., 2002 and 2004). Overall, NAFTA has not brought about significant changes in the U.S. sugar market because the Mexican exporters have been unable to significantly expand shipments to the United States. Rather, attention was poured into issue of HFCS and its immediate impact on the Mexico's sweetener market.

In this study, the direction of U.S.-Mexico sugar trade is examined using quantitative methods, with close attention to issues related to NAFTA and HFCS adoption in Mexico. Demand and supply analyses in both countries and a bilateral trade model using mathematical programming provide insights for the market balance in the future including political implications. Aggregated results from various simulations on the trade model are examined using a game theory analysis to investigate possible policy recommendations through assessing gainers and losers in sugar trade.

Problem Statement

The future outlook for the U.S.-Mexico sweetener market needs to be

quantitatively analyzed in a manner that includes influential factors such as trade agreements under NAFTA: trends in HFCS consumption in Mexico; and other related economic and political issues in the sweetener markets.

Researchable Questions

The study attempts to answer the following set of questions.






6


1. What was the impact of changes in the trade regime in the U.S. and Mexican
sweetener market since NAFTA was implemented in 1994?

2. How much surplus sugar can Mexico generate and how much sugar will cross
the border both under- and over-quota? What will happen after 2008 when all
the restrictions are eliminated on Mexican sugar?

3. What will be the impact of changes in Mexico's market on both the United
States and Mexico? How much influence will HFCS adoption cause in both the
U.S. and Mexican sweetener markets?

4. What will be the impact of changes in U.S. sugar policy on both the United
States and Mexico?

5. Is there alternative sugar policy for the United States to current price support?


Objectives

The primary objective of the study is to develop a bilateral trade model of the U.S.Mexico sugar industry that reflects provisions of NAFTA, as well as related market conditions in order to forecast the outlook of the sweetener market through various simulations, encompassing hypothetical changes in Mexican sweetener situations and U.S. sugar policy.

Secondly, the study aims to provide policy recommendations by examining

aggregated results from these simulations, paying attention to identify gainers and losers under different scenarios. By doing so, the study hopes to illustrate conflicts of interest among the various players in the U.S.-Mexico sweetener market.

Organization of the Study

The remainder of the dissertation is organized as follows. In chapter 2, the sugar industries in both the United States and Mexico are introduced in the context of the sweetener market in each country as well as the integrated market, paying close attention to historical and political perspectives. In chapter 3, the conceptual and theoretical






7


framework employed to analyze U.S.-Mexico sugar trade is presented. In chapter 4 and 5, empirical procedures as well as data set used in the study and the results form the empirical study are presented. Lastly, conclusions and implications for policy are discussed in chapter 6.














CHAPTER 2
SUGAR INDUSTRIES AND SWEETENER MARKETS IN THE UNITED STATES AND MEXICO

Sugar, one of the basic commodities with a long history of utilization, is traded in mature markets in many parts of the world with established business practices and networks. The recent trend towards freer markets in the international trade area has not left the industry unchanged. The sugar industries in the United States and Mexico are not exceptions. They have experienced more changes in the face of this recent trend towards rapid trade liberalization. In fact these two industries have become more economically inseparable than ever before as the sweetener markets in the United States and Mexico have been integrated under North American Free Trade Agreement (NAFTA). In this chapter, sugar industries in both the United States and Mexico are examined in the context of the sweetener market as well as the integrated market, paying close attention to historical and political perspectives.

First, the Mexican sugar industry and sweetener market are introduced with

fundamental characteristics of the structure and government involvement. The status of Mexico as a sugar exporter is also presented in conjunction with Mexico's relation to the U.S. market under the provisions of NAFTA. Next, the development and adoption of high fructose corn syrup (HFCS) is presented. Lastly, the U.S. sugar industry and sweetener market is introduced with emphasis on the current political environment surrounding that market.


8






9


The Mexican Sugar Industry and Sweetener Market

The Mexican sugar industry has a long history of playing an important role in the nation's economy and policy. A large number of small-scale sugarcane growers and antiquated sugar milling facilities still remain as the driving force of an industry under the protection of the government.

Mexico's Sugarcane Production

Most regions of Mexico have a suitable climate for sugarcane production, except for the northern region of the country where the climate tends to be cooler and drier. Sugarcane production is widely spread across the southern and coastal regions of the country under different environmental conditions. Production occurs at altitudes that vary from sea level to over 1,300 meters (4,333 feet) above sea level; annual average temperature from 17 to 35 'C (from 63 to 95 'F); and annual rainfall from 500 mm to over 3,000 mm (from 20 to over 118 inches) rainfall.

In the 2001/02 crop season, total net sugarcane production in Mexico was 41.5 million MT, with a yield of 4.9 million MT of raw sugar (COAAZUCAR, 2003a). Among producing regions, the state of Veracruz has the largest production accounting for 38.5 percent of national production (COAAZUCAR, 2003a). The total area harvested in Mexico was 610,121 ha in 2002 (COAAZUCAR, 2003a), making sugarcane the second largest agricultural crop by area, following only coffee (maize, wheat, alfalfa, beans, and oranges follow sugarcane) in 2002 (SAGARPA, 2003). Although sugarcane is a relatively low maintenance crop, sometimes referred to as the "lazy man's crop," varieties have been developed with higher disease resistance, higher sucrose content, and lower fiber content, yielding better sugar production. In order to avoid poor yields, sugarcane fields are usually replanted every six or seven years (Greene, 1998).






10


The total number of sugarcane growers is reported as approximately 158,000

(COAAZICAR, 2003b), which is equivalent to roughly 2 percent of the total labor force in the agricultural sector (INEGI, 2003). If related workers such as sugarcane cutters, cane-transport employees, factory workers and administrative, and technical and management personnel are included, total employment in the sugar sector exceeds 1,000,000 (Garcia Chaves et al., 2002) and accounts for more than 14 percent of agricultural labor.

Land area per grower ranges from less than 1 ha, which accounts for 3.6 percent of the total sugarcane area, to over 15 ha, which accounts for 17.5 percent of the land, averaging 3.9 ha per grower (COAAZUCAR, 2003b). When the number of growers is allotted to each land size category, a skewed distribution is revealed along the land scale spectrum with many small-land holders and a few large-land holders (Figure 2-1). A large number of small-scale sugarcane growers were created as a result of the Mexican revolution and the sugar program that evolved after the revolution: the communal land (ejido) has been divided and distributed among farmers since the revolution and the Mexican sugar program offers social security and medical services to each grower proving to be a large incentive for farmers to grow sugarcane. Mexico's Sugar Production

There are 60 operating sugar mills located across 15 states in the nation (Figure 22). Sugar mills are responsible not only for milling sugarcane, but also supervising sugarcane cultivation and organizing the harvest. This includes inspecting and advising on cultivation, scheduling harvest dates, pooling and arranging laborers, and providing trucks and drivers for the harvest.






I1


In the 2001/02 crop season, average sugarcane yield was 70.32 MT per hectare in Mexico, yielding 4,872,388 MT of sugar (COAAZUCAR, 2003a). Irrigation is one of the factors which influences cane yield; however, irrigation systems are found only in the area where less rainfall is expected (Figure 2-3): 30 percent of total sugarcane area has no irrigation system and 25 percent has full irrigation system (COAAZUCAR, 2003c).

Harvest is the most labor-intensive part of sugarcane production; the harvest season lasts for about six months starting between November and January and ending in June in most regions, depending on weather and size of enterprise. Harvest competes for grower's labor with other winter crops since many growers are also engaged in production of crops such as maize, vegetables and citrus. Most of the harvest is carried out manually; only 9 percent of total sugarcane processed at mills is harvested by machine; 27 out of 60 mills do not employ a machine harvester at all; however, a cane loader is used in most cases (COAAZUCAR, 2003d). Since the mills own machine harvesters, cane loaders, and trucks, growers do not need to own them; however, it means growers have no means to harvest and sell their sugarcane without the mill's assistance and coordination. Similar situations regarding grower's capacity in harvest are found in other crops such as citrus.

Upon harvest, sugarcane is bought by mills from growers and processed into sugar. Sugarcane quality is vital to the sugar production process; high sucrose content cane leads to high sugar production. Yet, it is often the case in Mexico that trucks endure long waiting times to unload cane due to limited milling capacities. The average wait time observed in 2001 was almost 30 hours across mills (COAAZUCAR, 2003e). The longer trucks wait, the lower the quality of sugarcane becomes. Scattered and fragmented






12


sugarcane fields and poor road conditions also contribute to longer transportation time, and hence decreased sugarcane quality.

Mexico's sugar production was approximately 4.8 million MT, raw equivalent, in 2002, ranking it seventh among all cane sugar-producing nations; Brazil and India are by far the largest cane sugar-producing nations, followed by China, the United States, Thailand and Australia (Table 2-1). Production in Mexico has been increasing for the past few decades (Figure 2-4).

The Mexican Sugar Industry and Government Involvement

Mexico's sugar production accounts for 0.5 to 0.7 percent of its gross domestic

product (Garcia Chaves et al., 2002; Farm Foundation, 2003). Since privatizing in the late 1980s, mills have neither successfully accumulated capital nor renewed their equipment leaving the industry financially vulnerable. In 2001, the Mexican government expropriated 27 mills, which represented approximately 50 percent of sugar production in Mexico. In February 2002, the Government of Mexico announced a National Sugar Policy for 2002 - 2006 which included a series of short- and long- term measures to help Mexico's ailing sugar industry with the main objective of regulating the sugar market and making the sugar sector profitable (USDA, 2002b). Today the sugar industry remains important in Mexico because it is considered crucial for maintaining social stability due to the large number of growers and related workers.

Among the public organizations that deal with the Mexican sugar industry,

Commite de la Agroindustria Azucarera (COAAZUCAR) plays an important role by monitoring and compiling sugarcane and sugar production data at each mill. Although the industry has been privatized, COAAZUCAR is in charge of determining the cane price. It took over the task from the former body, Azucar, S.A. which was dismantled in 1991






13


when sugar mills were privatized. Growers are paid by a fixed portion of the reference sugar price calculated from aggregated sugarcane harvested and processed in the specific mill. Details for the cane price setting formula are shown in equations [2.1] and [2.2] (Garcia Chaves et al, 2004):

Cane price / ton = (KARBE/ ton of cane)*(Price of KARBE)*(0.57) (2.1) KARBE/ ton of cane = (Pol) *(FF)*(FP)*(EBF)*(TF) (2.2)

where KARBE is kilogram of recoverable standard sugar basis (Pol 99.4 percent) for net ton of cane; Pol is polarization of cane (apparent percentage of sucrose in cane); FF is the fiber factor; FP is the purity factor; EBF is mill efficiency; and TF is the transformation factor.

As seen in equation [2.1], currently growers are paid 57 percent of the wholesale price per kilogram of standard sugar. Although in 1991 growers began to be paid according to the quality of cane produced as opposed to solely on weight as decreed in the amendments to Decreto Cafiero, the Sugarcane Growers Law, growers have little incentive to produce higher quality cane. The cane price is capped at 57 percent of average quality cane for the specific mill, not a price reflecting each grower's sugarcane. Thus, a main incentive for Mexican sugarcane growers is to receive social benefits and medical services from the government as opposed to producing quality cane.

The wholesale price of sugar produced at mills has been liberalized since 1997;

however, it still quotes a reference price calculated based on the formula published by the secretariat of Commerce (SECOFI). The price is determined by considering both the recent domestic price and the expected export price, which is the composite of the U.S. and world price (Garcia Chaves et al, 2004), as shown in equation [2.3].






14


Pr = cc Pn + (1- u) - Pe, (2.3)

where Pr is wholesale price per kilogram of standard sugar to be used as the reference for cane payment during the harvest; cc is expected portion of harvest to be consumed nationally (x equals one if expected consumption is greater than expected production); Pn is reference price for standard domestic sugar calculated by comparing the OctoberSeptember average price of the previous year with the current year; (1-u ) is expected surplus as a portion of production; and Pe, is expected export price of sugar, which is calculated by a weighed average of the U.S. price (Contract No. 14) and the world price (Contract No. 11) with corresponding export quantities.

The Instituto Medical y Seguro Social (IMSS), the Mexican Social Security

Institute, provides both pension and medical services to all the employees in Mexico as well as to small farmers who grow sugarcane (Greene, 1998). It is often the case that IMSS clinics are located to the next to the mills (Greene, 1998). Borrel (1991) as well as Buzzanell and Lord (1995) have pointed out this special relationship as a source of inefficiency in the Mexican sugar industry. Mexico's Sugar Consumption

Since Mexico is neither an importer of sugar nor producer of sugar beets, sugar

consumed in the country is derived solely from sugarcane grown domestically. The main use of sugarcane derivatives are shown in Figure 2.5. Sugarcane requires two steps in the refining process to obtain the refined sugar used by households and industries. Raw sugar, which is the product of the extraction process, is either stored or exported to other countries where refinery facilities are available. Mexico consumes two kinds of refined sugar, called standard sugar and white sugar. Standard sugar has a slight impurity,






15


whereas white sugar is as purely refined as ordinary refined sugar found elsewhere. Both kinds of refined sugar are indistinctly consumed by households as well as domestic bulk sugar users such as soft-drink manufacturers and confectionaries. This unique aspect is different from the U.S., where a purer form of sugar, equivalent to white sugar in Mexico, is what most households purchase. Molasses, the heavy dark viscous liquid residue discharged by the centrifugal from which no more sugar can be obtained by simple means (Polopolus and Alvarez, 1991) is utilized in rum making in Mexico.

Mexico's national total sugar consumption was approximately 4,500,000 MT, raw equivalent, in 2001 (FAO, 2003), making it the seventh largest sugar consumption country/ region in the world (Table 2-2). Per capita sugar consumption was 44.6 kg (98.5 lbs.), raw equivalent, in 2000; relatively high among other major sugar-producing countries (Table 2-2). When other kinds of sweeteners such as HFCS are included, the largest per capita consumer is the United States, followed by Cuba, Brazil, Australia, and Mexico (Table 2-2).

Mexico as a Sugar Exporter

Surplus raw sugar is either exported to the world market, primarily to the U.S. market due to higher price, or stored as stock. The magnitude of sugar exports from Mexico depends on the size of surplus determined by domestic production-consumption balance and the quota limitation imposed on all sugar imports entering the U.S. market.

Geographically Mexico holds a sugar export advantage to the United States. One of the main shipping ports in Mexico is Veracruz, located facing the Gulf of Mexico, only 830 miles from New Orleans and 1130 miles from South Florida where sugar refinery facilities are located. Furthermore, the state of Veracruz produces approximately 40 percent of Mexico's domestic production of sugar.






16


Another advantage Mexico attained is the preferred trade conditions under the

North American Free Trade Agreement (NAFTA) implemented in 1994. NAFTA created a freer trade environment among Mexico, the United States and Canada by eliminating tariffs. In terms of agricultural trade between Mexico and the United States, many tariffs were eliminated immediately, while others were scheduled to be phased out over periods of 5 to 15 years (USDA, 2001a). Mexico benefited from exporting its surplus sugar to the U.S. market at a higher price and at a lower tariff rate which will be reduced each year and eventually gives Mexico free and unlimited access to the U.S. market beginning in 2008 (Table 2-3).

The NAFTA agreement is a double-edged sword to the Mexican sugar industry, however, creating counteracting conditions in the Mexican sweetener market: one is the increased access to the U.S. market and the other increased access of high fructose corn syrup (HFCS) from the United States. The quantity of duty-free sugar exported from Mexico is limited by quotas which vary depending on Mexico's balance in the domestic sweetener market - not the sugar market; Mexico receives a larger quota if its domestic sugar production exceeds domestic consumption of sweetener, including HFCS (called "net surplus sweetener producer status") in two consecutive years. At the same time, the agreement gives U.S. HFCS producers free access to the Mexican sweetener market beginning in 2004 as a tariff imposed on HFCS imported from the United States is also being phased out. If Mexico fails to attain net surplus sweetener producer status, it receives a sugar import quota of only 7,258 MT rather than 250,000 MT and thus most of Mexican sugar exported to the U.S. market is subject to over-quota tariffs until 2008 when all the restrictions on Mexican sugar are lifted.






17


Development and Adoption of High Fructose Corn Syrup

Sweeteners are generally classified into two categories, caloric and non-caloric; common sweeteners in the former group are sucrose, invert sugar, lactose, maltose, and sorbitol; and aspartame and saccharine in the latter. Sucrose is found in various forms of sugar such as raw sugar, granulated sugar and brown sugar derived form sugarcane or sugar beets, or in honey and maple sugar. Invert sugar such as dextrose, glucose, fructose, and HFCS are made form starch through chemical processes. HFCS is produced by converting a portion of naturally occurring glucose in starch into fructose through a corn wet milling process (Congressional Research Service, 1999). Lactose, maltose and sorbitol are found naturally in certain kinds of food and give food a sweet taste. Noncaloric sweeteners, sometimes called artificial sweeteners, such as aspartame, are often used for special dietary purposes.

Commercially produced and rapidly adopted since the early 1970s in the United States, HFCS became an important player in the sweetener market among sugar substitutes (Figure 2-6). Production of HFCS has increased from 51,000 MT in 1970 to nearly 8.7 million MT in 2001 (Figure 2-7). HFCS production expanded during the 1980s as a substitute for sugar used in the soft-drinks. Today, about 75 percent of total HFCS and 90 percent of HFCS-55 (55 percent fructose) supplied in the United States are consumed in soft-drink market (Buzzanell, 2002; Congressional Research Service, 1999). HFCS-42 (42 percent fructose), which is roughly 90 percent as sweet as sugar, is used mainly in beverages (44 percent), processed food products (21 percent), and other products including cereal and bakery products (Buzzanell, 2002; Congressional Research Service, 1999). As a result, HFCS and two other corn-derived sweeteners, glucose syrup and dextrose, accounted for 55 percent of total U.S. caloric sweetener use in recent years






is


(Congressional Research Service, 1999). Due to its liquid form, HFCS is considered as a close substitute of sugar, though not a perfect one. Crystalline fructose, fine white crystals of pure fructose is slightly sweeter than sugar and is another potential substitute for sugar; however, crystalline fructose is more expensive than sugar and behaves differently from sugar in most baking and other manufactured food uses thus limiting its use as a sugar substitute (USDA, 1997).

HFCS is predominantly produced in the United States, which accounted for 74 percent of the world HFCS production in 2001(Buzzanel, 2002). HFCS is produced in wet-milling facilities located in corn growing regions in the U.S. HFCS producers outside the United States and their recent production levels are: Japan (766,000 MT), Canada (400,000 MT), Argentina (312,000 MT, estimated), Mexico (291,000 MT, estimated), and European Union (293,000 MT) (Buzzanel, 2002). In the United States, most HFCS is supplied and consumed domestically and only a small fraction is exported to Mexico and Canada, the NAFTA member economies (Figure 2-8 and 2-9). The difference in trade between these two countries is that Canada and the United States exchange a similar amount of HFCS across the border while Mexico is a net buyer. Although the quantity exported to Mexico, 122,800 MT in 2001, accounts for only 1.5 percent of total HFCS demanded in the United States, the quantity accounts for 60 percent of total HFCS export from the United States (Figure 2-10). This amount is equivalent to about 40 percent of Mexico's domestic production ability (291,000 MT, Buzzanell, 2002).

The introduction of HFCS into the Mexican sugar market brought about significant changes in the sugar consumption pattern in Mexico. Although HFCS is traded at a higher price than sugar in Mexico, it has been gaining an increasing share of the Mexican






19


sweetener market over the past several years. The Mexican sugar industry has struggled to supply sugar to the domestic market at a price competitive with HFCS. Prior to 1994, nearly all of the caloric sweetener consumption was derived from domestically produced sugar; however, implementation of NAFTA resulted in opening the door for HFCS consumption in Mexico. As a result, Mexican per capita sugar consumption has decreased slightly since 1991, while per capita sweetener consumption has been increasing (Figure 2-11). Soft-drink manufacturers are believed to account for about onethird of the total sugar domestically demanded (Buzzanell, 2002). Currently, Mexico's HFCS consumption accounts for approximately 12 percent of total consumption of sweetener (approximately 25 percent of indirect sugar consumption) in 2001.

In response to this threatening trend of replacing domestic sugar consumption with U.S. produced HFCS, the Mexican government imposed tariffs in 1996 on HFCS, based on a claim that U.S. companies were dumping HFCS at an unfair price and affecting the export volume and value of Mexican sugar. This action evolved into a trade dispute between the United States and Mexico and ended when the WTO panel ruled against Mexico's claim (Garcia et al., 2002 and 2004). Combined with a slump in production that occurred in 1999 and 2000, the Mexican sugar industry underwent an economic crisis. In September 2001, the Mexican government expropriated 27 of 60 Mexico's functioning sugar mills in order to maintain the industry (USDA, 2001b). Major events in sugar industry history in both Mexico and in the United States are summarized in Appendix A.

The U.S. Sugar Industry and Sweetener Market

Sugar production in the United States comes from two sources: sugarcane and

sugar beets. The main sugarcane production regions are Florida, Hawaii, Louisiana, and Texas where the climate is tropical or semi-tropical. Louisiana and Florida produce






20


approximately 48 and 45 percent of the total sugarcane production in 2001, respectively (USDA, 200 1a). Florida's sugar production from sugarcane made up 50 percent of total cane sugar production in the United States in 2001 (USDA, 2002a), exceeding that of Louisiana (40 percent) due to a higher sugar recovery rate from cane. Sugar beet production regions are classified into four regions: Great Lakes, Upper Midwest, Great Plains, and Far West. The Upper Midwest, which includes Minnesota and North Dakota, produces approximately 47 percent of total sugar beet production; the Far West, which includes California, Idaho, Oregon and, Washington, produces approximately 26 percent of total sugar beet production in 200 1(USDA 2002a). Total sugar production from sugarcane and sugar beets was 4,017,000 short tons (3,615,000 MT) and 4,000,000 short tons (3,600,000 MT), respectively (USDA, 2002a). The proportion of sugar produced in the United States from sugarcane and sugar beets is about equal.

Compared to Mexico, sugarcane production in the United States is regionally

concentrated and highly mechanized as well as vertically integrated. In the case of the Florida's sugar industry, sugarcane is grown areas concentrated in flat land in south Florida. Sugarcane growing activities such as planting, harvesting and transporting harvested crop are fully mechanized, unlike Mexico. Six raw sugar mills, which are located near the sugarcane fields, possess an average daily processing capacity of 20,750 tons of sugarcane (Alvarez and Polopolus, 2002a). Two sugar refineries are located adjacent to two sugar mills. U.S. Sugar, the country's largest sugar producer operating in Florida, owns a fully integrated cane sugar refinery facility that manages not only sugar refining but also packaging and warehousing. With the facility built next to the existing






21


sugar mill, this company is able to produce bagged sugar from sugarcane in a single plant location.

Sugar is one of several commodities protected in the Farm Bill, which includes

rice, cotton, dairy, tobacco, peanuts, grain such as wheat and corn, and soybeans (Alvarez and Polopolus, 2002b). The sugar program operates through a loan program and market stabilization price (MSP) without production or acreage restraints, differentiating it from other programs that include target prices or deficiency payments along with export enhancement programs (Alvarez and Polopolus, 2002b). Loans are issued as nonrecourse loans' and are available to processors of domestically grown sugarcane at a rate of 18 cents per pound and to processors of domestically grown sugar beets at 22.9 per pound of sugar, respectively (Haley and Suarez, 2002). Since there are no production restraints, import quotas and tariffs are the main policy instruments utilized to comply with the provision that the program has to operate "no cost" to the government (Alvarez and Polopolus, 2002b).2 In 1990, the United States committed to accept a minimum import quota of 1.256 million MT of sugar as a result of GATT.

The U.S. sweetener market has maintained a stable commodity balance, unlike in Mexico, even after NAFTA was implemented and during the trade dispute with Mexico over HFCS dumping. One of the reasons is the successful effort by the American Sugar Alliance (ASA), the sugar producer's alliance in the United States, which lobbies for the U.S. sugar program. The ASA is a strong coalition that includes sugarcane producers, sugar beet producers, and sugar processors, as well as corn producers and HFCS As long as the raw sugar tariff-quota is set higher than 1.5 million short tons (Haley and Suarez, 2002).
2 The "no cost" provision was not included in Food and Agriculture Improvement and Reform Act (FAIR) in 1996 (Haley and Suarez, 2002).






')2


manufacturers (Moss and Schmitz, 2002). The other entity in the U.S. sweetener market is Coalition for Sugar Reform (CSR), the industrial sugar user's coalition. CSR opposes U.S. sugar policy, but has been unsuccessful at bringing a lower sugar price to the market from which industrial sugar users as well as consumers would benefit. Many studies have shown that producers clearly gain while consumers and industrial users lose in the U.S. sugar program: an analysis conducted by the General Accounting Office (GAO, 2000) indicates that food manufacturers could be substantial gainers from elimination of the sugar program (Moss and Schmitz, 2002).

A history of sugar and HFCS prices are shown in Figure 2-12. Both sugar and

HFCS prices exhibit a continuing declining trend. The decline in both U.S. domestic and export prices of the HFCS price is due to sophistication of wet-milling technology in combination with decreases in the tariff schedule under NAFTA for the latter. HFCS has been marketed at a lower price than raw sugar in the U.S. market. Although the U.S. HFCS industry sells at a lower price than sugar, it still benefits from the sugar program because the price of sugar is maintained higher than it would be without the program. This structure confirms why the HFCS industry has been supporting the sugar program as a member of ASA; however, an opportunity for the HFCS industry to increase marketing overseas such as in Mexico may weaken the incentive to support the sugar program: if the sugar price drops as a result of Mexican sugar flowing into the U.S. market through the large import quota promised under NAFTA, the industry has to weigh both costs and benefits to the industry (Moss and Schmitz, 2002).

Although the economic implications of the price differential between sugar and HFCS are identified, estimation of a quantitative relationship for HFCS supply is not






23


straightforward. According to the analysis by Moss and Schmitz (2002), the relationship between sugar and HFCS prices has changed over time: HFCS price responded to wholesale sugar price from 1983 to 1996, but not from 1997 to 2001. Another analysis by Evans and Davis (2002) indicates that the estimated cross price elasticity of HFCS with respect to sugar demand was found to be insignificant. This implies that the HFCS price is set below the sugar price in order to attract bulk sweetener users but its behavior remains ambiguous and arbitrary. Furthermore, other research reported that HFCS price is not correlated with the price of corn (Offenbach, 1995), but others found that the HFCS price responded to the price of corn from 1983 to 1996 but not from 1997 to 2001 (Moss and Schmitz, 2002). Given the insignificant relationship between corn price and HFCS supply mentioned above, the indirect impact of changes in corn production on HFCS price would likely be small; however, any drastic changes in corn program or large changes in corn export may eventually affect HFCS prices. Corn statistics such as recent corn production for selected countries, industrial use of corn in the U.S., and the corn price in the U.S. are illustrated in Appendix B.






24


120,000 100,000
80,000 - - --- - -
C 60,000 o 40,000 . 20,000
0
V V V V V V V V V V V V V V V A
M r- 00 0
Individual Farm Size

re Growers


Figure 2-1. Distribution of Individual Farm Size of Mexican Sugarcane Growers Source: COAAZUCAR, 2003b









Tamaulipas


Sinaloa " San Luis Potosi
Yucatan
Nayarit ..""-Campeche .". Veracruz Jalisco .".. Tbsc
Colima Tabas.co"
Michoaca
M Chan Puebla """Quintana
Mexico City Roo
Morelos
Oaxaca Chiapas

- Indicates a processing facility (mill)


Figure 2-2. Map of Mexico's Sugar Producing and Processing States.






25


C

a
a-


100 90 80 70 60 50
40 30
20
10
0


0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000
Annual Rainfall [mm]


Figure 2-3. Annual Rainfall and Irrigation Rate in Sugarcane Fields Source: COAAZUCAR, 2003c


Table 2-1. Cane Sugar Production in Selected Countries, 1997-2000 Average.
Country Area Sugar Cane Sugar Sugarcane
Harvested Production Yield Yield Recovery
(1,000 ha) (1,000 MT) (MT/ha) (MT/ha) Rate (%)
Brazil 4,914 18,339 68.28 3.73 5.47
India 4,092 17,233 69.41 4.21 6.07
Cuba 1,086 3,814 32.69 3.51 10.77
China 1,064 6,532 75.08 6.14 8.18
Pakistan 1,029 3,064 46.57 2.98 6.38
Thailand 923 5,468 56.16 5.92 10.55
Mexico 627 4,807 76.46 7.66 10.02
Australia 409 5,281 90.94 12.91 14.22
United States 360 3,811 88.00 10.59 12.03
Colombia 391 2,227 86.25 5.70 6.61
Philippines 324 1,796 81.06 5.55 6.87
South Africa 315 2,684 72.57 8.54 11.76
World 19,307 90,340 65.12 4.68 7.19
Source: FAO 2003, USDA 2002a, USDA 2003b


--U


U

U






26


6.000.000 -
5,000,000
4,000.000
3,000,0002,000,000
1,000.000t 01
0 00 M' 0- W l Cl - n ' ' 0
oo 00 C' C' C' C C' C' C C ' C , Year


Figure 2-4. Mexican Sugar Production, 1988-2002 Source: COAAZUCAR 2003a



Sugarcane

1- Cane juice
Raw sugar ns s*- International market or stored

L Refined sugar

Standard sugar White sugar (purer) n=*Domestic market Other grades M~olass es


t Alcohol (rum)
Other by-products


Figure 2-5. Main Use of Sugarcane Derivatives in Mexico
Source: Adapted from Polopolus and Alvarez, 1991






27


Table 2-2. World Sugar Consumption in 2000
National total Per capita
consumption' consumption'
Sweetenerb Sugar Sweetener" Sugar Country region (1,000 MT) (1,000 MT) (kg) (kg)

India 26,234 18,101 25.59 17.66
United States 21,221 9,371 74.22 32.77
European Union (15 countries) 18,960 14,370 50.26 38.10 China 11,805 11,028 9.13 8.53
USSR, Former Area of 11,771 11,407 40.57 39.32
Brazil 10,007 9,620 57.99 55.75
Mexico 5,100 4,476 50.81 44.60
Japan 3,741 2,327 29.38 18.27
Philippines 2,135 1,975 27.68 25.61
Thailand 1,937 1,924 30.47 30.26
Canada 1,297 1,128 41.83 36.37
Australia 1,060 922 54.81 47.67
Cuba 711 710 63.24 63.18
World 168,632 133,401 27.49 21.75
a: Production + net import + change in stocks
b: Sugar (raw equivalent) plus other kinds of sweeteners
Source: FAO, 2003


Table 2-3 Quota and Tariff Schedule Imposed on Mexican Su2ar Exported to the U.S.


U.S. Import Quota (MT)
Year Mexico as a net surplus Mexico NOT as a net Over-Quota Tariff
producer surplus producer (raw cane, cents/pound)
1994 25,000 7,258 16.00 (Base)
1995 25,000 7,258 15.20
1996 25,000 7,258 14.80
1997 25,000 7,258 14.40
1998 25,000 7,258 14.00
1999 25,000 7,258 13.60
2000 250,000 7,258 12.09
2001 250,000 7,258 10.58
2002 250,000 7,258 9.97
2003 250,000 7,258 7.56
2004 250,000 7,258 6.04
2005 250,000 7,258 4.53
2006 250,000 7,258 3.02
2007 250,000 7,258 1.51
2008 and beyond 00 0.00
Source: USDA, 1999.







28


100.C 90
80 70 60 50
2 40 30

20 10
0
I Or-~ 00 C, 0 m 'I V' N 00 0
- N N N N 0000 00 00 00 00 00 00 00 00 O~0' 0

Ye ar


--- Total HFCS use per capita -*-- Refined sugar use per capita


Figure 2-6. U.S. Refined Sugar and HFCS Use per Capita Source: USDA, 1993


10,000
- 9,000 -8,000 7,000
6,000 5,000
4,000S 3,000 - -
2,000 3 1,000
0
0 1 r' 0 00 0 ' It~ ' 00 0 e'q 1' 0 00 0
r- N N N 00 00 00 00 00 C' "IC' 0 C'

Year




Figure 2-7. U.S. HFCS Production Source: USDA, 1993 and 2001 a.







29


Import mostlyy fromCanada) -1.6%


U.S. production
98.4%


Figure 2-8. U.S. HFCS Supply
Source: USDA, 2002a







O Export to 0 Export to
Mexico Canada
1.5% 1.0%c








[ Donvstic
consumption
97.5%


Figure 2-9. U.S. HIFCS Export
Source: USDA, 2002a







30


400.0 350.0 300.0 250.0
200.0 150.0 100.0
50.0 0.0


CII M '
01' <71 C C, C, C,


'' 0 a, C,
ON 01


cli


Year

--*-U.S. total export -W- U.S. export to Mexico



Figure 2-10. U.S. HFCS Export, 1992-2001
Source: USDA, 2002a


110 105 ' 100


95 > 901


85 80


HFCS


r- 0' 0' 0' 0- "' M' -It cc0 Year



-U- Consumption of sugar per capita -0- Consumption of sweetener per capital


Figure 2-11. Consumption of Sugar and HFCS per Capita in Mexico
Source: Garcia Chaves et al., 2004


0q


-





- - -


-







31


70 - - -- ---- - - ---- ____ _____0
-60

~40
-3O

S20 U
*~10

0
1994 1995 1996 1997 1998 1999 2000

Year

-1- World rawsugar price Ia US rawsuW price, dly fee paid

--US spot price for ICS-42, Mdost markets -6 Unit valoeof HKSexpotedto Mxico


Figure 2-12. Prices of Sugar and HFCS, 1994-2000 Source: USDA, 2001a














CHAPTER 3
CONCEPTUAL AND THEORETICAL FRAMEWORK

In this chapter the conceptual and theoretical framework employed to analyze trade between the United States and Mexico in the sugar market is discussed. In the conceptual framework, factors that influence the market and the trade system are specified and linked. Based on the conceptual framework, theoretical foundations are established in two parts: an analysis of the sugar market for each country and an analysis of the market balance in the U.S.-Mexico bilateral sugar trade system.

Conceptual Framework

The sugar market, one of the oldest and most common agricultural commodity

markets, is built upon sugarcane and sugar beet production and the resulting production of processed sugar from these raw materials. Being an essential commodity for a daily diet, sugar has been traded across borders for a long period of time. More recently, there have been major changes to the trading pattern due to the emergence of an alternative in the market. High fructose corn syrup (HFCS), now the most widely adopted sugar substitute, expanded the sugar market into a sweetener market. This is particularly true in the case of the United States, where HFCS now occupies nearly half of the sweetener market, and is becoming the case in Mexico as a result of the two markets becoming more closely linked by the North American Trade Agreement (NAFTA). Understanding this linkage between the United States and Mexico holds important clues to analyze economic and political impacts in the sweetener markets. To illustrate this linkage, the


32






33


flow of the goods in each market and factors that influence markets are illustrated in Figure 3-1.

In Mexico, sugar distributed in the market is supplied entirely by domestic sugar production, which is also provided entirely by domestic sugar cane growers. Sugarcane production depends on inputs such as labor from Mexican farmers, land, and agricultural chemicals, technology to grow and harvest sugarcane, and other factors such as weather and government support. Grower's behavior is also influenced by their relationship with mills. Sugar processing depends on inputs such as labor from mill workers, harvested sugarcane, and energy such as petroleum to run the facilities, technology to produce sugar from sugarcane, infrastructure, and government support. Note that government plays an important role to support both grower's and mill's activities. HFCS is primarily supplied by domestic production and the remained is imported from the United States. Sugar is consumed by households and bulk users (such as soft-drink manufacturers), while HFCS is consumed only by bulk users. Determinants of sweetener demand are income, tastes and preferences, price and other factors such as population growth.

In the United States, domestic sugar supply is derived from domestic production and supplemented by import from various origins. Unlike Mexico, domestic sugar is derived from both sugarcane and sugar beet production. Sugar mills produce sugar not only from domestically produced sugarcane and sugar beets, but also from imported raw sugar. HFCS is supplied only from domestic corn sweetener manufacturers and supplies roughly half of caloric sweetener consumption in the United States. HFCS is also exported, and Mexico is one of the main destinations.






34


The U.S.-Mexico bilateral trade system involves sugar exported from Mexico to the United States and HFCS exported from the United States to Mexico. Mexican sugar is exported to a variety of markets, but the primary destination is the U.S. sweetener market. Once Mexico exports the amount of sugar to the United States allowed under the U.S.import system, excess sugar is exported to the world sugar market. In total, more than 30 countries exported sugar to the United States under the allocated tariff-rate quota in 2002 (USDA, 2002a). Under the conditions of NAFTA, Mexico could be allocated 250,000 MT of the U.S. import quota given certain conditions (successful attainment of net surplus sweetener producer status), which would increase Mexico's share of the total U.S. import sugar quota allocation to 20 percent. A two-country trade model with quota imposed by a large country importer is illustrated in Figure 3-2 (a). By imposing a quota on Mexican excess supply of sugar (ESmx), the quantity exported to the United States is limited to Qq instead of Qf. Consequently, the price for imported sugar in the United Staes increases to Pq, us and the price for exported sugar from Mexico decreases to Pq, MX. This causes welfare loss in Mexico (area abcd) due to lower sugar export price and generates quota revenue in the United States (area efgh) collected by the U.S. quota holders or the government. Beginning in 2008 Mexico will have free access to the U.S. market. The same trade model without the quota system is illustrated in Figure 3-2 (b). Trade without distortion brings an increase in welfare in both countries. For simplicity, the producer price support policy is excluded from both Figures (a) and (b).






35


Theoretical Framework

Sweetener Market Analysis

The sweetener market in a country that consumes both sugar and HFCS can be expressed as follows:

SUGAR =fl (PSUGARZI) (3.1)

Q dHFCS =f2 (PHFCS,,Z2) (3.2)

SUGAR = h, SUGARA, WI) (3.3)

QAHFCS = h2 (PHFCS, W2) (3.4)

QdSUGAR SUGAR (3.5)

QdHFCS = QHFCS (3.6)

where Qd is aggregate quantity demanded, Q'is aggregate quantity supplied, P is a price, Z and W are vectors of other factors that influence aggregate demand and supply of sugar or HFCS, respectively. Equations [3.5] and [3.6] depict market clearing conditions for each commodity.

Sweetener demand

Sweetener demand is defined based on consumer demand theory derived from

utility maximizing behavior. Following Varian's demonstration (1992), aggregate sugar demand is derived from maximizing utility of aggregated consumers, including industry sugar users, by purchasing sugar:

max u (QSUGAR, QX)

s.t. QSUGAR *PSUGAR + QX*PX = n (3.7)

where Q is the quantity consumed, X represents all other goods consumed, P is a price, and m is national income. By solving maximization problem, the aggregate demand function is expressed as:






36


QSUGAR SUGARA, PX, in). (3.8)

Since the demand function is homogeneous of degree zero, it can be normalized by either price. Normalize by Px:

QSUGAR (PsUGARPX, mPX), (3.9)

and thus the aggregate demand for sugar is expressed as a function of real price of sugar and real income. By treating real income as one of demand shifters, aggregate demand is expressed as shown in equation [3.1]. Similarly the aggregate demand for HFCS can be expressed as a function of HFCS price and a vector that affects aggregate HFCS demand. Note that in reality, HFCS is consumed directly by industry users and its price is observed by them; households as final consumers consume HFCS indirectly through HFCS-contained goods.

Sweetener supply

Sweetener supply is defined based on firm supply theory derived from profit maximizing behavior. Although in reality sugar supply consists of two steps of production in reality, sugarcane and sugar production, a simple aggregate sugar supply equation at industry level is derived rather than two equations. This is suitable for two reasons: a single industry supply equation makes simulations in the bilateral trade model in the following procedure simple, and it is the industry supply price that the government is interested in supporting.

Suppose the industry faces a cost function given by:

C = C (QSUGAR, V ) (3.10)

where V is a set of other factors including input prices. The industry maximizes profits, assuming the market is competitive (the industry as price taker):






37


max /T = PSUGAR * QSIUGAR - C (QSUGAR,, V)- (3.11)

The first-order condition gives the supply relationship for the industry by equating output price to marginal cost. Expressed in a general form with a vector of other factors that influence aggregate supply:

Q'SUGAR SUGARR W1). (3.12)

Similarly HFCS supply from industry (QiHFCS ) is theoretically expressed as a

function of the HFCS price and the supply shifters; however, the quantitative relationship is expected to be insignificant: HFCS pricing is ambiguous and arbitrary as discussed in the previous chapter and thus the estimated relation does not likely represent the associated marginal cost curve.

U.S.-Mexico Bilateral Sugar Trade System

A spatial equilibrium model is used to portray the U.S.-Mexico bilateral sugar

trade system, following the Takayama and Judge formulation (1964). The model provides the optimal equilibrium price as well as quantities demanded and supplied at the equilibrium through maximization of welfare in each region, i.e. the sum of the consumer and producer surplus, given the demand and supply equations and the transportation cost among regions. Let Si (Yi) represent the inverse supply function (price-dependent form) for sugar in region i; Yi represents the quantity produced in region i; Dj (Qj) represents the inverse demand function (price-dependent form) for sugar in region j; and Qj represents the quantity consumed in region j.

i I J I J
Max D,(Q,)dQ- S(Y,)dZ, - TR X - TTRXX, (3.13)
,=1 r=1t1=1 j=1 i=1 ,=1


s.t. (X +XXi)iY,,Vi=1,..,I (3.14)






38


(X, + XX ) QVj =1. J (3.15)


X Quota,, IVi 1, Vj= (3.16)


Y,Q,, Xj and XXj 0

where X1 is the quantity of sugar shipped from supply region i to demand region j under quota limit, TRj is per unit transfer costs associated with Xj, XXii is the quantity of sugar shipped region i to demand region j over quota limit, TTRJ is per unit transfer costs associated with Xj. TRI0 is a compound transfer cost that includes per unit transportation cost (Tij) and per-unit tariffs imposed on exported sugar under quota limit (Tar ). Similarly, TTRj is a compound transfer cost that includes per unit transportation cost (T1) and per-unit tariffs imposed on exported sugar over the quota limit (over-quota tariff, OQTary ).

TR1 = Tj + Tar1 (3.17)

TTRj = T + OQTar0 (3.18)

Let specific inverse linear demand and supply functions for United States and Mexico be defined as follows:

PDUS IU1 + IU2*QDUs + ShifterD us (3.19)

Psus = IUUI + IUU2*Qsus + Shifters us (3.20)

PD T IMI + IM2C*Q x + ShifterDMx (3.21)

PsMx = IMM + IMMI,*QsMx + Shiftersmx (3.22)

where P D QD represent the price and the quantity demanded in each country, respectively; P5 and Qs represent the price and the quantity supplied in each country, respectively; ShifterD and Shifters represent the demand and supply shifters in each






39


country, respectively; IU1, 1UU1, lM, and IMM1 represent the intercepts in each inverse linear equation; and IU7, IUU2, IM2, and IMM2 represent coefficients associated with each quantity variable. The objective function of the U.S.-Mexico bilateral sugar trade system is expressed with equations [3.19], [3.20], [3.21], and [3.22] with constraints such as necessary conditions, i.e. demand-incoming shipment and supply-outgoing shipment balance as well as those specific to the bilateral trade model such as Mexico's quota allocation under NAFTA and U.S. price support:

Max J (IU; + IU2*QDus + ShifterDus) dQDuS


+ f (IMI + IM2*QTcMx + Shifter0Mx) dQTcmx - (IUU] + IUU2*QSus + Shiftersus) dQsus


-f (IMM1 + IMM2*QSMx + ShifterSMx) dQsmx

- (TMX, us)*X MX. us - (Tus, Mx )*X US, MX

- (Tmx, us + OQTarx, us)*XX Mx, us

- (TROw, US + PROW)*XROW, US

+ PROW *XMX, ROW (3.23)

where US, MX, ROW represent Mexico, the United States, and the rest of the world, respectively; Ti; is per unit transportation cost from i toj; OQTarMx, us is per unit per unit over-quota tariff imposed on Mexican sugar shipped to the United States; and PROW is world price of sugar. Note that the transportation cost within the country is assumed to be zero. The over-quota import from the rest of the world to the United States (XXROW, us) is ignored because of the high tariff rate imposed on the sugar from the rest of the world,






40


compared to sugar from Mexico: the first sugar that enters into the United States overquota must be from Mexico. The transfer cost of sugar from the rest of the world to the United States includes the price of sugar, i.e. sugar from Mexico and the rest of the world compete with each other to enter the U.S. market: the one with lower transfer cost enters the market first. The last term in the equation [3.23] considers Mexico's sales of sugar to the rest of the world. Equation [3.23] is rewritten in quadratic form as follows:

Max (1/2)* IU2*(QDUS )I + (1U1 + ShifterDUS)*QDUS

+ (1/2)* IM2*(QcMx )2 + (IM + Shifter"Mx)*QCMx

- (1/2)* IUU2*(Qsus )2 + (IUUI + Shiftersus)*Qsus

- (1/2)* IMM2*(QSMX )2 + (IMM + ShiftersMx)*Q smx

- (TMx, us) *X Mx, us - (Tus, MX )*X us, MX

- (TMx, us + OQTarx, us)*XX Mx, us

- (TROW US + PROW)*XROW, US

+ PROW *XMX, ROW (3.24)

Constraints are defined to balance the quantities shipped with the quantities

demanded as well as supplied; to impose a quota on imported sugar from Mexico; and to impose a quota on imported sugar from the rest of the world, which is equivalent to the U.S. minimum sugar import quota required under WTO agreements:

QDUS - X us, US - XMX, US - XX MX US - X ROW US 0 (3.25)

Q MX - XMX, MX - X us,MX 0 (3.26)

- Qsus + X us, us + X us, X 0 (3.27)

- Qs MX + X MX, MX + X MX, Us + XX MX, us + X MX, ROW 0 (3.28)






41


MX, s - Quota < 0 (3.29)

XROW, Us + XXmx, us - USMin > 0 (3.30)

where Quotais the quota Mexico is assigned according to the NAFTA provisions and USMin is the quota allocated to the rest of the world (the U.S. minimum sugar import quota less the quota assigned to Mexico). Equation [3.25] and [3.26] satisfy domestic demand; equation [3.27] and [3.28] regulate the quantities shipped from supply regions; and equations [3.29] and [3.30] define quotas imposed on Mexican sugar and from the rest of the world.

The implications of the model can be examined by writing the first-order

conditions obtained through using Kuhn-Tucker theorem. From equations from [3.24] to [3.30] the Lagrangian form (L) is given as:

L = (1/2)* IU2*(QDus)2 + (IU + ShifterDus. )*QDUS

+ (1/2)* IM2*(QTC MX)2 + (IM, + Shifter DMX)*QTCMX

- (1/2)* IUU)*(Qsus)2 + (IUU + Shiftersus)*Qsus

- (1/2)* IMM2*(Qsmx)2 + (IMM + Shifters Mx)*Q sx

- (TMxu s)*XMx, us - (Tus, Mx )*X us, Mx

- (TMx, us + OQTarMX, us)*XX Mx, us

- (TRow, US + PROW)*XROW, US

+ PROW *XMX ROW

+ Vus (X Us, Us + X MX, us + XX MX. US + X ROW. US - QDUS)

+ VMx (X MX, MX + X us, MX - QcMX)

+ Wus (Qsus - X us, us - X us, x)






42


+ WMX (Qs MX - XMX, MX - XMX, us - XXMX. us - XMX. ROW)

+ A (Quota - XMX, us)

+ o(X ROW, US + XXMX. us - USMin) (3.31)

where V, W, A, and a are Kuhn-Tucker multiplier associated with each constraint representing the imputed marginal value of price of sugar demanded, supplied, that of Mexican sugar exported under-quota and that of over-quota, respectively. Note that A is positive in sign and ois negative, reflecting the way their associated constraints are defined. Kuhn-Tucker conditions are expressed as follows:


D =IU2*Q Us +IUI +Shifter us - VUS 0 0 DUS 0, QDUS D 0
aQus uQLs

(3.32)


aL = -(IUU2*Qsus +IUUi+Shiftersus)+Wus *s 0, as *QS 0
US 3us

(3.33)


=IM2*QTCMx +IMI+ShifterDmX -VMX 0, T *QTCMX 0, QTC 0
aQmx QMX
(3.34)

*QS -0QM 0
as = (IMM2*Qsx +IMM M+Shiftersmx +WMX : 0, *Qsx =X 0, QsMx > 0 (3.35)


= Vs - Ws 0, L *X s, us =0, X U us s 0 (3.36)
axlSL, s ax us.us

a =VMX -Wmx 0, a *X MX MX =0, X MX, MX 0 (3.37)
aX MX.MX ax MXX






43


aL aL
= -TMX. us + Vus -WMx - A 0, *X x (s =0, XMX uis 0 aX TXJ s

(3.38)

= Tus, mx + VUs -WMx 0, *s =X USx 0
'SJMX SXM XS

(3.39)

aL aL
= -TMX, us - OQTarmx, us +Vus -WMx + c-< 0, *XMX, us =0, JXXMXI'S ax MYXUS


XMX, us 0 (3.40)


= -TROW, US -PROW + VUS +o* 0, *X ROW, US =0, X ROW us 0 aX ROW S OX RW.1S

(3.41)

DL ___L= PROW - WMX 0, aX *X MX, ROW =0, XMX, ROW 0
X MX ROW MXR0W

(3.42)

By solving equations above, the following conditions must hold at equilibrium:' VUS = pD US = WUS = PS us (3.43)

VMX = pDMX = WMX = pSMX (3.44)

PD US ! PS MX + TMX US + A (3.45)

pD MX 5 US + TUS, MX (3.46)

PD US MX + TMX, US + OQTarx, us -a (3.47)

PD us PROW + TROWu - (3.48)

PROW PDMX (3.49)


Assuming non-zero production and consumption in both the U.S. and Mexico.






44


The complementary slackness conditions indicate that if Mexico exports sugar to the U.S. under-quota (XMx, us > 0), then the demand price in the United States should not exceed the value of Mexican exporting sugar, which is equivalent to the sum of the supply price, transportation cost and the marginal value of exporting sugar under-quota (equation [3.45]). By the same token, if Mexico exports sugar to the United States overquota (XXux, us > 0), then the demand price in the United States should not exceed the sum of the Mexican supply price, transportation cost, tariff imposed on over-quota sugar and the marginal value of exporting sugar under-quota (equation [3.47]). The inequality of the prices expressed in equation [3.49] accords with reality. If both over-quota export from Mexico (XXmx. Mus) and import from the rest of the world (Xmx, us ) are greater than zero, the following must also hold from equations [3.47] and [3.48].

PSMX - PROW = tROW. US - MX US - OQTarMx us (3.50)

This equation implies that when the price difference between Mexico and the rest of the world is equal to the difference in transfer cost (transportation cost and tariff), both over-quota export from Mexico and import from the rest of the world occur at the same time. In other words, Mexico would export over-quota only if the transportation cost from the rest of the world is high enough to justify Mexico to do so.





Figure 3-1. Conceptual Framework for the Analysis in this Study


Relationship with mills Weather

A .... ........................

Growers Irrigation .n.u.World Sugar.
Inpu Wold S-arU.S. Government
..........-- .............-....-- M arket

Farm land
Mexican Sugarcane U.S. Sugarcane
Productionduction ..... . ............. ..............................W orld Sugarcane
Mexican Government Raw ugar Production
... ..... ..U.S. Sugar Beet

Technology..Mexican.Sugar.Quota System Raw Sugar Production
TechnologyMeiaSga
Production Process ---- .......:
..............----...... Raw Sugar.
......U.S. Sweetener
Infrastructure
- NAFTA Mexican Sweetener . - --- - - --.-.--7 Market
..... ......... H FC S
IncomeU.S. HFCS ......................... Production

..............----- - Investm ent
PopultionMexican HFCS
Population Pouto
Pro auction
......................................... U .S . C o rn

:Goods -...... : Influences Production










P


SNMX


a b

d c


Dmx


Q


P

S us e f QDus




Qq


Mexican Market


Mexico-U.S. Market


U.S. Market


(a)


P


SMX


DMx


Q


Mexican Market


P


ESMX


EDus





Qf

Mexico-U.S. Market


Q


S us





Li US


I Q


U.S. Market


(b)


Figure 3-2. Two-country Trade Model (a) with Quota System and (b) without Quota System.


46


P


I


Pq, Us Pq, mx


ESMX


EDus


Q


EDus'


Qq Qf


Q














CHAPTER 4
EMPIRICAL MODELS AND DATA SOURCE Based upon theoretical framework developed in chapter 3, this chapter focuses on the empirical procedures and data set used in the analysis of the U.S.-Mexican sweetener market. The empirical model consists of three components: (1) demand and supply analysis models for both the U.S. and Mexican sugar markets; (2) a bilateral sugar trade analysis model; and (3) game theory analysis. These models are ordered sequentially so as to utilize the results from the former. After the trade model is introduced, assumptions associated with the model, the methods of model calibration, and simulated scenarios are presented. Results from simulations on the trade model are aggregated to assess policy recommendations. Lastly, an overview of the data set used in the empirical models is provided.

Empirical Models

U.S. Sweetener Demand Model

Regression analysis is used to estimate sweetener demand utilizing quarterly timeseries data. The model is specified as a double-log (natural log) form that allows interpretation of estimated coefficients as elasticities associated with each variable. Based on the basic form of demand equation shown in equation [3.1], the U.S. sugar demand equation is specified as:

In(QSUGA R ,) = U1 + U)*In(PDSUGAR, d + U3*In(GDP,) + U4*In(POP,)

+ U5*QT R1+ U6*QTR2+ U7*QTR3+ U8*DHFCS + e, (4.1)

where QD is quantity of sugar demanded in each quarter in year t [1000 short tons], PD is


47






48


real retail price of refined sugar (deflated by Consumer Price Index (CPI)) [cents / pound], GDP is real per capita GDP (deflated by CPI [US$]), POP is population, QTRs are dummy variables representing quarters of the year, DHFCS is dummy variable for availability of high fructose corn syrup (HFCS) in the sweetener market (DHFCS=1 after t=1975), and e,is an error term. In theory, the price of HFCS would be preferred as prices of substitutes are expected to influence quantity demanded, however, limited availability of the data made this impossible, therefore, the dummy was used. It is expected that the elasticities associated with price (U2), the first quarter (U5) (compared to the omitted fourth quarter), and HFCS availability (U) to be negative and the elasticities of the remaining variables to be positive. Price elasticity is expected to be inelastic based on previous research. Previously reported own-price elasticities for sugar are -0.141 by Lopez (Lopez, 1990) and -0.73 by Petrolia and Kennedy (Petrolia and Kennedy, 2002). In the latter estimation, the U.S. wholesale refined beet sugar price reported at the Midwest Markets was used. In the process of regressions, serial correlation is anticipated and corrected by the Yule-Walker Method with appropriate lags assigned. Mexican Sweetener Demand Model

Estimation of Mexican sweetener demand was conducted in a similar manner to

that of the U.S. demand. The main difference was that the demand for two kinds of sugar is estimated separately in the Mexican model. Traditionally the demand for sugar in Mexico has been estimated by regressing the total consumption on sugar price and percapita income. Borrell (1991) estimated the price elasticity be -0.004 and the income elasticity at 0.5. This model was appropriate in early 1990s because there was no HFCS consumption in Mexico. In order to account for the entry of HFCS into the Mexican sweetener market, the demand for sweetener is estimated by disaggregating sugar






49


consumption into: sugar consumed by households, referred to as "direct consumption"; and sugar consumed by bulk sugar users, referred to as "indirect consumption". Only indirect consumption of sugar is expected to be influenced by HFCS consumption as they are the only consumers of HFCS. In addition to estimating sugar demand, it would have been preferable to directly estimate HFCS consumption, but this was precluded because of the relatively short time series available. Instead, total consumption of sweeteners (the sum of sugar and HFCS consumption) is analyzed by regressing on sugar price and other variables. A dummy variable that represents the availability of HFCS in the market is also added to the regression forms of indirect consumption of sugar and total consumption of sweeteners where substitution between sugar and HFCS occurs in the market.

In(QDCSUGAR, )= MI + M2*In(PDSUGAR, d) + M3*In(GDP,) + M4*ln(POP,)+ ut

(4.2)

In(QICSUGAR ,) =M5 + M6*In(PDSUGAR, ) + M7*In(GDP,) + Ms*In(POP,)

+ Ug *DHFCS + v, (4.3)

In(QTCSUGAR, ) = M10 + MII*In(PDSUGAR, ) + M12*In(GDP,) + M13 *In(POP)

+ U14 *DHFCS + w, (4.4)

where Q Cis direct consumption of sugar [MT], QIC is indirect consumption of sugar [MT], QTC is total consumption of sweeteners [MT], PD is real retail price of standard sugar (deflated by CPI [pesos / Kg]), GDP is real per capita GDP (deflated by CPI [pesos]), POP is population, DHFCS is dummy variable for availability of high fructose corn syrup (HFCS) in the sweetener market (DHFCS=1 after t=1992), and u,,v,,w, are error terms. It is expected that the elasticities associated with price (M2, M6, and M11 ) and






50


HFCS availability (M9 and M14) to be negative and the others to be positive. Price elasticities were expected to be inelastic. Serial correlation is anticipated and corrected by the Yule-Walker Method with appropriate lags assigned. U.S. Sweetener Supply Model

Regression analysis is used to estimate the U.S. sweetener supply utilizing timeseries data. Aggregate sugar supply is analyzed in three parts, i.e. total sugar supply, sugar supply from sugarcane and sugar supply from sugar beets. Sugarcane and sugar beet production are estimated separately as they have different production regions and a different refining process. Sugar production is used as the dependent variable. It is assumed that domestic sugar production is the primary source of sugar supply and carried-over stock from the previous periods is considered constant over the estimated time span. The model is specified as a double-log (natural) form and estimated coefficients can be read as elasticities associated with each variable.

In(QSTOTAL SUGAR, t) = UUJ + UU2*In(P TOTAL SUGAR, 1-1) + UU3*In(COSTt) + UU4*In(RCVTOTAL SUGAR, ) + UUs*In(Q TOTAL SUGAR, t-.) + UU6*TREND t + ee, (4.5)

In(QSCANE SUGAR, ) = UU7+ UUs*ln(P CANE SUGAR, t-1) + UU9*ln(COSTt)

+ UUIo*In(RCVCANE SUGAR, ) + UUi I*In(Q CANE SUGAR, t-1) + UU12*TREND t + uut (4.6)

II(Q SBEET SUGAR, ) = UU13 + UUi4*In(P dBEET SUGAR, t-1) + UUis*In(COSTt) + UU16 *In(RCVBEET SUGAR, ) + UU]7*In(Q SBEETSUGAR, t-1) + UU18*TREND, + vv, (4.7)

where Q s represents the total sugar quantity produced, sugar quantity produced from sugarcane, and sugar quantity produced from sugar beets [1000 short tons] in the three






51


equations, respectively; P d represents the real retail refined sugar price in the previous year (deflated by CPI), real raw sugar price at NY sugar exchange in the previous year (deflated by CPI), and real wholesale refined beet sugar price in the previous year (deflated by CPI [cents / pound]), respectively; COST is real total farm production expenses deflated by CPI (used as a proxy of sugar production cost [US$]); RCVs are sugar recovery rates during sugar refining process [%]; TREND is a trend variable that represents technology advancement; and ee, , uu, vv, are error terms. Prices in the previous year are used assuming that decision-making on sugar production relies on sugar crop production, recognizing growers decide their production plan with the price realized in the previous year. Retail refined sugar price is used for total sugar production estimation due to lack of wholesale price data. Total sugar recovery rate is expressed as average of recovery rate of cane sugar and beet sugar computed by weighing each production onto each recovery rate. Autoregressive term (production in the previous year) is added in the regression to calculate long-run elasticities. Elasticities associated with price production cost (UU3, UU9, and UU15) are expected to be negative and the others to be positive. Price elasticities are expected to be inelastic. Previously reported own-price elasticities for the U.S. sugar supply at industry level are 0.14 for cane sugar and 0.34 for beet sugar by Petrolia and Kennedy (2002), using the U.S. wholesale refined beet sugar price reported at midwest markets. Price elasticities for land allowed to sugar production estimated by Lopez (1990) were 0.103 for cane and 0.246 for beets. Both authors reported higher elactisities for beet sugar.






52


Mexican Sweetener Supply Model

Analysis of the aggregate supply of sugar in Mexico is conducted in a similar manner as the U.S. supply analysis except that Mexico produces sugar only from sugarcane.

In(Qs ,) = MMI + MM2 *IZ(P D ,) + MM3*In(COSTt) + MM4*In(DTt)

+ MM5*In(SUGLOSS,) + MM6*In(DURTN,) +MM7*TREND I + ww,

(4.8)

where Q ss are total sugar quantity produced [MT], P d is real wholesale standard sugar price in the previous year (deflated by CPI [pesos / Kg]), COST is real average production cost per ton of sugar realized at sugar mill deflated by CPI [pesos], DT is average mill operation downtime ratio observed at sugar mill [%], SUGLOSS is average sugar loss ratio observed during sugar production process at mill [%], DURTN is average duration of the harvest in each season [days], TREND is a trend variable that represents technology advancement; and ww, is error term. Price in the previous year is used assuming that decision-making on sugar production relies on sugar crop production, recognizing growers decide their production plan with the price realized in the previous year. An autoregressive term (production in the previous year) is not added because of limited length of data available for regression. It is expected that the elasticities associated with price production cost (MM), downtime (MM4), and sugar loss (MMs) to be negative and the others to be positive. Price elasticities are expected to be inelastic. Previously reported own- price elasticities for Mexican sugar supply at industry level is

0.67 by Petrolia and Kennedy (2002). Borrel (1991) estimated two sugarcane price elasticities by regressing sugarcane yield on sugar cane price in an Almon polynomial distributed lag model and by regressing sugarcane acreage on sugarcane price. From the






53


former regression, statistically significant estimates were 0.0286 (lag=3) and 0.0294 (lag=4) and from the latter, the estimate was not statistically significant. U.S.-Mexico Bilateral Sugar Trade Model

Based on the theoretical framework presented in chapter 3, a bilateral trade model of U.S.-Mexico sugar trade is developed and prepared for the empirical study. First, linear inverse demand and supply equations (price endogenous) are formulated using estimated elasticities from the previous analysis following the procedures by Spreen et al. (2000) in order to formulate the objective function in the trade model. Detailed derivation is noted in Appendix C. Although the double-log form is used in the demand and supply analysis because of its statistical advantages in estimation, the linear form of demand and supply equations are used in the trade analysis model for the following reasons: with the linear form, welfare can be measured, whereas welfare cannot be measured using log form equations due to the nature of the mathematical attributes; using the linear form makes it possible for the demand or supply curves to shift with their slopes held constant; and interpretation of Kuhn-Tucker conditions presented in chapter 3 is facilitated.

Sugar trade models have been developed by Koo and Taylor (2000) and Petrolia and Kennedy (2000). The former model incorporates sugar production, consumption and stock changes in seventeen sugar producing and consuming countries and the latter encompasses the United States, Mexico, and Cuba. In these models, market equilibrium is solved based on the market clearing condition, i.e. the sum of all countries' excess demand for sugar becomes zero as sugar price adjusts. These models are relatively simple and may be suitable for macroscopic analysis since they are able to include many economies without making the models large and complicated; however, the solutions derived from these models pay little attention to country's welfare or transfer costs of






54


goods. In addition, details of trade agreements are difficult to incorporate in these models. The fact that there exist numerous bilateral and regional trade agreements, a spatial equilibrium model has the advantage of enabling the various trade agreements to be tailored into the model. Lastly but most importantly, it is critical to consider HFCS when the sweetener markets in both the United States and Mexico are under scrutiny. Since this study is focused on the bilateral relation between the United States and Mexico, a spatial equilibrium is chosen to conduct a more microscopic analysis, incorporating detailed trade agreements and the impacts from HFCS consumption. Simplifying assumptions

Several simplifying assumptions are made to run the model. First, the quantities in the models are raw sugar equivalent. In doing so, the price difference along the vertical market channel is ignored and derived demand and supply curves are assumed to posses the same slopes as demand and supply of refined sugar.

Second, changes in sugar stocks in both counties are also ignored and hence the excess sugar supply from Mexico and the excess sugar demand from the United States are captured as the difference between domestic sugar demanded and supplied in each county, illustrated in the following. Formally, the U.S. sweetener market balance is expressed with sugar production, consumption, import, export, and change in stock as well as with those of HFCS:

Q SUGAR, t + SE,+ ST,+1+ HFCSD t QSSUGAR, t+ SI 1+ STg-j + HFCSs, (4.9) where QDSUGAR is the quantity of sugar demanded, QsSUGAR is the quantity of sugar supplied from domestic sugar production, SE is sugar export, SI is sugar import, STt+ is sugar stock carried over to the next year, ST,.; is sugar stock carried over from the previous year, and HFCSD and HFCSs are HFCS demand and supply related transactions,









respectively. QDSUGAR, t Is expressed in annual base converted from quarterly base. Since neither supply nor demand of HFCS is estimated, HFCSD and HFCSS are treated constant. This is justified by the fact that HFCS use in the United States has remained stable in recent years, accounting for roughly half of total U.S. caloric sweetener use. In addition, changes in stock have also remained stable given the nature of commodity demand and production practices, ST,-1 and ST,1 are set zero. SE is also set zero as sugar export from the United States is negligible given the production capacity. Equation [4.9] is then simplified and the quantity of forecasted sugar import in the United States (Si, ) is expressed with forecasted quantities of sugar demanded and supplied (6DSUGAR, and 6 SSUGAR, ) as:

SI, =Q SUGAR,t - Q SUGARR. (4.10)

Mexico's sweetener market balance is expressed in a similar fashion. The

differences are sugar demand in Mexico is estimated by total, direct and indirect sugar consumption and Mexico imports HFCS from the United States to meet its domestic demand for sweeteners. By ignoring stock changes and sugar import, the sweetener market balance is expressed as:

QDCSUGAR, t Q Q'CSUGAR, I + HFCSD ,+ SE, = QSSUGAR, t + HFCSs (4.11) where QDCSUGAR is the quantity of demanded by households (direct consumption), QICSUGAR is the quantity of sugar demanded by bulk users (indirect consumption), HFCSD is the quantity of HFCS demanded by bulk users, SE is sugar export, QSSUGAR is the quantity of sugar supplied from domestic sugar production, and HFCSs is the quantity of HFCS supplied from domestic production and import from the United States. The study is interested in forecasting sugar surplus available for exporting to the United States






56


considering Mexico's increasing demand for HFCS. Since HFCSD and HFCSs are not estimated directly, HFCS demand (HFCSD ) is forecasted based on the associated scenarios with details provided in the following section.' Once HFCS demand is forecasted, the quantity of excess sugar destined to export is calculated by considering the substitution between HFCS and indirect sugar consumption, which occurs at industry (bulk users) level. Let QICSUGAR, tbe HFCS forecast-adjusted indirect sugar consumption, then the sugar export forecast (SE,) is expressed with forecasted quantities of sugar demanded and supplied ( SUGAR, t, DCSUGAR, t ,and Q /CSUGAR, t) as:


SZ, = SUGAR, t DCSUGAR, t + QCSUGAR, t) (4.12)

In equation [4.34], it is also assumed that HFCS is consumed only to supplement sugar consumption and its demand is met by readily available HFCS from domestic production and import from the United States. Accordingly, NAFTA provisions are defined, depending on the Mexico's domestic sweetener balance; Mexico receives a larger quota for the following year if is attained net surplus producer status for two successive years:

$ DCSUGAR, t + ICSUGAR, t + HFCSD t< SUGAR, t, (4.13)

otherwise, Mexico receives a smaller quota until 2008 when access to the U.S. market is unlimited.

Model calibration

The model is calibrated by positioning the intercepts of inverse linear demand and supply equations and by adjusting average transportation cost from the rest of the world


Alternatively HFCS demand can be forecasted balancing the total consumption of sweetener equation with direct and indirect consumption sugar equations, yet it resulted in a poor forecast.






57


to the United States in order to iterate for solution from 2002 through 2015 using the mathematical programming software package GAMS.

Intercepts of inverse linear demand and supply equations in quantity-price space

(1U1, IUU1, IM1, and IMM in equations [4.14] through [4.17], respectively) are calibrated with the actual values realized in 2001 (base year) from equations [C.7b], [C8b], [C9b] and [C.10b] presented in Appendix C:

IU1 = PD US, 2001 - (IU2*QDus, 2001 + Shiftero US, 2001) (4.14)

IUU1 = PUs, 2001 - (IUU2 *Qsus, 2001 + Shifter s US, 2001). (4.15)

IMI = PDMX, 2001 - (lM2*QTCMX, 2001 + ShifterD mx, 2001) (4.16)

IMMI = PsMX, 2001 - (IMM2*QsMX, 2001 + ShiftersMX, 2001). (4.17)

By calibrating intercepts, both excess supply in Mexico (quantity presented as "A" on ESMx curve in Figure 4-1) and excess demand in the United States (quantity presented as "B" on EDus curve in Figure 4-1) are set to correspond to the actual volume realized in 2001. Slopes for both inverse demand and inverse supply curves in both countries are held constant, yet these curves shift over the forecast horizon according to the scenarios proposed in the following section.

In order to represent Mexican sugar supply and export capacity adequately, the model is further calibrated by adjusting the average transportation cost from the rest of the world to the United States, based on the relationship expressed in equation [3.50] in chapter 3. To do so, the transportation cost is calibrated so that Mexico exports sugar over-quota at the minimum amount. This calibration procedure resulted in a rather high transportation cost from the rest of the world to the United States; however, it insinuates the irrational behavior of the Mexican sugar industry which has been suffering from






58


financial stress, vividly illustrated by the mill expropriation by the government in 2001, and producing and exporting surplus sugar to the rest of the world at the same time. Simulated Scenarios

Scenarios are formulated by considering alternative assumptions related to

Mexico's sweetener market situations affected by continued gains in the productivity of the sugar industry, HFCS consumption, as well as a policy lever and U.S. sugar policy levers. Each scenario carries a combination of Mexico's sweetener market situation and U.S. sugar policy. To compare the impacts of scenarios, a baseline scenario is defined where status quo is maintained (Table 4-1). In the baseline scenario, it is assumed that shifts in sugar demand and supply in both countries continues at the average rates observed in recent years and that the U.S. government maintains price support and allocates quota among exporters in a flexible manner, abiding by the WTO minimum import requirement.2

Assumptions related to Mexico's sweetener market situations are summarized in Table 4-2. Four situations are proposed: higher sugar production, higher HFCS adoption, a combination of both and introduction of tax on HFCS as a Mexico's alternative policy lever. The rates of increase in production and HFCS adoption are defined relative to the baseline. Impacts of Mexico's tax on HFCS is based on the forecast by Haley and Suarez (2003) where Mexican HFCS consumption drops significantly in 2002 and 2003 due to a tax imposed on beverages that contain HFCS. The impact of this tax is assumed significant considering soft-drink manufacturers currently account for about one-third of domestic sugar consumption in Mexico (Buzzanell, 2002). In all situations, U.S. demand


2 Current allocation system is based on historical trade shares (Skully, 1998).






59


and supply are held at "Baseline" level. The duration of harvest is held constant because it is partly affected by the weather. Also, the assumption where Mexico's HFCS adoption and tax on HFCS occur at the same time is not considered since the scenario becomes incompatible to simulate.

The scenarios associated with "HFCS adoption" are consistent with an increasing trend since 1994 when NAFTA went into effect and speculation of expansion of HFCS production in Mexico. If Mexico continues to increase HFCS consumption, it may follow a similar path as was seen in the United States in early 1980s when soft-drink manufacturers decided to switch to HFCS from sugar. HFCS manufacturers, who are mostly the U.S. corporations, see this phenomenon as a business opportunity in Mexico. With HFCS capital-intensive facilities, existing HFCS plants in Mexico are operated by the firms based in the United States. The "HFCS Adoption" case assumes that HFCS will be adopted in a linear fashion until HFCS consumption is 50 percent of total indirect sweetener consumption in 2008 and its share remains constant for the rest of the forecast horizon. The 50 percent share of indirect consumption of sweetener is equivalent to about 27 percent of total consumption of sweeteners. In 2001 (base year), share of HFCS in indirect sweetener consumption and total sweetener consumption were 25.3 percent and 11.6 percent, respectively. Comparisons of HFCS and indirect sugar consumption forecast for "Baseline" scenario, "HFCS adoption", and "Tax on HFCS" situations are illustrated in Figure 4.2.

Assumptions related to U.S. sugar policy levers are summarized in Table 4-3. Two kinds of policy levers are considered: one is to stabilize the demand price and the other is to allocate quotas to exporters. For the former, two possible sugar policies are used as






60


alternatives to the current price support. In one policy, instead of directly supporting sugar price, the government indirectly supports a sugar price through buying up excess sugar in the market to assure that sugar price will not fall due to excess supply from overseas. In this assumption, the government buys up as much sugar as the net sugar import, i.e. total sugar import less minimum import requirement. In the bilateral trade model, the cost of buying up excess sugar is not included in the objective function, assuming that the government does not spend ex ante cost. Also for simplicity, sugar storage costs incurred by the government are ignored. The second policy option is to introduce sugar production controls in both the United States and Mexico. This option requires cooperation from the Mexican sugar industry: defection by either party will result in an unsuccessful outcome. In this way, both the United States and Mexico are assumed to control their sugar production according to forecasted demand, avoiding excess supply of sugar. In other words, the idea implies that the United States is willing to import sugar from Mexico as long as Mexico cooperates to reduce its sugar production to meet the sugar demand in both countries and also that Mexico can avoid excess surplus sugar which cannot be sold anywhere except in the world market. This sugar policy does not involve financial support from the government: the cost of the program is zero.

U.S. policy levers related to quota allocations to exporters are treated with two different approaches: the U.S. government allocates import quotas in a flexible manner between Mexico and the rest of the world (status quo); and minimum quotas are maintained (the remainder of the minimum import requirement less that allocated to Mexico) to the rest of the world no matter how much Mexico exports to the United






61


States. Both policies abide by the minimum import requirement under WTO; however, political feasibility is assumed to be quite different.

Scenarios to simulate are formulated by combining specific Mexico's sweetener market situation and the U.S. sugar policy. A list of 16 scenarios is shown in Table 4-4.

Finally, special simulations are prepared in order to further examine the effects of Mexico's production improvement and HFCS adoption on the Mexican sugar industry and welfare. In addition to the assumption regarding Mexico's production improvement (an additional 1 percent to the baseline) and HFCS adoption (share of indirect sweetener consumption by HFCS at 50 percent) summarized in Table 4-2, simulations are conducted by changing production improvement rate at additional 0.5 and 1.5 percent as well as HFCS adoption to achieve a market share of 30, 40, and 45 percent. Game Theory Analysis

In order to assess policy recommendations using aggregated results from the

various simulations, an analysis based upon game theory is introduced. The basis of the game used in the study is a non-zero-sum game with mixed strategies. Non-zero-sum means the sum of the pay-offs in each pair of strategies is not zero; in other words, one player's winning does not necessarily cause the other to lose. Mixed strategies means a player chooses a strategy to play with probability (Morris, 1994; Mas-Colell et al., 1995). Since the game is non-zero-sum, both cooperative and non-cooperative games are considered. While a cooperative game allows players to make binding agreements about how they will play or about sharing pay-offs, a non-cooperative game does not. In the latter case, the game is played by two parties: the U.S. and the Mexican governments who hold the strategies and make decisions on behalf of the economy as a whole. In the former case, the game goes through the process of considering the possible pay-offs to






62


two coalitions formed among five parties (payees), i.e. three industries (the U.S. HFCS industry, the U.S. sugar industry and the Mexican sugar industry) and two countries; however, the final decisions are made by the governments who hold the strategies.

Each country plays with multiple strategies that correspond to Mexico's market

situations and the U.S. policy levers presented in the previous section (Table 4-5). In this game setting, a combination of strategies formulates a scenario. In the game, it is assumed that if the U.S. government introduces production control, Mexico always cooperates: a defection by either party is not considered. Also, the U.S. government chooses only flexible quota allocations to the rest of the world for simplicity. In the case of Mexico's strategies, HFCS adoption is treated as a strategy although it is neither a positive strategy nor controllable by the Mexican government.

Pay-offs from each scenario are calculated for each payee. Pay-offs to the

industries are expressed as present values of accumulated revenue between 2002 and 2015, assuming a three percent discount rate each year. The HFCS price is held constant at the average U.S. export price to Mexico realized between 1992 and 2001. Pay-offs to each country are expressed as present values of accumulated welfare, i.e. the sum of consumer and producer surplus. In doing so, changes in welfare are calculated only from the sugar market, assuming that sugar is a primal source of sweetener. Sugar cannot be substituted with HFCS for certain products due to the liquid form of HFCS. Also, sugar is preferred for certain products to HFCS due to flavor given to the final products. U.S. welfare is adjusted with tariff revenue from Mexico, the cost of the sugar program and the cost of buying up excess sugar. The cost of the sugar program is calculated by multiplying the price difference between the support price (loan rate for raw sugar, 18






63


cents per pound) and the U.S. equilibrium price with the production quantity guaranteed at the loan rate. This cost is captured by comparing between scenarios with and without the price support, ceteris paribus. The cost of buying up excess sugar is calculated by multiplying the U.S. net sugar import, i.e. total sugar import less 1,256,000 MT of minimum import requirement with the U.S. equilibrium price. In the case of the policy to buy excess sugar policy, sugar storage costs incurred by the government are ignored. Note that since the values were converted in terms of U.S. dollars prior to simulations, the exchange rate realized in the base year (2001) was implicitly used for calculating payoffs.

All the games proposed are solved through a two-players pay-off matrix (twodimensions) with the two governments are the decision-makers. When coalitions are formed, their individual pay-offs are pooled, assuming that the total pay-off is redistributed among them (Morris, 1994). In doing so, it is also assumed that industry revenue and each nation's welfare can be added together. After pay-offs are calculated for each party or coalition, these values are indexed as a relative gain or loss to the payoffs in baseline scenario to facilitate the decision process. Mixed-strategy games are first simplified by eliminating dominated strategies (strategies that played with zero probability) and then solved through maximizing the expected pay-off to each player from the game (Morris, 1994; Varian, 1992).

Sources of Data

Data for the Mexican sugar industry was obtained from the website of Comit6 de la Agroindustria Azucarera (COAAZUCAR, Sugar Agro-Industry Committee). The committee is in charge of monitoring sugar cane and sugar production at each mill as well as determining the cane price paid to farmers in the country. The latter task was






64


taken over from the former government body after the privatization of the milling sector. The committee carries an extensive data set regarding not only physical production and price but also detailed productivity and efficiency indicators such as sugar and fiber content in cane, mill downtime and sugar production loss during the process across 60 operating mills. Data for the U.S. sugar industry was obtained from the Sugar and Sweetener Situation and Outlook Yearbook and other publications by the Economic Research Service, USDA. Historic data for population were obtained form the website of the U.S. Bureau of the Census; and those for GDP, exchange rates and consumer price index were taken from OECD documents.

Data sources used in the study are summarized in Tables 4-6 through 4-10. For demand analysis, aggregate time-series data at the national level are used in the regressions for both the United States and Mexico. Missing data were found in U.S. sugar consumption and related prices in 1991 and filled with average values of two adjacent years in order to maintain the data continuity.

For the supply analysis, aggregate time-series data at industry level are used in the regression for both the United States and Mexico. No missing data were found in the data set. Nominal values of old Mexican pesos before devaluation found in the data are converted into current pesos. Local units are preserved during the estimation of elasticities in demand and supply analyses. In bilateral trade model, different local units are converted into common units such as metric tons and US dollars so that the model can achieve the equilibrium point in the system.






65


P


Dmx SMX


- - -.. . . P2001, MX
(calibration point)



Q
Mexican Market

Fi


P


ESmx
B
A >

Q


Mexico-U.S. Market

gure 4-1. Image of Model Calibration


P


(calibration point)
P2001, US D.us Sus


Q


U.S. Market






66


Table 4-1. Assumptions for Baseline (Status quo) Scenario

Category Shifters Country Assumptions


United States and Mexico


Will increase at the average real GDP annual growth rate realized between 1997 and 2001 (2.23 percent for the United States and 2.91 percent for Mexico).
Will increase similarly as the forecast by the U.S. Bureau of the Census. (Annual growth rate: 0.88 percent for the United states and 1.15 percent for Mexico).
Will decrease at the average annual reduction rate realized between 1997 and 2001 (1.10 percent for the United States and 1.80 percent for Mexico).


Recovery rate United Will increase at the average annual
States improvement rate realized between 1997 and 2001 (0.48 percent).


Sugar loss


Duration of harvest
HFCS
consumption


Mexico


Will decrease at the average annual reduction rate realized between 1997 and 2001 (1.00 percent).
Will decrease at the average annual reduction rate realized between n1997 and 2001 (2.00 percent).
Held constant

Will be consumed maintaining the same share in indirect consumption of sweetener realized in 2001 (25.3 percent)


Price support The U.S. government maintains the price
U.S. support (current policy) at 18 cents per pound
sugar (loan rate).
policy Flexible allocation The U.S. government allocates quotas in a flexible manner between Mexico and the rest of the world.


Market situations


GDP


Population



Production cost


Downtime






67


Table 4-2. Assumptions for Mexican Sweetener Market Situations

Situations Shifters Assumptions
(code)
High Production cost, Will improve an additional 1 percent to the
Mexican Downtime, "Baseline".
Production Sugar loss
(P)
High HFCS HICS It is assumed that HFCS consumption will increase in
Adoption consumption a linear fashion until it replaces 50 percentage of
(A) indirect consumption of sugar in 2008. After 2008, its
share remains at 50 percent.
High Production cost, The same as "High Mexican Production"
Mexican Downtime,
Poduction - Sugar loss High HFCS HFCS The same as "HFCS Adoption"
adoption consumption
(PA)
Tax on It is assumed that HFCS consumption for 2002
HFCS HFCS through 2004 will drop due to tax (Haley and Suarez,
(T) consumption 2003) and remain at reduced consumption level (7
percent to the indirect consumption of sugar) for the rest of the forecast horizon.3
Other shifters are the same as the Baseline.


3 A 20-percent tax on beverages that contain HFCS was introduced on January 1, 2002; suspended on March 5 by the president's decision; and then reimposed on July 16, 2002 with the decision by Mexico's Supreme Court of Justice (USDA, 2002a).







68


6,000,000 -T


5,000,000 4,000,000 3,000,000 2,000,000 1,000,000


V--+-0 HFC 0s p0 (a l o 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0) 0 0) 80 0; 0 S
Cl Cl rq rq Cl C l Cl C l Cl C l Cl4
Year

-4-Indirect consumption of sugar (Baseline)
-It Indirect consumption of sugar (Tax on HFCS)
-a-Indirect consumption of sugar (High HFCS adoption)
-0- HFCS consumption (Baseline)
-E- HFCS consumption (Tax on HFCS)
-A- HFCS consumption (High HFCS adoption)


Figure 4-2. Forecasted Indirect Sugar and HFCS Consumption in Mexico under Alternative Scenarios






69


Table 4-3. Assumptions for U.S. Sugar Policies

Categories Policies Assumptions
(code)
Stabilization Price support The U.S. government maintains the price support
of the (status quo) at 18 cents per pound (loan rate, 396.48 US$ per
demand price (S) MT).
Buying up excess The U.S. government abandons the price support
sugar in the market and buys up excess sugar in the market instead.
(B)
Production control The U.S. government abandons the price. Instead,
(C) the U.S. and Mexican governments collectively
control the sugar production in such a way that the sum of quantities demanded in two countries is primarily met by the sum of the quantities supplied from two countries.
Quota Flexible allocation The U.S. government allocates quotas in a
allocations (status quo) flexible manner between Mexico and the rest of
(F) the world.
Minimum quota The U.S. government reserves the minimum
allocations for the quotas (the remaining minimum import
rest of the world requirement less allocated to Mexico) for the rest
(M) of the world.






70


Table 4-4. Listing of Examined Scenarios

U.S. sugar policies Scenario Mexican market
number situations
(code) Stabilization of the Quota allocations
demand price

I Baseline
(baseline) (B) Price support Flexible
2 High production (S) allocations
(P-S-F) (P) (F)
3 HFCS adoption
(A-S-F) (A)
4 High prod.-HFCS adop.
(HA-S-F) (PA)
5 Baseline
(B-B-F) (B) Buying up excess
6 High production sugar in the market
(P-B-F) (P) (B)
7 HFCS adoption
(A-B-F) (A)
8 High prod.-HFCS adop.
(PA-B-F) (PA)
9 Baseline
(B-C-F) (B) Production control
10 High production (C)
(P-C-F) (P)
11 HFCS adoption
(A-C-F) (A)
12 High prod.-HFCS adop.
(PA-C-F) (PA)
13 Price support Minimum quotas
(PA-S-M) High prod.-HFCS adop. (S) allocation to the
14 (PA) Buying up excess rest of the world
(PA-B-M) sugar in the market (M)
(B)
15 Production control
(PA-C-M) (C)
16 Tax on HFCS Price support Flexible
(T-S-F) (T) (S) allocations (F)






71


Table 4-5. Strategies for the Sugar Trading Game


Country


Strategies


United States Strategy 1 Maintains price support (status Flexible quota quo) allocations among
Mexico and the
Strategy 2 Abandons price support and buys rest of the world up excess sugar in the market Strategy 3 Introduces production control with Mexico

Mexico Strategy 1 Maintains the current policy (status quo)


Strategy 2 Higher sugar production


Strategy 3 Higher HFCS adoption


Strategy 4 Higher sugar production and higher HFCS adoption


Strategy 5 Introduces tax on HFCS






72


Table 4-6. Data Sources for U.S. Demand

Data Unit Source

Consumption of sugar 1000 Sugar Statistical Compendium by (Stock
(dependent variable) short #91006, 1970-1990) and Sugar and
tons Sweetener Situation and Outlook Retail price of refined sugar cents / Yearbook (SSS-2002, 1992-2002) by pound Economic Research Service, U.S.
Department of Agriculture
Gross Domestic Product US$ Statistic database by Organization for
(GDP) Economic Co-operation and Development
(OECD)
Population persons International Database by U.S. Bureau of
the Census, U.S. Department of Commerce
Data length: 1970-2002, quarterly


Table 4-7. Data Sources for Mexican Demand

Data Unit Source

Direct consumption of sugar metric (dependent variable) tons
[MT] Azucar S.A. de C.V. Estadistica Indirect consumption of sugar Azucareras (1970-1989), database by
(dependent variable) Financiera Nacional Azucarera, S.N.C de
Total consumption of sugar C.V. and Comite de la Agroindustria
(dependent variable) Azucarera (COAAZUCAR) (1990-1999)
Retail price of standard sugar pesos / kg
Gross Domestic Product pesos Statistic database by Organization for
(GDP) Economic Co-operation and Development
(OECD)
Population persons International Database by U.S. Bureau of
the Census, U.S. Department of Commerce
Data length: 1970-1999






73)


Table 4-8. Data Sources for U.S. Supply

Data Unit Source

Production of sugar, total 1000 (dependent variable) short
Production of cane sugar tons
(dependent variable) Sugar Statistical Compendium by (Stock
Production of beet sugar #91006, 1970-1990) and Sugar and
(dependent variable) Sweetener Situation and Outlook
Retail price of refined sugar cents / Yearbook (SSS-2002, 1980-2002) by Wholesale price of refined pound Economic Research Service, U.S.
Department of Agriculture Sugar recovery rate, total percent Beet sugar recovery rate Cane sugar recovery rate
Total farm production million Database by Economic Research Service,
expenses US$ U.S. Department of Agriculture
Data length: 1960-2002


Table 4-9. Data Sources for Mexico Supply

Data Unit Source

Production of sugar MT
(dependent variable)
Wholesale price of standard pesos / sugar kg
Cost of sugar production per pesos / Database by Comite de la Agroindustria ton of sugar produced ton of Azucarera (COAAZUCAR)
sugar
Downtime observed at mills percent

Sugar loss during the process percent Duration of the harvest days
Data length: 1988-2000






74


Table 4-10. Data Sources for Miscellaneous

Data Source

U.S. Consumer Price Index (CPI) (1982- Database by the Bureau of Labor 1984=100) Statistics, U.S. Department of Labor
Mexico Consumer Price Index (CPI) (1994= 100) Database by the Banco de Mexico
Exchange rate (US$- Mexican pesos)
Quota-tariff rates Agricultural Outlook (1999) by Economic
Research Service, U.S. Department of Agriculture
Sugar per unit transportation cost (bagged, Personal communication with an exporter seaborne freight rate) of sugar
Loan rate for raw sugar in the U.S. Haley and Suarez (2002)
Guarantee price for raw sugar in Mexico Database by Comite de la Agroindustria Azucarera (COAAZUCAR)














CHAPTER 5
EMPIRICAL RESULTS AND INTERPRETATION

This chapter presents and discusses the empirical results for each of three analyses:

(1) demand and supply analysis for both the U.S. and Mexican sugar markets, (2) bilateral sugar trade analysis, and (3) game theory analysis.

Demand and Supply Analyses

Results for the U.S. demand and supply analysis are summarized in Table 5-1. In the demand equation, signs of estimates associated with each significant variable were as expected. Significant estimates at the 95 percent confidence level were associated with price, the dummy variables for quarter 1 and 3, and the dummy variable for HFCS availability. The estimated price elasticity of demand was inelastic. The significant estimate associated with the HFCS dummy variable implies that HFCS replaces sugar as a substitute in the market to some degree.

In the supply equation for the United States, estimates associated with trend and production in the previous year (autoregressive term) were significant at the 95 percent confidence level for all three models, i.e. total, beet and cane sugar supply regression models. Estimates associated with sugar recovery rate were insignificant in all models. Estimates associated with price and cost were significant at the 95 confidence level for total and beet sugar supply regressions, but not for the cane sugar supply equation. Two possible reasons why cane sugar production does not respond to the refined sugar price but beet sugar production does are: (1) cane sugar has two steps in the refinery process while beet sugar has one and (2) sugarcane is a perennial crop while sugar beets is an


75






76


annual crop. These results reflect these differences and imply that sugar beet production is more sensitive to price changes. Although cane sugar production is assumed to respond to raw sugar price, the coefficient of the price variable is not significant. Estimated price elasticities were both inelastic for total sugar supply and beet sugar supply as anticipated. The long-run supply price elasticities were calculated dividing the estimated price elasticities (short-run) by (1- X) where X represents the estimates for production in the previous year (autoregressive term). Computation yielded all inelastic long-run price elasticities: 0.3875 and 0.6764 for total sugar and beet sugar, respectively.

Results for the demand and supply analysis for Mexico are summarized in Table 52. In the demand equation, signs of significant estimates associated with each variable were consistent with a priori expectations. The only statistically significant estimate among the three price elasticities was direct consumption, and it was inelastic. The population variable accounted for most of the explanatory power of consumption in all models. Significant estimates associated with GDP in indirect sugar and total sweetener consumption indicate that consumers tend to consume more sugar through sugarcontaining products as their income increases.

In the supply equation for Mexico, the signs of estimates associated with each

variable corresponded with a priori expectations. The estimate associated with price was inelastic. While reduction in production cost and factory downtime indicated an increase in production, the length of sugarcane harvest duration was almost perfectly correlated to sugar output from the mills. The coefficient for the variable representing sugar loss during the process was not significantly related to sugar output implying that the degree of sugar loss was not as critical as other factors such as production cost and factory






77


downtime for mills trying to improve their production efficiency. A positive and significant estimate associated with the trend variable indicates technology related to sugar production at sugar mills had been advancing during the time period covered in this study (1988-2000).

Bilateral Sugar Trade Analysis

The results from the bilateral sugar trade analysis are divided into two parts: U.S.

sugar import forecasts and the comparison of pay-offs among industries and countries. Results and discussion are guided by comparing the impacts from changes in Mexican sweetener market situations and from changes in U.S. sugar policy.

U.S. sugar import forecasts for the selected eight scenarios (scenario 1, 2, 3, 4, 8, 12, 13, and 16 in Table 4-4) are shown in Figures 5-1 through 5-8. Comparisons among scenarios 1, 2, 3, 4, and 16 illustrate the impacts from changes in the Mexican sweetener market situations; comparisons among scenarios 4, 8, and 12 illustrate the impacts from changes in U.S. sugar price stabilization policy; and comparison between scenarios 4 and 13 illustrates the impacts from changes in U.S. quota allocation policy.

The results from the baseline scenario (Scenario 1) shows that if status quo

surrounding the sugar industries in both countries is maintained over the forecast horizon, Mexico will not likely attain a net surplus sweetener producer status and hence will miss the opportunity to benefit from exporting sugar under a larger quota allocation (250,000 MT) under NAFTA (Figure 5-1). This is due to growing domestic sugar demand relative to domestic sugar production. For comparison, forecasts for Mexican sugar consumption and production in 2008 by this study are 5.4 million and 5.8 million MT, respectively and those by Koo and Taylor (2000) are 5.3 million and six million MT, respectively.






78


When Mexico expands sugar production (Scenario 2), net surplus sweetener

producer status will still not be attained (Figure 5-2). Yet, Mexico will generate enough surplus sugar to export over-quota (before 2008) and quota-free (after 2008), resulting in significant impacts on the U.S. market. In total, Mexican sugar will take up about onethird (Scenario 1) or more than half (Scenario 2) of the U.S. minimum import requirement at peak in 2008. The amount of export will decline in later years due to expanding domestic sugar consumption in Mexico.

In Scenario 3, when Mexico adopts HFCS at a higher rate, Mexico's sugar export swells as a result of substitution between sugar and HFCS in the domestic market. This result includes direct impacts on the Mexican sweetener market as well as extended impacts on the U.S. sweetener market (Figure 5-3). Although Mexico will not attain net surplus sweetener producer status, over-quota export will reach over 1.2 million MT by 2007 and will remain over one million MT until 2014. This export quantity will take up almost the entire U.S. minimum import requirement and as a result, sugar export from the rest of the world will be marginalized. Similar results are drawn when increases in Mexico's production and HFCS adoption are combined (Scenario 4). A slightly larger scale of Mexico's exports than Scenario 3 is shown in Figure 5-4.

The policy followed by the U.S. government in its allocation of its import quota has a large impact on sugar exports from both Mexico and the rest of the world. The aforementioned large-scale export of Mexican sugar is possible only if the U.S. imports the minimum amount of sugar and allocates sugar quotas in a flexible manner among exporters. This allocation method may cause friction with the other countries that export sugar to the United States since Mexican sugar has potential to take up a large portion of






79


the U.S. quota. Thus, allocating such a large portion of the quota to one country may not be a feasible policy option in the United States. In Scenario 13, it is conjectured that the U.S. government maintains minimum quotas (the remainder of the minimum import requirement less allocated to Mexico) for the rest of the world no matter how much Mexico exports. As shown in Figure 5-5, Mexico's over-quota and quota-free export will be dampened because of Mexico's comparative disadvantage to the rest of the world, while the export from the rest of the world remains over 1.2 million MT over the entire forecast horizon.

By contrast, a change in the U.S. government's policy on stabilization of the

domestic price does not pose much impact on sugar exports to the U.S. market. With the same Mexican sweetener market situations (high production-high HFCS adoption), the U.S. policy options of price support, buying up excess sugar in the market, and production controls are compared (Figures 5-4, 5-6 and 5-7). The results indicate that all three scenarios bring about the similar trends in Mexican sugar export. Yet, U.S. production control with Mexico causes an overall increase in quota-free export after 2008. This is because the Mexican sugar price is maintained lower relative to the price support or buying up excess sugar scenarios (Figures 5-9, 5-10, and 5-11). The prices of sugar in both countries will converge in the integrated market without a U.S. price support (Figure 5-10) while the difference in prices are kept after 2008 with production control (Figure 5-11): the price difference corresponds to the transportation cost from Mexico to the United States.

Mexico's tax on HFCS brings about very different results. In this scenario,

Mexico is unable not only to export either under-quota, over-quota or quota-free but also






80


to generate a sugar surplus after 2006 due to expanding domestic sugar consumption accelerated by higher sugar consumption by bulk users who chose sugar over HFCS. Mexico cannot produce enough sugar to meet domestic demand after 2008 and will import sugar from the United States, resulting in a higher domestic sugar price than that of the U.S. (Figure 5-12). This extreme case would occur only if the impact of tax on HFCS lingers over the forecast horizon as assumed in this study; however, the chance for Mexico to enjoy exporting sugar would be slim.

In spite of fluctuating imports from Mexico and the rest of the world, U.S. domestic sugar consumption and production will remain relatively unchanged. Sugar demand and supply forecasts for both the United States and Mexico for baseline and high productionhigh HFCS adoption scenarios are shown in Figures 5-13 and 5-14, respectively. In either scenario, U.S. demand and supply are forecasted to remain approximately at nine million and 8.8 million MT over the forecast horizon, respectively.

The results from the various simulations showed that the U.S. sugar price will gradually decline but will not dip below the support price level (396.48 US$ per MT) before 2008 if the United States accepts most of the imported sugar from Mexico rather than from the rest of the world by allocating quotas in a flexible manner to the exporters, no matter which U.S. demand price stabilization policy is in place (Figures 5-9, 5-10, and 5-11). This implies that the Mexican sugar price contributes to maintain a high sugar price in the integrated U.S.-Mexico sugar market; in other words, accommodating Mexican sugar can act as an alternative form of price support in the United States. On the other hand if the U.S. government maintains minimum quotas for the rest of the world no matter how much Mexico exports and abandons the price support, U.S. equilibrium price






81


will fall below the support price (Figures 5-15 and 5-16). This is a result of cheaper sugar from the rest of the world flowing into the U.S. market. In reality, the U.S. sugar price will face downward pressure from importing world sugar as well as political pressure from the rest of the world, considering the likelihood that many sugar exporting countries will not easily give up their existing shares of the U.S. import quotas.

Pay-offs to the industries and countries also portray interesting contrasts among

scenarios. The results from three sets of selected scenarios are summarized in Tables 5-3, 5-4, and 5-5. In these tables, the present values of accumulated pay-offs are expressed in billion of dollars and those values are indexed relative to the baseline scenario inside the brackets.

The impact of changes in Mexican sweetener market on pay-offs to the industries and the two nations' welfare is illustrated in Table 5-3. The listed five scenarios are based on the assumptions that the U.S. government maintains price support and allocates quotas between Mexico and the rest of the world in a flexible manner. It is clear that the U.S. HFCS industry will become better off if Mexico adopts HFCS: revenue for the U.S. HFCS industry increases by 78 percent; and that the industry will become worse off if Mexico introduces taxes on HFCS: revenue for the industry decreases by 72 percent. The U.S. sugar industry does not gain from either Mexico's increase in sugar production or HFCS adoption since either change in Mexican sweetener market generates a larger Mexican sugar surplus that is destined for the U.S. market. Tariff revenue to the U.S. government from Mexican sugar increases as Mexico increases sugar production or HFCS adoption. Large tariff revenue from high HFCS adoption scenario, which is unexpectedly larger than the high production-high HFCS adoption scenario, is due to






82


forecasted larger exports in 2008 and 2009, which consequently brings a lower U.S. sugar price accompanied by a higher cost of the price support. Overall, U.S. welfare becomes worse off with changes in Mexican sweetener market for two reasons: (1) reduced producer surplus caused by a lower producer price as a result of increased Mexican export and (2) increased net costs even though increased tariff revenue is expected as mentioned above. The Mexican sugar industry gains from expanding sugar production, but the gains are dissipated when HFCS is adopted at higher rates, given the assumption that HFCS price held constant. In Mexico's HFCS tax scenario, the Mexican sugar industry gains not from exporting to the U.S. market but from domestic sales at higher prices. All the entities except for U.S. HFCS industry benefit from this policy; however, the policy lever may not be acceptable in the international trade environment. In fact, the Mexican government swung its decisions in the past: a 20-percent tax on beverages that contain HFCS was introduced on January 1, 2002; suspended on March 5 by the president's decision; and then reimposed on July 16, 2002 with the decision by Mexico's Supreme Court of Justice (USDA, 2002a).

The impact of changes in the U.S. price stabilization policy is illustrated in Table 54. The compared three scenarios (Scenario 4, 8, and 12) are based on the assumption that Mexico increases sugar production as well as HFCS adoption and that the U.S. government allocates quotas between Mexico and the rest of the world in a flexible manner. Alternative sugar policies to the price support only bring about improvement in U.S. welfare; sugar industries in both countries and Mexico's welfare become worse off, posing a larger negative impact on the Mexican sugar industry and welfare than the sugar industry in its own country. When the U.S. government switches sugar policy from






83


supporting price directly to buying up excess sugar in the market, the cost of the program is expected to diminish, resulting in welfare improvement. Yet, if the U.S. government were to switch policies from the price support to buying up excess sugar, the timing to do so will be important so as to minimize the cost incurred by the government; the cost of buying up excess sugar will rise immediately after policies are switched while the cost of the price support will not because the U.S. sugar price will be maintained relatively high in the early stage of the forecast horizon. In practice, storage costs need to be considered. When the United States controls sugar production with Mexico, welfare improves by increased consumer surplus and tariff revenue combined with zero program cost incurred by the government. This scenario also demonstrates that U.S. welfare becomes better off at the expense of the sugar industries in both countries and Mexico's welfare. When the policy of buying up excess sugar and the production control policy are compared, the U.S. government could pursue the latter in light of the nation's welfare; however, the gain is very small compared to the loss born by industries and Mexico. Both alternative policies do not satisfy pareto optimality and the overall loss outweighs the gain by the United States as a nation.

The impact of changes in U.S. price stabilization policy is illustrated in Table 5-5 with the assumption that the U.S. government maintains the minimum quotas for the rest of the world no matter how much Mexico exports. When the U.S. government introduces alternative policies to price support, U.S. welfare gains but both the U.S. and Mexican sugar industries as well as Mexico's welfare lose to a larger degree. An extreme result comes from Scenario 13. In this scenario, the U.S. sugar price support becomes extremely costly if the U.S. government reserves the minimum quota for the rest of the






84


world. As a result of cheap sugar from the rest of the world flowing into the U.S. market, the U.S. price will fall far below the support price. The loss born by the Mexican sugar industry particularly stands out: nearly a half of its expected revenue disappears due to restricted access to the U.S. market. When fostering the sugar industry in its own country and Mexico as a neighboring trade partner, the U.S. government needs to accept sugar from Mexico more than from the rest of the world. This requires sensitive negotiations among exporters.

Overall, the Mexican sugar industry and welfare are more prone to changes in the Mexican sweetener market and U.S. sugar policy than the U.S. sugar industry and U.S. welfare. Among tested changes, Mexico's HFCS adoption and U.S.'s quota allocation policy hold the greatest effects.

Effects of Mexico's production improvement and HFCS adoption on the Mexican sugar industry and welfare are illustrated in Figures 5-17, 5-18, and 5-19. The rate of production improvement ranges from the average rate realized between 1997 and 2001 (baseline) to additional 0.5, 1.0, and 1.5 percent to the average rate; the HFCS adoption ranges from 25 percent share of indirect sweetener consumption (baseline) to 30, 40, 45, and 50 percent (HFCS adoption situation). The effects on the Mexican sugar industry are shown in Figure 5-17. Expected accumulated revenue increases as sugar production improves an additional I and 1.5 percent above the average rate. An increase in HFCS adoption also brings about an increase in revenue until the share attained by HFCS climbs up to 40 percent. When HFCS share reaches 50 percent, revenue shrinks compared to the baseline. Increased revenue caused by HFCS adoption is due to higher domestic sugar price and decreased revenue is due to decreased quantity of sugar demanded. The effects






85


on Mexico's welfare differ from those on the industry (Figure 5-18). Production improvement increases welfare, but not HFCS adoption, assuming that sugar is a primary and preferred source of sweetener. While the gains to the Mexican sugar industry from production improvement at 1.5 percent to the baseline (US$ 0.65 billion) can make up the loss to the industry itself from HFCS adoption, even at a 50 percent share (US$ 0.42 billion), gains to Mexico's welfare cannot. The net loss to the nation from HFCS adoption at 30 percent adoption share is forecasted US$ 5.12 billion (loss of US$ 5.41 billion from welfare and a gain of US$ 0.29 billion from the industry).' This loss is far greater than the sum of expected gains from production improvement at additional 1.5 percent, which is US$ 3.42 billion (gains of US$ 2.77 billion from welfare and US$ 0.65 billion from the industry). The Mexican government faces difficulties allowing faster HFCS adoption to happen in the domestic market (Figure 5-19).

Game Theory Analysis

Results from the game theory analysis are summarized in Tables 5-6 through 5-13. The analysis is based on the game setting played by the United States and Mexico with multiple strategies that correspond to Mexico's market situation (the Mexican government's strategy set) and the U.S. policy levers (the U.S. government's strategy set) in order to assess gainers and losers from trade. Calculated actual pay-offs to five payees (the U.S. HFCS industry, the U.S. sugar industry, Mexican sugar industry, U.S. costadjusted welfare, and Mexico's welfare) for each combination of U.S. and Mexico's strategies of the game are shown in Table 5-6. Values in the other tables are indexed relative to the baseline scenario. Three different forms of coalitions are considered:


Assuming that industry's revenue and nation's welfare can be added together.






86


coalition of countries, coalition of the sugar industries, and grand coalition that includes all five entities. Since the inclusion of the U.S. HFCS industry into a coalition is a determinant factor, both coalitions with and without the U.S. HFCS industry are also considered.

Pay-offs to the U.S. HFCS industry and the government of Mexico's welfare

fluctuate more than the other three payees (Table 5-6). Among the five strategies facing Mexico, introduction of tax on HFCS returns the best pay-offs to the Mexican sugar industry and welfare and thus the Mexican government would always choose this strategy. Yet, this policy lever may not be acceptable in the international trade environment. In the following games, Mexico's tax strategy is excluded.

Indexed pay-offs are shown in Table 5-7. Values in corresponding cells are

categorized in four ways: no change (100), gain (over 100, cells shaded in gray), slight loss (over 95 and under 100), and loss (less than 95, cells shaded with stripes). In this game, there is neither dominant strategy (a strategy that is chosen over other strategies) nor pure strategy (a single strategy chosen by the government that brings about improvement in the industry or industries as well as welfare in its own country: this strategy is played with probability of I for either U.S. or Mexican government). By inspecting the tendency of pay-offs, some interesting contrasts among payees rise to the surface without solving mathematically. In the United States, the government always prefers a production control strategy to the other two but the sugar industry is always better off with the price support strategy, no matter which strategy Mexico chooses. As Mexico increases sugar production or HFCS adoption, the U.S. sugar industry likely loses while the U.S. government and the HFCS industry never become worse off. The






87


U.S. sugar industry will lobby against a production control strategy for fear that it may lose up to five percent (high production-high adoption market situation). On the contrary, the Mexican government's choice of strategy will be accepted by the industry in most cases, except for the case of where the U.S. government buys sugar in the baseline scenario. Since HFCS adoption harms both welfare and the industry, the Mexican government will continue to struggle to suppress HFCS adoption in its market.

Results from the game played by two coalitions of countries (the United States and Mexico) are shown in Table 5-8. In this game, the U.S. HFCS industry is excluded from the game. As expected, the high production strategy becomes Mexico's pure strategy and thus the game is solved when the U.S. government chooses a strategy: production control policy (cells shaded with vertical stripes). This combination of strategies is coincidentally the best choice for the U.S. coalition, but not for the Mexican coalition. The Mexican government would prefer the U.S. government to choose either the price support or buying up excess sugar strategies.

Results from the game played by the sugar industry coalition and the government coalition are shown in Table 5-9. The high production strategy becomes a pure strategy for Mexico when a decision is made by the government coalition. The solution of the game is thus determined when the U.S. government chooses the strategy of buying up excess sugar. This result differs from the game played by the country coalitions mentioned above. By a government cooperating with the other government rather than with the industry in its own country, the Mexican government improves its pay-off by US $4.75 billion; the U.S. government loses by US $0.24 billion; and the Mexican sugar industry loses US $1.27 billion (Table 5-4). Since the loss born by the U.S. government is






88


relatively small, the choice by the Mexican government is made at the expense of its own sugar industry. On the contrary, the industry coalition would lobby for the U.S. production control strategy, assuming that the Mexican sugar industry promises to compensate the loss of US $0.01 billion (US $10 million) born by the U.S. sugar industry with expected gain of US $1.27 billion (Table 5-6). With this contrast of the results between a country coalition (Table 5-8) and a government/industry coalition (Table 5-9), the U.S. government would likely form a coalition with its own industry while the Mexican government would try to form a coalition with the U.S. government.

Results from the game played by the grand coalition that includes all players except the U.S. HFCS industry are shown in Table 5-10. The solution is the Mexican government high production strategy and the U.S. government buying excess sugar strategy. Although this solution does not satisfy pareto optimality for all four payees, the overall loss is minimal: the loss incurred by the U.S. sugar industry is US$ 0.31 billion (Table 5-6) and is theoretically compensated by the total gain of US$ 2.29 billion (US$

0.36 billion from the U.S. government, US$ 0.13 billion from the Mexican sugar industry, and US$ 1.80 billion from Mexico's welfare, Table 5-6) This solution corresponds to the government/industry coalition seen in Table 5-9. In other words, if the U.S. government sees the benefit from pooling gains with Mexico rather than with its industry and if redistribution of gains is possible among governments and industries, no one loses from the game. Such an arrangement and agreement in practice would be expected to be difficult to reach and implement.

When the U.S. HFCS industry is included in coalitions, the same solutions are reached: the combination of Mexican government high production strategy and U.S.






89


government production control strategy for country coalitions (Table 5-11), Mexican government high production strategy and the U.S. government buying up excess sugar strategy for government/industry coalitions, and Mexican government high production strategy and the U.S. government buying up excess sugar strategy for grand coalition. Yet, pooled gains become larger due to inclusion of gains from the U.S. HFCS industry, the pay-off matrices offer additional factors that need to be taken into the policy assessment process.

Results from the game played by country coalitions are shown in Table 5-11.

Although the solution is the same as the one without the U.S. HFCS industry included, pay-offs expected from Mexican government's other strategies become more attractive to the U.S. coalition. It is obvious that the U.S. coalition would receive a better pay-off if Mexico chooses the HFCS adoption strategy. This fact gives the U.S. coalition a strong incentive to influence Mexican government's choice of strategy.

Intensified conflict of interests between the industry coalition and the government coalition is illustrated in Table 5-12. Without the HFCS industry in a coalition, the industry coalition prefers the U.S. production control strategy to the buying up excess sugar strategy for the sake of slight gains. When gains from the HFCS industry are pooled, the industry coalition will lobby for strategies that involve Mexico's HFCS adoption, no matter which strategy the U.S. government plays. This is possible only if the industry coalition promises to compensate the Mexican sugar industry for the loss. If redistribution of gains among industries is feasible, the Mexican sugar industry may prefer being compensated to expecting protection from the Mexican government's strategy. Ultimately, if the Mexican government chooses to overlook the impact on




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E6IP5ION4_K7T0NE INGEST_TIME 2015-04-02T21:14:02Z PACKAGE AA00029867_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

ANALYSIS OF U.S.-MEXICO SUGAR TRADE: IMPACTS OF THE NORTH AMERICAN FREE TRADE AGREEMENT (NAFTA) AND PROJECTIONS FOR THE FUTURE By DAISUKE SANO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004

PAGE 2

Copyright 2004 by Daisuke Sano

PAGE 3

To my parents and sister

PAGE 4

ACKNOWLEDGMENTS I would like to express my gratitude to Dr. Thomas H. Spreen, chair of my supervisory committee, for providing me with resources throughout my program of studies. My dissertation would not have been possible without his guidance, unflagging enthusiasm and helpful ideas. I also would like to express my gratitude to Dr. Lisa A. House, cochair of my supervisory committee, for her generous and patient support during my course of studies. Under her supervision, my research has been completed in an efficient and successful manner. Special appreciation is extended to Dr. Chris O. Andrew for his encouragement throughout the program and for broadening my horizons. His inspiration has been a compass after the direction of my career had been shifted. I am also deeply grateful to Dr. Luis R. Garcia for providing me with insights and detailed data which were indispensable for this study; to Dr. Terry L. McCoy for his contributing to the depth of my study; and to Dr. Kenneth L. Buhr for his sharing suggestions for the breadth of my study. Kind support from Dr. Jeffrey R. Burkhardt and staff in the department for their various services are also appreciated. I cannot fail to thank John S. Lander, who opened my eyes to studying overseas, for his unconditional support and wisdom. Friendship from my colleagues in the program is a cherished treasure. Last but not least, the financial support received from the Food and Resource Economics Department throughout my program has been greatly appreciated. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS iv LIST OF TABLES vii LIST OF FIGURES ix ABSTRACT xii CHAPTER 1 INTRODUCTION 1 Background 1 Problem Statement 5 Researchable Questions 5 Objectives 6 Organization of the Study 6 2 SUGAR INDUSTRIES AND SWEETENER MARKETS IN THE UNITED STATES AND MEXICO 8 The Mexican Sugar Industry and Sweetener Market 9 Mexico's Sugarcane Production 9 Mexico's Sugar Production 10 The Mexican Sugar Industry and Government Involvement 12 Mexico's Sugar Consumption 14 Mexico as a Sugar Exporter 15 Development and Adoption of High Fructose Com Syrup 17 The U.S. Sugar Industry and Sweetener Market 19 3 CONCEPTUAL AND THEORETICAL FRAMEWORK 32 Conceptual Framework 32 Theoretical Framework 35 Sweetener Market Analysis 35 Sweetener demand 35 Sweetener supply 36 U.S.-Mexico Bilateral Sugar Trade System 37 V

PAGE 6

4 EMPIRICAL MODELS AND DATA SOURCE 47 Empirical Models 47 U.S. Sweetener Demand Model 47 Mexican Sweetener Demand Model 48 U.S. Sweetener Supply Model 50 Mexican Sweetener Supply Model 52 U.S.-Mexico Bilateral Sugar Trade Model 53 Simplifying assumptions 54 Model calibration 56 Simulated Scenarios 58 Game Theory Analysis ^1 Sources of Data ^-^ 5 EMPIRICAL RESULTS AND INTERPRETATION 75 Demand and Supply Analyses 75 Bilateral Sugar Trade Analysis 77 Game Theory Analysis ^5 6 CONCLUSIONS AND IMPLICATIONS FOR POLICY 115 Conclusions and Implications for Policy 115 Impact of Changes in Trade Regime 116 Mexico's Export Potential 1 17 The Impact of Changes in Mexican Market Situation 1 17 The Impact of Changes in U.S. Sugar Policy 118 Alternative Sugar Policy by the United States 120 Limitation of the Study and Suggestions for Future Research 122 APPENDIX A MAJOR EVENTS IN THE SUGAR INDUSTRY HISTORY IN MEXICO AND THE U.S 123 B CORN STATISTICS 125 C DERIVATION OF INVERSE LINEAR EQUATIONS 128 BIOGRAPHICAL SKETCH 136 vi

PAGE 7

LIST OF TABLES Table ms& 2-1 Cane Sugar Production in Selected Countries, 1997-2000 Average 25 2-2 World Sugar Consumption in 2000 27 2-3 Quota and Tariff Schedule Imposed on Mexican Sugar Exported to the U.S 27 4-1 Assumptions for Baseline (Status quo) Scenario 66 4-2 Assumptions for Mexican Sweetener Market Situations 67 4-3 Assumptions for U.S. Sugar Policies 69 4-4 Listing of Examined Scenarios 70 4-5 Strategies for the Sugar Trading Game 71 4-6 Data Sources for U.S. Demand 72 4-7 Data Sources for Mexican Demand 72 4-8 Data Sources for U.S. Supply 73 4-9 Data Sources for Mexico Supply 73 410 Data Sources for Miscellaneous 74 51 Summary of the U.S. Supply-Demand Analysis 91 5-2 Summary of the Mexican Supply-Demand Analysis 92 5-3 Impact of Changes in Mexican Sweetener Market on Pay-offs to the Industries and Nation's Welfare [Billion US$] 101 5-4 Impact of Changes in U.S. Price Stabilization Policy on Pay-offs to the Industries and Nation's Welfare [BilHon US$] Flexible Quota Allocations 102 5-5 Impact of Changes in U.S. Quota Allocation Policy on Pay-offs to the Industries and Nation's Welfare [Billion US$] Minimum Quota Allocations 103 vii

PAGE 8

5-6 Pay-off Matrix for the Trade Policy Game [Billion US$] 107 5-7 Indexed Pay-off Matrix for the Trade Policy Game without Coalitions [Baseline = 100] 108 5-8 Indexed Pay-off Matrix for the Trade Policy Game Played by Two Coalitions of Countries without the U.S. MFCS industry [Baseline=100] 109 5-9 Indexed Pay-off Matrix for the Trade Policy Game Played by the Industry Coalition and the Government Coalition without the U.S. HFCS industry [Baseline=100] HO 5-10 Indexed Pay-off Matrix for the Trade Policy Game Played by the Grand Coalition without the U.S. HFCS industry [Baseline=100] 11 1 5-11 Indexed Pay-off Matrix for the Trade Policy Game Played by Two Coalitions of Countries with the U.S. HFCS industry [Baseline=100] 112 5-12 Indexed Pay-off Matrix for the Trade Policy Game Played by the Industry Coalition and the Government Coalition with the U.S. HFCS industry [Baseline=100] 113 5-13 Indexed Pay-off Matrix for the Trade Policy Game Played by the Grand Coalition with the U.S. HFCS industry [Baseline=100] 1 14 C-1 Coefficients for Inverse Linear Functions -U.S.129 C-2 Coefficients for Inverse Linear Functions -Mexico130 viii

PAGE 9

LIST OF FIGURES Figure page 2-1 Distribution of Individual Farm Size of Mexican Sugarcane Growers 24 2-2 Map of Mexico's Sugar Producing and Processing States 24 2-3 Annual Rainfall and Irrigation Rate in Sugarcane Fields 25 2-4 Mexican Sugar Production, 1988-2002 26 2-5 Main Use of Sugarcane Derivatives in Mexico 26 2-6 U.S. Refined Sugar and HFCS Use per Capita 28 2-7 U.S. HFCS Production 28 2-8 U.S. HFCS Supply 29 2-9 U.S. HFCS Export 29 2-10 Transition of U.S. HFCS Export 30 2-11 Consumption of Sugar and HFCS per Capita in Mexico 30 212 Transition of Prices of Sugar and HFCS 31 31 Conceptual Framework for the Analysis in this Study 45 32 Two-country Trade Model (a) with Quota System and (b) without Quota System. 46 41 Image of Model Calibration 65 42 Forecasted Indirect Sugar and HFCS Consumption in Mexico under Alternative Scenarios 68 51 U.S. Sugar Import Forecast (Scenario 1 "Baseline") 93 5-2 U.S. Sugar Import Forecast (Scenario 2 "P-S-F") 93 5-3 U.S. Sugar Import Forecast (Scenario 3 "A-S-F') 94 ix

PAGE 10

5-4 U.S. Sugar Import Forecast (Scenario 4 "PA-S-F") 94 5-5 U.S. Sugar Import Forecast (Scenario 13 "PA-S-M") 95 5-6 U.S. Sugar Import Forecast (Scenario 8 "PA-B-F") 95 5-7 U.S. Sugar Import Forecast (Scenano 12 "PA-C-F") 96 5-8 U.S. Sugar Import Forecast (Scenario 16 "T-S-F") 96 5-9 Forecasted Equilibrium Sugar Prices in the U.S. and Mexican Markets (Scenario 4 "PA-S-F') 97 5-10 Forecasted Equilibrium Sugar Prices in the U.S. and Mexican Markets (Scenario 8 "PA-B-F") 97 5-11 Forecasted Equilibrium Sugar Prices in the U.S. and Mexican Markets (Scenario 12 "PA-C-F") 98 5-12 Forecasted Equilibrium Sugar Prices in the U.S. and Mexican Markets (Scenario 16 'T-S-F") 98 5-13 Forecasted Sugar Demand and Supply for both the United States and Mexico (Scenario 1 "Baseline") 99 5-14 Forecasted Sugar Demand and Supply for both the United States and Mexico (Scenario 4 "P-S-F") 99 5-15 Forecasted Equilibrium Sugar Prices in the United States and Mexico (Scenario 14 "PA-B-M") 100 5-16 Forecasted EquiHbrium Sugar Prices in the United States and Mexico (Scenario 15 "PA-C-M") 100 5-17 Absolute Effects of Production Improvement and HFCS Adoption on Pay-off to the Mexican Sugar Industry 104 5-18 Absolute Effects of Production Improvement and HFCS Adoption on Pay-off to Mexico's Welfare 105 5-19 Relative Gain and Loss of Pay-offs to the Mexican Sugar Industry and Welfare Caused by Production Improvement and HFCS Adoption 106 B-I Recent Com Production and Consumption for Selected Countries 126 B-2 Food and Industrial Com Use in the U.S., 1980-2002 126 B-3 Com Price (No.2 Yellow) in Chicago Market, 1981-1998 127

PAGE 11

B-4 Exports of Products Made from Com in 2002 127 xi

PAGE 12

CHAPTER 1 INTRODUCTION This chapter introduces the research problem for U.S. -Mexico sugar trade. Background, the researchable questions, and objectives are provided, followed by the organization of the dissertation. Background Trade issues surrounding the world sugar market are often seen as classic examples in agricultural economics, yet the market still provides us with important questions today. In the case of the U.S.-Mexico sugar trade, the main issues boil down to two aspects: the provisions of North American Free Trade Agreement (NAFTA) and the role of high fructose com syrup (HFCS), a substitute for sugar, in sweetener markets. In the following section, a summary of this area, focusing on these two aspects, is provided. NAFTA was implemented in 1994, creating a freer trade environment among Mexico, the United States, and Canada by eliminating tariffs. Among other regional trade agreements involving North America, Latin and Caribbean countries, it is the least ambitious on paper of the major trade agreements, but it has been the most successful adhering to the negotiated schedule in lowering tariffs (McCoy, 2002). In terms of agricultural trade between Mexico and the United States, many tariffs were eliminated immediately while others being phased out over periods of 5 to 15 years (USDA, 2001c). In addition to a transition period of up to 15 years for certain products, NAFTA has 1

PAGE 13

special safeguards to protect import-sensitive crops, including sugar, which are defined under side agreements between Mexico and the United States (USDA, 2001c). When trading sugar with the United States, the Mexican sugar industry faces two counteracting conditions under the NAFTA regime: increased access to the U.S. market ^ which would facilitate sugar exports at favorable prices; and the pressure of increased imports of MFCS from the United States which have been gaining an increasing share of Mexico's sweetener market since 1994. Under the provisions of NAFTA, both an overquota tariff for Mexican sugar which enters into the United States and a tariff on exported HFCS which enters the Mexican market are regulated in such a way that both tariffs will be reduced to zero by 2008 and 2004, respectively. In addition to the rules of the tariffs, Mexican sugar is subject to U.S. import quota allocations. Mexico is allowed to access two kinds of quotas, depending on Mexico's domestic balance in the sweetener market: if Mexico's sugar production exceeds its sweetener consumption (the sum of sugar and HFCS consumption in two consecutive years -"net surplus sweetener producer status"), Mexico receives 25,000 MT of sugar import quota; and if not, Mexico receives 7,258 MT of quota. Additionally beginning in 2000, the sugar import quota expands from 25,000 MT to 250,000 Mt as long as Mexico satisfies the conditions of a net surplus sweetener producer. Mexico can export over the 7,258 MT quota without attaining net surplus sweetener, but any sugar exported in this scenario would be subject to taxation in the form of tariffs as mentioned above. In 2008 when all the restrictions, i.e. both tariffs and quotas, are lifted, Mexico will have free and unlimited access to the U.S. sugar market. In Mexico, the sugar industry has played an important role in the economy and the politics of the country. In spite of experiencing drastic economic and political changes.

PAGE 14

3 including NAFTA, devaluation, privatization of the sugar cane processing industry in the 1990s, and several changes in the policy regime, sugar production has shown steady expansion over the past 10 years: Mexican sugar production expanded from 3.2 million MT in 1990 to 4.7 million MT in 2000 (COAAZUCAR, 2003a). A significant amount of surplus sugar destined to export has been generated since 1995, ranging from 200,000 MT in 1995 to over 1.1 million MT in 1998 (COAAZUCAR, 2003a). These records may appear favorable; however, Mexico stood to benefit little from NAFTA. From 1996 through 1999 Mexico successfully received a 25,000 MT import quota as a result of attaining net surplus producer status, yet it did not enjoy the expanded quota (250,000 MT) from 2000 through 2002 (USDA, 2003a), the amount equivalent to 20 percent of the U.S. minimum sugar import requirement under GATT, because Mexico's production fell short relative to its sweetener consumption. This indicates that Mexico missed the opportunity to export sugar under-quota even though it generated a significant surplus. Combined with a slump in production that occurred in 1999 and 2000, the Mexican sugar industry underwent an economic crisis. In September 2001, the Mexican government expropriated 27 of 60 of Mexico's functioning sugar mills in order to maintain the industry (USDA, 2002b). Today, the circumstances surrounding the sugar industry remain unfavorable. At the industry level, many mills are financially vulnerable and suffer from low efficiency of production due to old technology or poor infrastructure. Foreign investment has not been successfully encouraged to provide capital for needed investments in new capital equipment. At the farm level, production efficiency is low due to fragmented farmland, which is a result of the Ejido system (Mexico's agrarian law) and social security program specifically tailored to sugar cane growers. Lack of credit

PAGE 15

4 and old technology also contribute to low productivity. Although the price of sugar at the wholesale level has been privatized, the sugar price paid to growers is still controlled by a government agency, and hence farmers have little incentive to grow sugar cane other than to receive social benefits from the Mexican sugar program. At the national level, the Mexican government faces a dilemma between gaining competitiveness in the international market and maintaining social stability through offering employment and financial supports to the livelihood of a large number of growers and related workers. Overall, there has been little benefit to the Mexican sugar industry resulting from NAFTA. The U.S. sugar market, where a large quantity of sugar is traded by a large number of sellers, has maintained commodity balance by assigning tariffs and import quotas to foreign sellers and maintaining domestic price support through the U.S. sugar program. As a result of GATT, the United States committed to accept a minimum import quota of 1.256 million MT of sugar in 1990; however, the U.S. sugar market has been maintained unchanged until today through successful lobbying efforts by the American Sugar Alliance (ASA), the sugar producers' primary alliance. In the meantime, HFCS had been gaining its share in the U.S. sweetener market since the early 1970s when commercial production of HFCS became possible by the advancement of wet-milling technology. Today, more than 50 percent of caloric sweetener consumption in the United States is derived from com syrup including HFCS (Congressional Research Service, Library of Congress, 1999). A similar phenomenon appears to be beginning in Mexico. The implementation of NAFTA resulted in opening the door for HFCS consumption in Mexico where nearly all caloric sweetener

PAGE 16

5 consumption was derived from domestically-produced sugar before 1994. Reflecting this threatening trend of replacing domestic sugar consumption with HFCS, in 1996, the Mexican government imposed tariffs on HFCS claiming that U.S. companies were dumping HFCS at an unfair price and affecting the export volume and value of Mexican sugar. This action evolved into a trade dispute between the United States and Mexico and ended when the WTO panel ruled against Mexico's claim (Garcia Chaves et al., 2002 and 2004). Overall, NAFTA has not brought about significant changes in the U.S. sugar market because the Mexican exporters have been unable to significantly expand shipments to the United States. Rather, attention was poured into issue of HFCS and its immediate impact on the Mexico's sweetener market. In this study, the direction of U.S. -Mexico sugar trade is examined using quantitative methods, with close attention to issues related to NAFTA and HFCS adoption in Mexico. Demand and supply analyses in both countries and a bilateral trade model using mathematical programming provide insights for the market balance in the future including political implications. Aggregated results from various simulations on the trade model are examined using a game theory analysis to investigate possible policy recommendations through assessing gainers and losers in sugar trade. Problem Statement The future outlook for the U.S.-Mexico sweetener market needs to be quantitatively analyzed in a manner that includes influential factors such as trade agreements under NAFTA; trends in HFCS consumption in Mexico; and other related economic and political issues in the sweetener markets. Researchable Questions The study attempts to answer the following set of questions.

PAGE 17

6 1. What was the impact of changes in the trade regime in the U.S. and Mexican sweetener market since NAFTA was implemented in 1994? 2. How much surplus sugar can Mexico generate and how much sugar will cross the border both underand over-quota? What will happen after 2008 when all the restrictions are eliminated on Mexican sugar? 3. What will be the impact of changes in Mexico's market on both the United States and Mexico? How much influence will HFCS adoption cause in both the U.S. and Mexican sweetener markets? 4. What will be the impact of changes in U.S. sugar policy on both the United States and Mexico? 5. Is there alternative sugar policy for the United States to current price support? Objectives The primary objective of the study is to develop a bilateral trade model of the U.S.Mexico sugar industry that reflects provisions of NAFTA, as well as related market conditions in order to forecast the outlook of the sweetener market through various simulations, encompassing hypothetical changes in Mexican sweetener situations and U.S. sugar policy. Secondly, the study aims to provide policy recommendations by examining aggregated results from these simulations, paying attention to identify gainers and losers under different scenarios. By doing so, the study hopes to illustrate conflicts of interest among the various players in the U.S.-Mexico sweetener market. Organization of the Study The remainder of the dissertation is organized as follows. In chapter 2, the sugar industries in both the United States and Mexico are introduced in the context of the sweetener market in each country as well as the integrated market, paying close attention to historical and political perspectives. In chapter 3, the conceptual and theoretical

PAGE 18

7 framework employed to analyze U.S. -Mexico sugar trade is presented. In chapter 4 and 5, empirical procedures as well as data set used in the study and the results form the empirical study are presented. Lastly, conclusions and implications for policy are discussed in chapter 6.

PAGE 19

CHAPTER 2 SUGAR INDUSTRIES AND SWEETENER MARKETS IN THE UNITED STATES AND MEXICO Sugar, one of the basic commodities with a long history of utilization, is traded in mature markets in many parts of the world with established business practices and networks. The recent trend towards freer markets in the international trade area has not left the industry unchanged. The sugar industries in the United States and Mexico are not exceptions. They have experienced more changes in the face of this recent trend towards rapid trade liberalization. In fact these two industries have become more economically inseparable than ever before as the sweetener markets in the United States and Mexico have been integrated under North American Free Trade Agreement (NAFTA). In this chapter, sugar industries in both the United States and Mexico are examined in the context of the sweetener market as well as the integrated market, paying close attention to historical and political perspectives. First, the Mexican sugar industry and sweetener market are introduced with fundamental characteristics of the structure and government involvement. The status of Mexico as a sugar exporter is also presented in conjunction with Mexico's relation to the U.S. market under the provisions of NAFTA. Next, the development and adoption of high fructose com syrup (HFCS) is presented. Lastly, the U.S. sugar industry and sweetener market is introduced with emphasis on the current political environment surrounding that market. 8

PAGE 20

9 The Mexican Sugar Industry and Sweetener Market The Mexican sugar industry has a long history of playing an important role in the nation's economy and policy. A large number of small-scale sugarcane growers and antiquated sugar milling facilities still remain as the driving force of an industry under the protection of the government. Mexico's Sugarcane Production Most regions of Mexico have a suitable climate for sugarcane production, except for the northern region of the country where the climate tends to be cooler and drier. Sugarcane production is widely spread across the southern and coastal regions of the country under different environmental conditions. Production occurs at altitudes that vary from sea level to over 1,300 meters (4,333 feet) above sea level; annual average temperature from 17 to 35 °C (from 63 to 95 °F); and annual rainfall from 500 mm to over 3,000 mm (from 20 to over 118 inches) rainfall. In the 2001/02 crop season, total net sugarcane production in Mexico was 41.5 million MT, with a yield of 4.9 million MT of raw sugar (COAAZUCAR, 2003a). Among producing regions, the state of Veracruz has the largest production accounting for 38.5 percent of national production (COAAZUCAR, 2003a). The total area harvested in Mexico was 610,121 ha in 2002 (COAAZUCAR, 2003a), making sugarcane the second largest agricultural crop by area, following only coffee (maize, wheat, alfalfa, beans, and oranges follow sugarcane) in 2002 (SAGARPA, 2003). Although sugarcane is a relatively low maintenance crop, sometimes referred to as the "lazy man's crop," varieties have been developed with higher disease resistance, higher sucrose content, and lower fiber content, yielding better sugar production. In order to avoid poor yields, sugarcane fields are usually replanted every six or seven years (Greene, 1998).

PAGE 21

10 The total number of sugarcane growers is reported as approximately 158,000 (COAAZICAR, 2003b), which is equivalent to roughly 2 percent of the total labor force in the agricultural sector (INEGI, 2003). If related workers such as sugarcane cutters, cane-transport employees, factory workers and administrative, and technical and management personnel are included, total employment in the sugar sector exceeds 1,000,000 (Garcia Chaves et al., 2002) and accounts for more than 14 percent of agricultural labor. Land area per grower ranges from less than 1 ha, which accounts for 3.6 percent of the total sugarcane area, to over 15 ha, which accounts for 17.5 percent of the land, averaging 3.9 ha per grower (COAAZUCAR, 2003b). When the number of growers is allotted to each land size category, a skewed distribution is revealed along the land scale spectrum with many small-land holders and a few large-land holders (Figure 2-1). A large number of small-scale sugarcane growers were created as a result of the Mexican revolution and the sugar program that evolved after the revolution: the communal land (ejido) has been divided and distributed among farmers since the revolution and the Mexican sugar program offers social security and medical services to each grower proving to be a large incentive for farmers to grow sugarcane. Mexico's Sugar Production There are 60 operating sugar mills located across 15 states in the nation (Figure 22). Sugar mills are responsible not only for milling sugarcane, but also supervising sugarcane cultivation and organizing the harvest. This includes inspecting and advising on cultivation, scheduling harvest dates, pooling and arranging laborers, and providing trucks and drivers for the harvest.

PAGE 22

11 In the 2001/02 crop season, average sugarcane yield was 70.32 MT per hectare in Mexico, yielding 4,872,388 MT of sugar (COAAZUCAR, 2003a). Imgation is one of the factors which influences cane yield; however, irrigation systems are found only in the area where less rainfall is expected (Figure 2-3): 30 percent of total sugarcane area has no irrigation system and 25 percent has full irrigation system (COAAZUCAR, 2003c). Harvest is the most labor-intensive part of sugarcane production; the harvest season lasts for about six months starting between November and January and ending in June in most regions, depending on weather and size of enterprise. Harvest competes for grower's labor with other winter crops since many growers are also engaged in production of crops such as maize, vegetables and citrus. Most of the harvest is carried out manually; only 9 percent of total sugarcane processed at mills is harvested by machine; 27 out of 60 mills do not employ a machine harvester at all; however, a cane loader is used in most cases (COAAZUCAR, 2003d). Since the mills own machine harvesters, cane loaders, and trucks, growers do not need to own them; however, it means growers have no means to harvest and sell their sugarcane without the mill's assistance and coordination. Similar situations regarding grower's capacity in harvest are found in other crops such as citrus. Upon harvest, sugarcane is bought by mills from growers and processed into sugar. Sugarcane quality is vital to the sugar production process; high sucrose content cane leads to high sugar production. Yet, it is often the case in Mexico that trucks endure long waiting times to unload cane due to limited milling capacities. The average wait time observed in 2001 was almost 30 hours across mills (COAAZUCAR, 2003e). The longer trucks wait, the lower the quality of sugarcane becomes. Scattered and fragmented

PAGE 23

sugarcane fields and poor road conditions also contribute to longer transportation time, and hence decreased sugarcane quality. Mexico's sugar production was approximately 4.8 million MT, raw equivalent, in 2002, ranking it seventh among all cane sugar-producing nations; Brazil and India are by far the largest cane sugar-producing nations, followed by China, the United States, Thailand and Australia (Table 2-1). Production in Mexico has been increasing for the past few decades (Figure 2-4). The Mexican Sugar Industry and Government Involvement Mexico's sugar production accounts for 0.5 to 0.7 percent of its gross domestic product (Garcia Chaves et al., 2002; Farm Foundation, 2003). Since privatizing in the late 1980s, mills have neither successfully accumulated capital nor renewed their equipment leaving the industry financially vulnerable. In 2001, the Mexican government expropriated 27 mills, which represented approximately 50 percent of sugar production in Mexico. In February 2002, the Government of Mexico announced a National Sugar Policy for 2002 2006 which included a series of shortand longterm measures to help Mexico's ailing sugar industry with the main objective of regulating the sugar market and making the sugar sector profitable (USDA, 2002b). Today the sugar industry remains important in Mexico because it is considered crucial for maintaining social stability due to the large number of growers and related workers. Among the public organizations that deal with the Mexican sugar industry, Commite de la Agroindustria Azucarera (COAAZUCAR) plays an important role by monitoring and compiling sugarcane and sugar production data at each mill. Although the industry has been privatized, COAAZUCAR is in charge of determining the cane price. It took over the task from the former body, Azucar, S.A. which was dismantled in 1991

PAGE 24

13 when sugar mills were privatized. Growers are paid by a fixed portion of the reference sugar price calculated from aggregated sugarcane harvested and processed in the specific mill. Details for the cane price setting formula are shown in equations [2.1] and [2.2] (Garcia Chaves et al, 2004): Cane price / ton = (KARBE/ ton of cane)*(Price of KARBE)*(0.57) (2.1) KARBE/ ton of cane = (Pol) *(FF)*(FP)*(EBF)*(TF) (2.2) where KARBE is kilogram of recoverable standard sugar basis (Pol 99.4 percent) for net ton of cane; Pol is polarization of cane (apparent percentage of sucrose in cane); FF is the fiber factor; FP is the purity factor; EBF is mill efficiency; and TF is the transformation factor. As seen in equation [2.1], currently growers are paid 57 percent of the wholesale price per kilogram of standard sugar. Although in 1991 growers began to be paid according to the quality of cane produced as opposed to solely on weight as decreed in the amendments to Decreto Canero, the Sugarcane Growers Law, growers have little incentive to produce higher quality cane. The cane price is capped at 57 percent of average quality cane for the specific mill, not a price reflecting each grower's sugarcane. Thus, a main incentive for Mexican sugarcane growers is to receive social benefits and medical services from the government as opposed to producing quality cane. The wholesale price of sugar produced at mills has been liberalized since 1997; however, it still quotes a reference price calculated based on the formula published by the secretariat of Commerce (SECOFI). The price is determined by considering both the recent domestic price and the expected export price, which is the composite of the U.S. and world price (Garcia Chaves et al, 2004), as shown in equation [2.3].

PAGE 25

Pr = a.Pn + (l-a)*P'x (2-3) where Pr is wholesale price per kilogram of standard sugar to be used as the reference for cane payment during the harvest; a is expected portion of harvest to be consumed nationally (a equals one if expected consumption is greater than expected production); Pn is reference price for standard domestic sugar calculated by comparing the OctoberSeptember average price of the previous year with the current year; (1-a ) is expected surplus as a portion of production; and P% is expected export price of sugar, which is calculated by a weighed average of the U.S. price (Contract No. 14) and the world price (Contract No. 11) with corresponding export quantities. The Instituto Medical y Seguro Social (IMSS), the Mexican Social Security Institute, provides both pension and medical services to all the employees in Mexico as well as to small farmers who grow sugarcane (Greene, 1998). It is often the case that IMSS clinics are located to the next to the mills (Greene, 1998). Borrel (1991) as well as Buzzanell and Lord (1995) have pointed out this special relationship as a source of inefficiency in the Mexican sugar industry. Mexico's Sugar Consumption Since Mexico is neither an importer of sugar nor producer of sugar beets, sugar consumed in the country is derived solely from sugarcane grown domestically. The main use of sugarcane derivatives are shown in Figure 2.5. Sugarcane requires two steps in the refining process to obtain the refined sugar used by households and industries. Raw sugar, which is the product of the extraction process, is either stored or exported to other countries where refinery facilities are available. Mexico consumes two kinds of refined sugar, called standard sugar and white sugar. Standard sugar has a slight impurity.

PAGE 26

15 whereas white sugar is as purely refined as ordinary refined sugar found elsewhere. Both kinds of refined sugar are indistinctly consumed by households as well as domestic bulk sugar users such as soft-drink manufacturers and confectionaries. This unique aspect is different from the U.S., where a purer form of sugar, equivalent to white sugar in Mexico, is what most households purchase. Molasses, the heavy dark viscous liquid residue discharged by the centrifugal from which no more sugar can be obtained by simple means (Polopolus and Alvarez, 1991) is utilized in rum making in Mexico. Mexico's national total sugar consumption was approximately 4,500,000 MT, raw equivalent, in 2001 (FAO, 2003), making it the seventh largest sugar consumption country/ region in the world (Table 2-2). Per capita sugar consumption was 44.6 kg (98.5 lbs.), raw equivalent, in 2000; relatively high among other major sugar-producing countries (Table 2-2). When other kinds of sweeteners such as HFCS are included, the largest per capita consumer is the United States, followed by Cuba, Brazil, Australia, and Mexico (Table 2-2). Mexico as a Sugar Exporter Surplus raw sugar is either exported to the world market, primarily to the U.S. market due to higher price, or stored as stock. The magnitude of sugar exports from Mexico depends on the size of surplus determined by domestic production-consumption balance and the quota limitation imposed on all sugar imports entering the U.S. market. Geographically Mexico holds a sugar export advantage to the United States. One of the main shipping ports in Mexico is Veracruz, located facing the Gulf of Mexico, only 830 miles from New Orleans and 1 130 miles from South Florida where sugar refinery facilities are located. Furthermore, the state of Veracruz produces approximately 40 percent of Mexico's domestic production of sugar.

PAGE 27

16 Another advantage Mexico attained is the preferred trade conditions under the North American Free Trade Agreement (NAFTA) implemented in 1994. NAFTA created a freer trade environment among Mexico, the United States and Canada by eliminating tariffs. In terms of agricultural trade between Mexico and the United States, many tariffs were eliminated immediately, while others were scheduled to be phased out over periods of 5 to 15 years (USDA, 2001a). Mexico benefited from exporting its surplus sugar to the U.S. market at a higher price and at a lower tariff rate which will be reduced each year and eventually gives Mexico free and unlimited access to the U.S. market beginning in 2008 (Table 2-3). The NAFTA agreement is a double-edged sword to the Mexican sugar industry, however, creating counteracting conditions in the Mexican sweetener market: one is the increased access to the U.S. market and the other increased access of high fructose com syrup (HFCS) from the United States. The quantity of duty-free sugar exported from Mexico is limited by quotas which vary depending on Mexico's balance in the domestic sweetener market not the sugar market; Mexico receives a larger quota if its domestic sugar production exceeds domestic consumption of sweetener, including HFCS (called "net surplus sweetener producer status") in two consecutive years. At the same time, the agreement gives U.S. HFCS producers free access to the Mexican sweetener market beginning in 2004 as a tariff imposed on HFCS imported from the United States is also being phased out. If Mexico fails to attain net surplus sweetener producer status, it receives a sugar import quota of only 7,258 MT rather than 250,000 MT and thus most of Mexican sugar exported to the U.S. market is subject to over-quota tariffs until 2008 when all the restrictions on Mexican sugar are lifted.

PAGE 28

: 17 Development and Adoption of High Fructose Corn Syrup Sweeteners are generally classified into two categories, caloric and non-caloric; common sweeteners in the former group are sucrose, invert sugar, lactose, maltose, and sorbitol; and aspartame and saccharine in the latter. Sucrose is found in various forms of sugar such as raw sugar, granulated sugar and brown sugar derived form sugarcane or sugar beets, or in honey and maple sugar. Invert sugar such as dextrose, glucose, fructose, and HFCS are made form starch through chemical processes. HFCS is produced by converting a portion of naturally occurring glucose in starch into fructose through a com wet milling process (Congressional Research Service, 1999). Lactose, maltose and sorbitol are found naturally in certain kinds of food and give food a sweet taste. Noncaloric sweeteners, sometimes called artificial sweeteners, such as aspartame, are often used for special dietary purposes. Commercially produced and rapidly adopted since the early 1970s in the United States, HFCS became an important player in the sweetener market among sugar substitutes (Figure 2-6). Production of HFCS has increased from 51,000 MT in 1970 to nearly 8.7 million MT in 2001 (Figure 2-7). HFCS production expanded during the 1980s as a substitute for sugar used in the soft-drinks. Today, about 75 percent of total HFCS and 90 percent of HFCS-55 (55 percent fructose) supplied in the United States are consumed in soft-drink market (Buzzanell, 2002; Congressional Research Service, 1999). HFCS-42 (42 percent fructose), which is roughly 90 percent as sweet as sugar, is used mainly in beverages (44 percent), processed food products (21 percent), and other products including cereal and bakery products (Buzzanell, 2002; Congressional Research Service, 1999). As a result, HFCS and two other corn-derived sweeteners, glucose syrup and dextrose, accounted for 55 percent of total U.S. caloric sweetener use in recent years

PAGE 29

18 (Congressional Research Service, 1999). Due to its liquid form, HFCS is considered as a close substitute of sugar, though not a perfect one. Crystalline fructose, fine white crystals of pure fructose is slightly sweeter than sugar and is another potential substitute for sugar; however, crystalline fructose is more expensive than sugar and behaves differently from sugar in most baking and other manufactured food uses thus limiting its use as a sugar substitute (USDA, 1997). HFCS is predominantly produced in the United States, which accounted for 74 percent of the world HFCS production in 2001(Buzzanel, 2002). HFCS is produced in wet-milling faciUties located in com growing regions in the U.S. HFCS producers outside the United States and their recent production levels are: Japan (766,000 MT), Canada (400,000 MT), Argentina (312,000 MT, estimated), Mexico (291,000 MT, estimated), and European Union (293,000 MT) (Buzzanel, 2002). In the United States, most HFCS is supplied and consumed domestically and only a small fraction is exported to Mexico and Canada, the NAFTA member economies (Figure 2-8 and 2-9). The difference in trade between these two countries is that Canada and the United States exchange a similar amount of HFCS across the border while Mexico is a net buyer. Although the quantity exported to Mexico, 122,800 MT in 2001, accounts for only 1.5 percent of total HFCS demanded in the United States, the quantity accounts for 60 percent of total HFCS export from the United States (Figure 2-10). This amount is equivalent to about 40 percent of Mexico's domestic production ability (291,000 MT, Buzzanell, 2002). The introduction of HFCS into the Mexican sugar market brought about significant changes in the sugar consumption pattern in Mexico. Although HFCS is traded at a higher price than sugar in Mexico, it has been gaining an increasing share of the Mexican

PAGE 30

19 sweetener market over the past several years. The Mexican sugar industry has struggled to supply sugar to the domestic market at a price competitive with HFCS. Prior to 1994, nearly all of the caloric sweetener consumption was derived from domestically produced sugar; however, implementation of NAFTA resulted in opening the door for HFCS consumption in Mexico. As a result, Mexican per capita sugar consumption has decreased slightly since 1991, while per capita sweetener consumption has been increasing (Figure 2-1 1). Soft-drink manufacturers are believed to account for about onethird of the total sugar domestically demanded (Buzzanell, 2002). Currently, Mexico's HFCS consumption accounts for approximately 12 percent of total consumption of sweetener (approximately 25 percent of indirect sugar consumption) in 2001. In response to this threatening trend of replacing domestic sugar consumption with U.S. produced HFCS, the Mexican government imposed tariffs in 1996 on HFCS, based on a claim that U.S. companies were dumping HFCS at an unfair price and affecting the export volume and value of Mexican sugar. This action evolved into a trade dispute between the United States and Mexico and ended when the WTO panel ruled against Mexico's claim (Garcia et al., 2002 and 2004). Combined with a slump in production that occurred in 1999 and 2000, the Mexican sugar industry underwent an economic crisis. In September 2001, the Mexican government expropriated 27 of 60 Mexico's functioning sugar mills in order to maintain the industry (USDA, 2001b). Major events in sugar industry history in both Mexico and in the United States are summarized in Appendix A. The U.S. Sugar Industry and Sweetener Market Sugar production in the United States comes from two sources: sugarcane and sugar beets. The main sugarcane production regions are Florida, Hawaii, Louisiana, and Texas where the climate is tropical or semi-tropical. Louisiana and Florida produce

PAGE 31

20 approximately 48 and 45 percent of the total sugarcane production in 2001, respectively (USDA, 2001a). Florida's sugar production from sugarcane made up 50 percent of total cane sugar production in the United States in 2001 (USDA, 2002a), exceeding that of Louisiana (40 percent) due to a higher sugar recovery rate from cane. Sugar beet production regions are classified into four regions: Great Lakes, Upper Midwest, Great Plains, and Far West. The Upper Midwest, which includes Minnesota and North Dakota, produces approximately 47 percent of total sugar beet production; the Far West, which includes California, Idaho, Oregon and, Washington, produces approximately 26 percent of total sugar beet production in 2001 (USDA 2002a). Total sugar production from sugarcane and sugar beets was 4,017,000 short tons (3,615,000 MT) and 4,000,000 short tons (3,600,000 MT), respectively (USDA, 2002a). The proportion of sugar produced in the United States from sugarcane and sugar beets is about equal. Compared to Mexico, sugarcane production in the United States is regionally concentrated and highly mechanized as well as vertically integrated. In the case of the Florida's sugar industry, sugarcane is grown areas concentrated in flat land in south Florida. Sugarcane growing activities such as planting, harvesting and transporting harvested crop are fully mechanized, unlike Mexico. Six raw sugar mills, which are located near the sugarcane fields, possess an average daily processing capacity of 20,750 tons of sugarcane (Alvarez and Polopolus, 2002a). Two sugar refineries are located adjacent to two sugar mills. U.S. Sugar, the country's largest sugar producer operating in Florida, owns a fully integrated cane sugar refinery facility that manages not only sugar refining but also packaging and warehousing. With the facility built next to the existing

PAGE 32

21 sugar mill, this company is able to produce bagged sugar from sugarcane in a single plant location. Sugar is one of several commodities protected in the Farm Bill, which includes rice, cotton, dairy, tobacco, peanuts, grain such as wheat and com, and soybeans (Alvarez and Polopolus, 2002b). The sugar program operates through a loan program and market stabilization price (MSP) without production or acreage restraints, differentiating it from other programs that include target prices or deficiency payments along with export enhancement programs (Alvarez and Polopolus, 2002b). Loans are issued as nonrecourse loans' and are available to processors of domestically grown sugarcane at a rate of 18 cents per pound and to processors of domestically grown sugar beets at 22.9 per pound of sugar, respectively (Haley and Suarez, 2002). Since there are no production restraints, import quotas and tariffs are the main policy instruments utilized to comply with the provision that the program has to operate "no cost" to the government (Alvarez and Polopolus, 2002b).^ In 1990, the United States committed to accept a minimum import quota of 1.256 million MT of sugar as a result of GATT. The U.S. sweetener market has maintained a stable commodity balance, unlike in Mexico, even after NAFTA was implemented and during the trade dispute with Mexico over HFCS dumping. One of the reasons is the successful effort by the American Sugar Alliance (ASA), the sugar producer's alliance in the United States, which lobbies for the U.S. sugar program. The ASA is a strong coalition that includes sugarcane producers, sugar beet producers, and sugar processors, as well as com producers and HFCS ' As long as the raw sugar tariff-quota is set higher than 1.5 million short tons (Haley and Suarez, 2002). ^ The "no cost" provision was not included in Food and Agriculture Improvement and Reform Act (FAIR) in 1996 (Haley and Suarez, 2002).

PAGE 33

, 22 manufacturers (Moss and Schmitz, 2002). The other entity in the U.S. sweetener market is Coahtion for Sugar Reform (CSR), the industrial sugar user's coahtion. CSR opposes U.S. sugar poHcy, but has been unsuccessful at bringing a lower sugar price to the market from which industrial sugar users as well as consumers would benefit. Many studies have shown that producers clearly gain while consumers and industrial users lose in the U.S. sugar program: an analysis conducted by the General Accounting Office (GAO, 2000) indicates that food manufacturers could be substantial gainers from elimination of the sugar program (Moss and Schmitz, 2002). A history of sugar and HFCS prices are shown in Figure 2-12. Both sugar and HFCS prices exhibit a continuing declining trend. The decline in both U.S. domestic and export prices of the HFCS price is due to sophistication of wet-milling technology in combination with decreases in the tariff schedule under NAFTA for the latter. HFCS has been marketed at a lower price than raw sugar in the U.S. market. Although the U.S. HFCS industry sells at a lower price than sugar, it still benefits from the sugar program because the price of sugar is maintained higher than it would be without the program. This structure confirms why the HFCS industry has been supporting the sugar program as a member of ASA; however, an opportunity for the HFCS industry to increase marketing overseas such as in Mexico may weaken the incentive to support the sugar program: if the sugar price drops as a result of Mexican sugar flowing into the U.S. market through the large import quota promised under NAFTA, the industry has to weigh both costs and benefits to the industry (Moss and Schmitz, 2002). Although the economic implications of the price differential between sugar and HFCS are identified, estimation of a quantitative relationship for HFCS supply is not

PAGE 34

23 straightforward. According to the analysis by Moss and Schmitz (2002), the relationship between sugar and HFCS prices has changed over time: HFCS price responded to wholesale sugar price from 1983 to 1996, but not from 1997 to 2001. Another analysis by Evans and Davis (2002) indicates that the estimated cross price elasticity of HFCS with respect to sugar demand was found to be insignificant. This implies that the HFCS price is set below the sugar price in order to attract bulk sweetener users but its behavior remains ambiguous and arbitrary. Furthermore, other research reported that HFCS price is not correlated with the price of com (Offenbach, 1995), but others found that the HFCS price responded to the price of com from 1983 to 1996 but not from 1997 to 2001 (Moss and Schmitz, 2002). Given the insignificant relationship between com price and HFCS supply mentioned above, the indirect impact of changes in com production on HFCS price would likely be small; however, any drastic changes in com program or large changes in com export may eventually affect HFCS prices. Com statistics such as recent com production for selected countries, industrial use of com in the U.S., and the com price in the U.S. are illustrated in Appendix B.

PAGE 35

24 G O u
PAGE 36

25 100 T 90 " 80 — 70 « 60 4 06 G 50 o 'S 40 30 i-i 20 10 0 1-T m-' m — mimm — — ^ — 500 1,000 1.500 2,000 2,500 Annual Rainfall [mm] 3,000 3.500 4,000 Figure 2-3. Annual Rainfall and Irrigation Rate in Sugarcane Fields Source: COAAZUCAR, 2003c Table 2-1. Cane Sugar Production in Selected Countries, 1997-2000 Average. Country Area Sugar Cane Sugar Sugarcane Harvested Production Yield Yield Recovery (1,000 ha) (1,000 MT) (MT/ha) (MT/ha) Rate (%) Brazil 4,914 18,339 68.28 3.73 5.47 India 4,092 17,233 69.41 4.21 6.07 Cuba 1,086 3,814 32.69 3.51 10.77 China 1,064 6,532 75.08 6.14 8.18 Pakistan 1,029 3,064 46.57 2.98 6.38 Thailand 923 5,468 56.16 5.92 10.55 Mexico 627 4,807 76.46 7.66 10.02 Australia 409 5,281 90.94 12.91 14.22 United States 360 3,811 88.00 10.59 12.03 Colombia 391 2,227 86.25 5.70 6.61 Philippines 324 1,796 81.06 5.55 6.87 South Africa 315 2,684 72.57 8.54 11.76 Wodd 19,307 90,340 65.12 4.68 7.19 Source: FAO 2003, USDA 2002a, USDA 2003b

PAGE 37

26 6.000.000 H g 5.000,000 O 4,000.000 u 3 3.000.000 o £ 2,000,000 u W) 1,000,000 O — o o o o o o ^ — — . — rj (S Year Figure 2-4. Mexican Sugar Production, 1988-2002 Source: COAAZUCAR 2003a Sugarcane Cane juice IRa^v sugar Iiitemational market or stored l*Refined sugar Standai'd sugar 1^ White sugar (purer) Other grades Molasses Alcohol (rum) Other by-products Domestic market t Figure 2-5. Main Use of Sugarcane Derivatives in Mexico Source: Adapted from Polopolus and Alvarez, 1991

PAGE 38

27 Table 2-2. World Sugar Consumption in 2000 National total Per capita consumption" consumption" Coiintrv/ rppion \^WUllll jl l^glWll ITSweetener Sugar " b Sweetener Sugar (1,000 MT) (1,000 MT) (kg) (kg) India 26,234 18,101 25.59 17.66 United States 21,221 9,371 74.22 32.77 European Union (15 countries) 18,960 14,370 50.26 38.10 China 11,805 11,028 9.13 8.53 USSR, Former Area of 11,771 11,407 40.57 39.32 Brazil 10,007 9,620 57.99 55.75 Mexico 5,100 4,476 50.81 44.60 Japan 3,741 2,327 29.38 18.27 Philippines 2,135 1,975 27.68 25.61 Thailand 1,937 1,924 30.47 30.26 Canada 1,297 1,128 41.83 36.37 Australia 1,060 922 54.81 47.67 Cuba 711 710 63.24 63.18 World 168,632 133,401 27.49 21.75 a: Production + net import + change in stocks b: Sugar (raw equivalent) plus other kinds of sweeteners Source: FAO, 2003 Table 2-3. Quota and Tariff Schedule Imposed on Mexican Sugar Exported to the U.S. Year U.S. Import Quota (MT) Over-Quota Tariff (raw cane, cents/pound) Mexico as a net surplus producer Mexico NOT as a net surplus producer 1994 25,000 7,258 16.00 (Base) 1995 25,000 7,258 15.20 1996 25,000 7,258 14.80 1997 25,000 7,258 14.40 1998 25,000 7,258 14.00 1999 25,000 7,258 13.60 2000 250,000 7,258 12.09 2001 250,000 7,258 10.58 2002 250,000 7,258 9.97 2003 250,000 7,258 7.56 2004 250,000 7,258 6.04 2005 250,000 7,258 4.53 2006 250,000 7,258 3.02 2007 250,000 7,258 1.51 2008 and beyond oo 0.00 Source: USD A, 1999.

PAGE 39

28 Year •Total HFCS use per capita Refined sugar use per capita Figure 2-6. U.S. Refined Sugar and HFCS Use per Capita Source: USD A, 1993 Figure 2-7. U.S. HFCS Production Source: USD A, 1993 and 2001a.

PAGE 40

29 Import (mostly from Canada) 1.6% \ MiiSjimf^ U.S. production 98.4% Figure 2-8. U.S. HFCS Supply Source: USDA, 2002a Export to ^Export to B Domestic consumption 97.5% Figure 2-9. U.S. HFCS Export Source: USDA, 2002a

PAGE 41

30 400.0 Year •U.S. total export U.S. export to Mexico Figure 2-10. U.S. HFCS Export, 1992-2001 Source: USDA, 2002a -Consumption of sugar per capita — — Consumption of sweetener per capita Figure 2-11. Consumption of Sugar and HFCS per Capita in Mexico Source: Garcia Cliaves et al., 2004

PAGE 42

31 70 T 1995 19% 1997 1998 1999 2(XX) Year -B—Worldrawa^ price -•— US rawsu^ price, city fee paid -A— US spot price for HPCS42, Mdv\est markrts -A— Uiit valiE of HKSecpoted to Nfexioo Figure 2-12. Prices of Sugar and HFCS, 1994-2000 Source: USD A, 2001a

PAGE 43

CHAPTER 3 CONCEPTUAL AND THEORETICAL FRAMEWORK In this chapter the conceptual and theoretical framework employed to analyze trade between the United States and Mexico in the sugar market is discussed. In the conceptual framework, factors that influence the market and the trade system are specified and linked. Based on the conceptual framework, theoretical foundations are established in two parts: an analysis of the sugar market for each country and an analysis of the market balance in the U.S.-Mexico bilateral sugar trade system. Conceptual Framework The sugar market, one of the oldest and most common agricultural commodity markets, is built upon sugarcane and sugar beet production and the resulting production of processed sugar from these raw materials. Being an essential commodity for a daily diet, sugar has been traded across borders for a long period of time. More recently, there have been major changes to the trading pattern due to the emergence of an alternative in the market. High fructose com syrup (HFCS), now the most widely adopted sugar substitute, expanded the sugar market into a sweetener market. This is particularly true in the case of the United States, where HFCS now occupies nearly half of the sweetener market, and is becoming the case in Mexico as a result of the two markets becoming more closely linked by the North American Trade Agreement (NAFTA). Understanding this linkage between the United States and Mexico holds important clues to analyze economic and political impacts in the sweetener markets. To illustrate this linkage, the 32

PAGE 44

33 flow of the goods in each market and factors that influence markets are illustrated in Figure 3-1. In Mexico, sugar distributed in the market is supplied entirely by domestic sugar production, which is also provided entirely by domestic sugar cane growers. Sugarcane production depends on inputs such as labor from Mexican farmers, land, and agricultural chemicals, technology to grow and harvest sugarcane, and other factors such as weather and government support. Grower's behavior is also influenced by their relationship with mills. Sugar processing depends on inputs such as labor from mill workers, harvested sugarcane, and energy such as petroleum to run the facilities, technology to produce sugar from sugarcane, infrastructure, and government support. Note that government plays an important role to support both grower's and mill's activities. HFCS is primarily supplied by domestic production and the remained is imported from the United States. Sugar is consumed by households and bulk users (such as soft-drink manufacturers), while HFCS is consumed only by bulk users. Determinants of sweetener demand are income, tastes and preferences, price and other factors such as population growth. In the United States, domestic sugar supply is derived from domestic production and supplemented by import from various origins. Unlike Mexico, domestic sugar is derived from both sugarcane and sugar beet production. Sugar mills produce sugar not only from domestically produced sugarcane and sugar beets, but also from imported raw sugar. HFCS is supplied only from domestic com sweetener manufacturers and supplies roughly half of caloric sweetener consumption in the United States. HFCS is also exported, and Mexico is one of the main destinations.

PAGE 45

34 The U.S. -Mexico bilateral trade system involves sugar exported from Mexico to the United States and HFCS exported from the United States to Mexico. Mexican sugar is exported to a variety of markets, but the primary destination is the U.S. sweetener market. Once Mexico exports the amount of sugar to the United States allowed under the U.S. import system, excess sugar is exported to the world sugar market. In total, more than 30 countries exported sugar to the United States under the allocated tariff-rate quota in 2002 (USDA, 2002a). Under the conditions of NAFTA, Mexico could be allocated 250,000 MT of the U.S. import quota given certain conditions (successful attainment of net surplus sweetener producer status), which would increase Mexico's share of the total U.S. import sugar quota allocation to 20 percent. A two-country trade model with quota imposed by a large country importer is illustrated in Figure 3-2 (a). By imposing a quota on Mexican excess supply of sugar (ESmx), the quantity exported to the United States is limited to Qq instead of Qf. Consequently, the price for imported sugar in the United Staes increases to Pq, us and the price for exported sugar from Mexico decreases to Pq, mx. This causes welfare loss in Mexico (area abed) due to lower sugar export price and generates quota revenue in the United States (area efgh) collected by the U.S. quota holders or the government. Beginning in 2008 Mexico will have free access to the U.S. market. The same trade model without the quota system is illustrated in Figure 3-2 (b). Trade without distortion brings an increase in welfare in both countries. For simplicity, the producer price support policy is excluded from both Figures (a) and (b).

PAGE 46

35 Theoretical Framework Sweetener Market Analysis The sweetener market in a country that consumes both sugar and HFCS can be expressed as follows: SUGAR =fl (PsUGAR„Zl) (3-1) HFCS =f2 (Phfcs„Z2) (3-2) &SUGAR = h,(PsUGAR, Wj) (3.3) (/hfcs = h2 (Phfcs,. W2) (3.4) SUGAR = SUGAR (3-5) (2^ HFCS = (^HFCS (3-6) where is aggregate quantity demanded, Q' is aggregate quantity supplied, F is a price, Z and W are vectors of other factors that influence aggregate demand and supply of sugar or HFCS, respectively. Equations [3.5] and [3.6] depict market clearing conditions for each commodity. Sweetener demand Sweetener demand is defined based on consumer demand theory derived from utility maximizing behavior. Following Varian's demonstration (1992), aggregate sugar demand is derived from maximizing utility of aggregated consumers, including industry sugar users, by purchasing sugar: max u ( QsuGAR , Qx) s.t. QsuGAR *PsuGAR + Qx*Px = m (3.7) where Q is the quantity consumed, X represents all other goods consumed, P is a price, and m is national income. By solving maximization problem, the aggregate demand function is expressed as:

PAGE 47

36 QSUGAR (PsUGAR , Px, m). (3.8) Since the demand function is homogeneous of degree zero, it can be normahzed by either price. Normalize by Px ' and thus the aggregate demand for sugar is expressed as a function of real price of sugar and real income. By treating real income as one of demand shifters, aggregate demand is expressed as shown in equation [3.1]. Similarly the aggregate demand for HFCS can be expressed as a function of HFCS price and a vector that affects aggregate HFCS demand. Note that in reality, HFCS is consumed directly by industry users and its price is observed by them; households as final consumers consume HFCS indirectly through HFCS-contained goods. Sweetener supply Sweetener supply is defined based on firm supply theory derived from profit maximizing behavior. Although in reality sugar supply consists of two steps of production in reality, sugarcane and sugar production, a simple aggregate sugar supply equation at industry level is derived rather than two equations. This is suitable for two reasons: a single industry supply equation makes simulations in the bilateral trade model in the following procedure simple, and it is the industry supply price that the government is interested in supporting. Suppose the industry faces a cost function given by: where Visa set of other factors including input prices. The industry maximizes profits, assuming the market is competitive (the industry as price taker): QsuGAR (PsuGAR / Px, / Px), (3.9) C=C{QsUGAR,V) (3.10)

PAGE 48

37 max n = Psugar * Qsugar C (Qsugar., V ). (3.11) The first-order condition gives the supply relationship for the industry by equating output price to marginal cost. Expressed in a general form with a vector of other factors that influence aggregate supply: ffsuGAR (Psugar. Wi). (3.12) Similarly HFCS supply from industry {(^hfcs ) is theoretically expressed as a function of the HFCS price and the supply shifters; however, the quantitative relationship is expected to be insignificant: HFCS pricing is ambiguous and arbitrary as discussed in the previous chapter and thus the estimated relation does not likely represent the associated marginal cost curve. U.S.-Mexico Bilateral Sugar Trade System A spatial equilibrium model is used to portray the U.S.-Mexico bilateral sugar trade system, following the Takayama and Judge formulation (1964). The model provides the optimal equilibrium price as well as quantities demanded and supplied at the equilibrium through maximization of welfare in each region, i.e. the sum of the consumer and producer surplus, given the demand and supply equations and the transportation cost among regions. Let 5, (K,) represent the inverse supply function (price-dependent form) for sugar in region i; Y\ represents the quantity produced in region i; Dj {Qj) represents the inverse demand function (price-dependent form) for sugar in region j; and Qj represents the quantity consumed in region j. MaxXj/),(e,)^e, -Y^js,(Y,)dZ, -YX^R.X,^ -ZE7T/?,XX, (3.13) i=\ 1=1 /=i i=i 1=1 >=i s.t. X(^, +^,)^i',>Vj = l,..,/ (3.14)

PAGE 49

38 2;(X,, + XX,,)>e,,V; = l,...,7 (3.15) 1=1 X„ 0 where Xy is the quantity of sugar shipped from supply region i to demand region j under quota Hmit, TRij is per unit transfer costs associated with X,;,, XXy is the quantity of sugar shipped region i to demand region j over quota limit, TTRij is per unit transfer costs associated with Xy. TRij is a compound transfer cost that includes per unit transportation cost (Tij) and per-unit tariffs imposed on exported sugar under quota limit [Tanj ). Similarly, TTRij is a compound transfer cost that includes per unit transportation cost (Ty) and per-unit tariffs imposed on exported sugar over the quota limit (over-quota tariff, OQTar.j ). TR,j = T.j + Tarij (3.17) TTR,j = T,j + OQTanj (3.18) Let specific inverse linear demand and supply functions for United States and Mexico be defined as follows: P^us = lU, + lU2*(fvs + Shifter'' us (3.19) I^us = IUU, + IUU2*Q^us+ Shifter^ us (3.20) P'^MX = Ml + /M2*(2%x + Shifter'' MX (3.21) P^Mx = IMMi + IMM2 *Q^Mx + Shifter^ MX (3 .22) where P°,and Q'' represent the price and the quantity demanded in each country, respectively; F^ and Q'^ represent the price and the quantity supplied in each country, respectively; Shifter'' and Shifter^ represent the demand and supply shifters in each

PAGE 50

39 country, respectively; Wi, lUUi, IMi, and IMMi represent the intercepts in each inverse Hnear equation; and IU2, IUU2, IM2, and IMM2 represent coefficients associated with each quantity variable. The objective function of the U.S.-Mexico bilateral sugar trade system is expressed with equations [3.19], [3.20], [3.21], and [3.22] with constraints such as necessary conditions, i.e. demand-incoming shipment and supply-outgoing shipment balance as well as those specific to the bilateral trade model such as Mexico's quota allocation under NAFTA and U.S. price support: Max j {lUi + IU2*Q^ us + Shifter^ us) us + j iIM, + IM2*Q^^ MX + Shifter^ Mx)dQ^^ MX j (lUUi + IUU2*Q^us + Shifter^ us) d^us J (IMM, + IMM2*Q^MX + Shifter^ Mx) d^MX (Tmx. us )*X MX. us (Tus. Mx)*X us, MX (Tmx us + OQTarMx us)*XXmx, us (Trow, us Prow )*Xrow, us + Prow *Xmx row (3.23) where US, MX, ROW represent Mexico, the United States, and the rest of the world, respectively; Ty is per unit transportation cost from i to j; OQTarMx. us is per unit per unit over-quota tariff imposed on Mexican sugar shipped to the United States; and Prow is world price of sugar. Note that the transportation cost within the country is assumed to be zero. The over-quota import from the rest of the world to the United States {XXrow. us) is ignored because of the high tariff rate imposed on the sugar from the rest of the world.

PAGE 51

40 compared to sugar from Mexico: the first sugar that enters into the United States overquota must be from Mexico. The transfer cost of sugar from the rest of the world to the United States includes the price of sugar, i.e. sugar from Mexico and the rest of the world compete with each other to enter the U.S. market: the one with lower transfer cost enters the market first. The last term in the equation [3.23] considers Mexico's sales of sugar to the rest of the world. Equation [3.23] is rewritten in quadratic form as follows: Max iV2)*IU2*(Q^usf +iIUi + Shifter'' us )*Q'' us + (l/2)*/M2*fQ%x)' + (/M/ + Shifter'' mx)*Q''' MX (1/2)* IUU2*( us f + UUU,+ Shifter^ us)*Q^ us (1/2)* IMM2*(Q'mx )' + (IMM, + Shifter^ mx)*Q' MX (Tmx, us )*X MX, us (TuS. MX )*X us, MX (Tmx, us + OQTarMx, us)*XXmx, us (Trow, us + Prow )*Xrow, us + Prow *Xmx, ROW (3-24) Constraints are defined to balance the quantities shipped with the quantities demanded as well as supplied; to impose a quota on imported sugar from Mexico; and to impose a quota on imported sugar from the rest of the world, which is equivalent to the U.S. minimum sugar import quota required under WTO agreements: Q us -Xus, us Xmx. us XX MX, us Xrow, t/s ^ 0 (3.25) Q^^mx -Xmx.mx X us, MX ^0 (3.26) Q^us + Xus,us+ X us, MX ^ 0 (3.27) Q^MX + Xmx, MX + Xmx, us + XX MX, us + Xmx. row ^ 0 (3.28)

PAGE 52

41 X MX. usQuota < 0 (3.29) Xrow. us + XX MX. us USMin > 0 (3.30) where Quotais the quota Mexico is assigned according to the NAFTA provisions and USMin is the quota allocated to the rest of the world (the U.S. minimum sugar import quota less the quota assigned to Mexico). Equation [3.25] and [3.26] satisfy domestic demand; equation [3.27] and [3.28] regulate the quantities shipped from supply regions; and equations [3.29] and [3.30] define quotas imposed on Mexican sugar and from the rest of the world. The implications of the model can be examined by writing the first-order conditions obtained through using Kuhn-Tucker theorem. From equations from [3.24] to [3.30] the Lagrangian form (L) is given as: L= i\/2)*IU2*(Q''usf +(IUi + Shifter^ us. )*Q'' us + (l/2)*/M2*(!2''W + (/M/ + Shifter'' MX )*Q''' MX (1/2)* IUU2*(Q^usf + iIUU,+ Shifter^ vs)*^ us (1/2)* IMM2*(Q'Mxf + (IMMj + Shifter' MX )*Q' MX (Tmx, us )*X MX. us (Tus.Mx)*X US.MX {Tmx, us + OQTarMx, us)*XXmx. us (Trow, us + P row )*Xrow, us + Prow *Xmx, row + Vus (X us, us + Xmx, us + XX MX, us + Xrow, us us ) + Vmx(XmX,MX + X USMX Q^^Mx) + Wus (Q^us Xus.us -X us, Mx)

PAGE 53

42 + Wa/x (Q^mx Xmx. MX Xmx. us XX MX, us Xmx. row ) + A (Quota -X MX, us) + a(XRow, us + XX MX us USMin) (3.31) where V, W, A, and crare Kuhn-Tucker multiplier associated with each constraint representing the imputed marginal value of price of sugar demanded, supplied, that of Mexican sugar exported under-quota and that of over-quota, respectively. Note that A is positive in sign and cris negative, reflecting the way their associated constraints are defined. Kuhn-Tucker conditions are expressed as follows: % =1^2*0^ us +IU,+Shifier''us Vus < 0. ^ *Q''us = 0, Q^'vs > 0 (3.32) -(IUU2*^us +IUU,+Shifter^us )+Wus < 0, *Q^us = 0, Q'us > 0 ^Qus ^Qus (3.33) ^ =IM2*Q'''mx +IMi+Shijief^Mx -Vmx < 0. *Q'''mx = 0, G%x > 0 (3.34) -{IMM2*Q'mx +IMMi+Shifter^Mx)+WMx < 0 . *Q'mx = 0, Q'mx > 0 oQmx ^Qmx (3.35) = Vus Wus < 0, , *X us, us =0, X us, us>0 (3.36) ^X[js,us ^Xus,us =Vmx -Wmx < 0, *Xmx MX =0, Xmx. mx>0 (3.37) ^

PAGE 54

43 = -Tmx, us + Vus -Wmx /I ^ 0, *Xmx, us =0, Xmx, as ^ 0 = -Tus. MX + ^us -Wmx ^ 0, — — *X us, mx -0, X vsmx ^ 0 (3.38) (3.39) = -Tmx. us OQTarMx, us +Vus -Wmx + cr< 0, —~ — *Xmx, us =0, ^XX dX Xmx. us>0 (3.40) = -Trow, us -Prow + Vus +<7< 0, — — *Xrow. us -0, Xrow. us^O ROW, us ROW, us = Prow Wmx ^ 0, *Xmx, row =0 , Xmx, row ^ 0 (3.41) MX ,ROW MX MOW (3.42) By solving equations above, the following conditions must hold at equilibrium:' Vus = P^ us = Wus = P^ us (3.43) Vmx = P^MX = Wmx = P" mx (3.44) P^us
PAGE 55

44 The complementary slackness conditions indicate that if Mexico exports sugar to the U.S. under-quota {Xm, us > 0), then the demand price in the United States should not exceed the value of Mexican exporting sugar, which is equivalent to the sum of the supply price, transportation cost and the marginal value of exporting sugar under-quota (equation [3.45]). By the same token, if Mexico exports sugar to the United States overquota (XX MX, us>0), then the demand price in the United States should not exceed the sum of the Mexican supply price, transportation cost, tariff imposed on over-quota sugar and the marginal value of exporting sugar under-quota (equation [3.47]). The inequality of the prices expressed in equation [3.49] accords with reality. If both over-quota export from Mexico {XX mx. us ) and import from the rest of the world (Xmx. us ) are greater than zero, the following must also hold from equations [3.47] and [3.48]. P^m Prow = tRow. us Imx. us OQTarMx. us (3-50) This equation implies that when the price difference between Mexico and the rest of the world is equal to the difference in transfer cost (transportation cost and tariff), both over-quota export from Mexico and import from the rest of the world occur at the same time. In other words, Mexico would export over-quota only if the transportation cost from the rest of the world is high enough to justify Mexico to do so.

PAGE 57

46 Figure 3-2. Two-country Trade Model (a) with Quota System and (b) without Quota System.

PAGE 58

CHAPTER 4 EMPIRICAL MODELS AND DATA SOURCE Based upon theoretical framework developed in chapter 3, this chapter focuses on the empirical procedures and data set used in the analysis of the U.S.-Mexican sweetener market. The empirical model consists of three components: (1) demand and supply analysis models for both the U.S. and Mexican sugar markets; (2) a bilateral sugar trade analysis model; and (3) game theory analysis. These models are ordered sequentially so as to utilize the results from the former. After the trade model is introduced, assumptions associated with the model, the methods of model calibration, and simulated scenarios are presented. Results from simulations on the trade model are aggregated to assess policy recommendations. Lastly, an overview of the data set used in the empirical models is provided. Empirical Models U.S. Sweetener Demand Model Regression analysis is used to estimate sweetener demand utilizing quarterly timeseries data. The model is specified as a double-log (natural log) form that allows interpretation of estimated coefficients as elasticities associated with each variable. Based on the basic form of demand equation shown in equation [3.1], the U.S. sugar demand equation is specified as: SUGAR, t) = U]+ U2*In(I^ SUGAR. ,) + V3*ln(GDP,} + V4*In(P0P,} + U5*QTR1+ U6*QTR2+ U7*QTR3+ U8*DHFCS + e, (4.1) where (2^ is quantity of sugar demanded in each quarter in year t [1000 short tons], is 47

PAGE 59

48 real retail price of refined sugar (deflated by Consumer Price Index (CPI)) [cents / pound], GDP is real per capita GDP (deflated by CPI [US$]), POP is population, QTRs are dummy variables representing quarters of the year, DHFCS is dummy variable for availability of high fructose com syrup (HFCS) in the sweetener market {DHFCS=\ after t=1975), and e,is an error term. In theory, the price of HFCS would be preferred as prices of substitutes are expected to influence quantity demanded, however, limited availability of the data made this impossible, therefore, the dummy was used. It is expected that the elasticities associated with price {Ui), the first quarter {U5) (compared to the omitted fourth quarter), and HFCS availability {Us) to be negative and the elasticities of the remaining variables to be positive. Price elasticity is expected to be inelastic based on previous research. Previously reported own-price elasticities for sugar are -0.141 by Lopez (Lopez, 1990) and -0.73 by Petrolia and Kennedy (Petrolia and Kennedy, 2002). In the latter estimation, the U.S. wholesale refined beet sugar price reported at the Midwest Markets was used. In the process of regressions, serial correlation is anficipated and corrected by the Yule-Walker Method with appropriate lags assigned. Mexican Sweetener Demand Model Estimation of Mexican sweetener demand was conducted in a similar manner to that of the U.S. demand. The main difference was that the demand for two kinds of sugar is estimated separately in the Mexican model. Traditionally the demand for sugar in Mexico has been estimated by regressing the total consumption on sugar price and percapita income. Borrell (1991) estimated the price elasticity be -0.004 and the income elasticity at 0.5. This model was appropriate in early 1990s because there was no HFCS consumption in Mexico. In order to account for the entry of HFCS into the Mexican sweetener market, the demand for sweetener is estimated by disaggregating sugar

PAGE 60

49 consumption into: sugar consumed by households, referred to as "direct consumption"; and sugar consumed by bulk sugar users, referred to as "indirect consumption". Only indirect consumption of sugar is expected to be influenced by HFCS consumption as they are the only consumers of HFCS. In addition to estimating sugar demand, it would have been preferable to directly estimate HFCS consumption, but this was precluded because of the relatively short time series available. Instead, total consumption of sweeteners (the sum of sugar and HFCS consumption) is analyzed by regressing on sugar price and other variables. A dummy variable that represents the availability of HFCS in the market is also added to the regression forms of indirect consumption of sugar and total consumption of sweeteners where substitution between sugar and HFCS occurs in the market. In(Q^suGAR. ,) = M, + M2*ln(P^ SUGAR. ,} + M,*In(GDP,) + M4*In(P0P,)+ u, (4.2) SUGAR, t) = M5 + M6*In(P" SUGAR, ,) + M7*ln(GDP,) + M8*In(P0P,) + U9*DHFCS + v, (4.3) In(Q^^ SUGAR, t) = Mio + Mu*ln(P^ SUGAR, t) + Mj2*In(GDP,) + M]3*In(P0P,) + Ui4*DHFCS + w, (4.4) where (2 is direct consumption of sugar [MT], (2 is indirect consumption of sugar [MT], is total consumption of sweeteners [MT], is real retail price of standard sugar (deflated by CPI [pesos / Kg]), GDP is real per capita GDP (deflated by CPI [pesos]), POP is population, DHFCS is dummy variable for availability of high fructose com syrup (HFCS) in the sweetener market {DHFCS=l after t=1992), and i<,,v,,w,are error terms. It is expected that the elasticities associated with price (M2, Me, and M77 ) and

PAGE 61

50 HFCS availability (M9 and Mu) to be negative and the others to be positive. Price elasticities were expected to be inelastic. Serial correlation is anticipated and corrected by the Yule-Walker Method with appropriate lags assigned. U.S. Sweetener Supply Model Regression analysis is used to estimate the U.S. sweetener supply utilizing timeseries data. Aggregate sugar supply is analyzed in three parts, i.e. total sugar supply, sugar supply from sugarcane and sugar supply from sugar beets. Sugarcane and sugar beet production are estimated separately as they have different production regions and a different refining process. Sugar production is used as the dependent variable. It is assumed that domestic sugar production is the primary source of sugar supply and carried-over stock from the previous periods is considered constant over the estimated time span. The model is specified as a double-log (natural) form and estimated coefficients can be read as elasticities associated with each variable. lni(^ TOTAL SUGAR. ,) = UU l + UU2nn(P\oTAL SUGAR, + UU3*In(C0ST,) + UU4*In(RCVT0TAL SUGAR, t) + UU 5*In(Q^ TOTAL SUGAR, t-l) + UU6*TREND , + eet (4.5) In(Q^ CANE SUGAR.,) = UU7+ UU8*ln(P ''cANE SUGAR, + UU9*ln(C0ST,) + UUio*In(RCVcANE SUGAR.,) + UUn*In(Q^ CANE SUGAR., -1) + UU,2*TREND , + UU, (4.6) In(Q BEET SUGAR, t) = UUu + UUj4*In(P 'beet sugar, ,-i) + UUj5*In(COST0 + UUi6*In(RCV beet SUGAR, ,) + UUi7*In(Q \eetsugar. ,-i) + UUi8*TREND , + vv, (4.7) where Q represents the total sugar quantity produced, sugar quantity produced from sugarcane, and sugar quantity produced from sugar beets [1000 short tons] in the three

PAGE 62

51 equations, respectively; P represents the real retail refined sugar price in the previous year (deflated by CPI), real raw sugar price at NY sugar exchange in the previous year (deflated by CPI), and real wholesale refined beet sugar price in the previous year (deflated by CPI [cents / pound]), respectively; COSTh real total farm production expenses deflated by CPI (used as a proxy of sugar production cost [US$]); RCVs are sugar recovery rates during sugar refining process [%]; TREND is a trend variable that represents technology advancement; and ee, , uu,, vv, are error terms. Prices in the previous year are used assuming that decision-making on sugar production relies on sugar crop production, recognizing growers decide their production plan with the price realized in the previous year. Retail refined sugar price is used for total sugar production estimation due to lack of wholesale price data. Total sugar recovery rate is expressed as average of recovery rate of cane sugar and beet sugar computed by weighing each production onto each recovery rate. Autoregressive term (production in the previous year) is added in the regression to calculate long-run elasticities. Elasticities associated with price production cost {UU3, UUg, and UU15) are expected to be negative and the others to be positive. Price elasticities are expected to be inelastic. Previously reported own-price elasticities for the U.S. sugar supply at industry level are 0.14 for cane sugar and 0.34 for beet sugar by Petrolia and Kennedy (2002), using the U.S. wholesale refined beet sugar price reported at midwest markets. Price elasticities for land allowed to sugar production estimated by Lopez (1990) were 0.103 for cane and 0.246 for beets. Both authors reported higher elactisities for beet sugar.

PAGE 63

^ i wf f • : • ,?«rit— • • '-^^ "^JPFT^T*." " 52 Mexican Sweetener Supply Model Analysis of the aggregate supply of sugar in Mexico is conducted in a similar manner as the U.S. supply analysis except that Mexico produces sugar only from sugarcane. In((f,) = MMi + MM2*In(P ^,.1) + MM3*In(C0ST,) + MM4*In(DT,) + MM5*In(SUGL0SS,) + MM6*In(DURTN,) +MM7*TREND, +ww, (4.8) where Q are total sugar quantity produced [MT], F is real wholesale standard sugar price in the previous year (deflated by CPI [pesos / Kg]), COST is real average production cost per ton of sugar realized at sugar mill deflated by CPI [pesos], DT is average mill operation downtime ratio observed at sugar mill [%], SUGLOSS is average sugar loss ratio observed during sugar production process at mill [%], DURTN is average duration of the harvest in each season [days], TREND is a trend variable that represents technology advancement; and ww, is error term. Price in the previous year is used assuming that decision-making on sugar production relies on sugar crop production, recognizing growers decide their production plan with the price realized in the previous year. An autoregressive term (production in the previous year) is not added because of limited length of data available for regression. It is expected that the elasticities associated with price production cost (MM.?), downtime (MM4), and sugar loss (MM5) to be negative and the others to be positive. Price elasticities are expected to be inelastic. Previously reported ownprice elasticities for Mexican sugar supply at industry level is 0.67 by Petrolia and Kennedy (2002). Borrel (1991) estimated two sugarcane price elasticities by regressing sugarcane yield on sugar cane price in an Almon polynomial distributed lag model and by regressing sugarcane acreage on sugarcane price. From the

PAGE 64

53 former regression, statistically significant estimates were 0.0286 (lag=3) and 0.0294 (lag=4) and from the latter, the estimate was not statistically significant. U.S.-Mexico Bilateral Sugar Trade Model Based on the theoretical framework presented in chapter 3, a bilateral trade model of U.S.-Mexico sugar trade is developed and prepared for the empirical study. First, linear inverse demand and supply equations (price endogenous) are formulated using estimated elasticities from the previous analysis following the procedures by Spreen et al. (2000) in order to formulate the objective function in the trade model. Detailed derivation is noted in Appendix C. Although the double-log form is used in the demand and supply analysis because of its statistical advantages in estimation, the linear form of demand and supply equations are used in the trade analysis model for the following reasons: with the linear form, welfare can be measured, whereas welfare cannot be measured using log form equations due to the nature of the mathematical attributes; using the linear form makes it possible for the demand or supply curves to shift with their slopes held constant; and interpretation of Kuhn-Tucker conditions presented in chapter 3 is facilitated. Sugar trade models have been developed by Koo and Taylor (2000) and Petrolia and Kennedy (2000). The former model incorporates sugar production, consumption and stock changes in seventeen sugar producing and consuming countries and the latter encompasses the United States, Mexico, and Cuba. In these models, market equilibrium is solved based on the market clearing condition, i.e. the sum of all countries' excess demand for sugar becomes zero as sugar price adjusts. These models are relatively simple and may be suitable for macroscopic analysis since they are able to include many economies without making the models large and complicated; however, the solutions derived from these models pay little attention to country's welfare or transfer costs of

PAGE 65

54 goods. In addition, details of trade agreements are difficult to incorporate in these models. The fact that there exist numerous bilateral and regional trade agreements, a spatial equilibrium model has the advantage of enabling the various trade agreements to be tailored into the model. Lastly but most importantly, it is critical to consider HFCS when the sweetener markets in both the United States and Mexico are under scrutiny. Since this study is focused on the bilateral relation between the United States and Mexico, a spatial equilibrium is chosen to conduct a more microscopic analysis, incorporating detailed trade agreements and the impacts from HFCS consumption. Simplifying assumptions Several simplifying assumptions are made to run the model. First, the quantities in the models are raw sugar equivalent. In doing so, the price difference along the vertical market channel is ignored and derived demand and supply curves are assumed to posses the same slopes as demand and supply of refined sugar. Second, changes in sugar stocks in both counties are also ignored and hence the excess sugar supply from Mexico and the excess sugar demand from the United States are captured as the difference between domestic sugar demanded and supplied in each county, illustrated in the following. Formally, the U.S. sweetener market balance is expressed with sugar production, consumption, import, export, and change in stock as well as with those of HFCS: Q"" SUGAR. ,+ SE,+ ST.^i + HFCS'' , =Q' sugar. t+ SI ,+ ST,.i + HFCS' , (4.9) where Q'^ sugar is the quantity of sugar demanded, sugar is the quantity of sugar supplied from domestic sugar production, SE is sugar export, SI is sugar import, ST,+i is sugar stock carried over to the next year, ST,.i is sugar stock carried over from the previous year, and HFCS'^ and HFCS^ are HFCS demand and supply related transactions,

PAGE 66

55 respectively. Q'^ sugar. I '^s expressed in annual base converted from quarterly base. Since neither supply nor demand of HFCS is estimated, HFCS^ and HFCS^ are treated constant. This is justified by the fact that HFCS use in the United States has remained stable in recent years, accounting for roughly half of total U.S. caloric sweetener use. In addition, changes in stock have also remained stable given the nature of commodity demand and production practices, ST,.] and 57,+/ are set zero. SE is also set zero as sugar export from the United States is negligible given the production capacity. Equation [4.9] is then simplified and the quantity of forecasted sugar import in the United States ( 57, ) is D expressed with forecasted quantities of sugar demanded and supplied ( Q sugar, t andQ^ SUGAR, t) as: si, = Q SUGAR. , Q 'sugar, , (4.10) Mexico's sweetener market balance is expressed in a similar fashion. The differences are sugar demand in Mexico is estimated by total, direct and indirect sugar consumption and Mexico imports HFCS from the United States to meet its domestic demand for sweeteners. By ignoring stock changes and sugar import, the sweetener market balance is expressed as: Q'^'^sugar. , + Q"^sugar. t + MFCS'", + SE, = Q^sugar, , + HFCS^ , (4.11) where Q sugar is the quantity of demanded by households (direct consumption), (^^ sugar is the quantity of sugar demanded by bulk users (indirect consumption), HFCS^ is the quantity of HFCS demanded by bulk users, SE is sugar export, sugar is the quantity of sugar supplied from domestic sugar production, and HFCS^ is the quantity of HFCS supplied from domestic production and import from the United States. The study is interested in forecasting sugar surplus available for exporting to the United States

PAGE 67

56 D S considering Mexico's increasing demand for HFCS. Since MFCS'' and MFCS' are not estimated directly, HFCS demand {MFCS^ ) is forecasted based on the associated scenarios with details provided in the following section.' Once HFCS demand is forecasted, the quantity of excess sugar destined to export is calculated by considering the substitution between HFCS and indirect sugar consumption, which occurs at industry (bulk users) level. Let Q'^ sugar, r be HFCS forecast-adjusted indirect sugar consumption, then the sugar export forecast ( SE, ) is expressed with forecasted quantities of sugar demanded and supplied {Q^ sugar, t, Q^^ sugar, t ,and Q'^ sugar, t) as: SE, = Q ^ sugar, I (Q ^'^ sugar, t + Q sugar, t ) (4. 12) In equation [4.34], it is also assumed that HFCS is consumed only to supplement sugar consumption and its demand is met by readily available HFCS from domestic production and import from the United States. Accordingly, NAFTA provisions are defined, depending on the Mexico's domestic sweetener balance; Mexico receives a larger quota for the following year if is attained net surplus producer status for two successive years: Q '''' SUGAR, t + Q sugar, , + MFCS'" , < Q ^ SUGAR, t , (4.13) otherwise, Mexico receives a smaller quota until 2008 when access to the U.S. market is unlimited. Model calibration The model is calibrated by positioning the intercepts of inverse linear demand and supply equations and by adjusting average transportation cost from the rest of the worid ' Alternatively HFCS demand can be forecasted balancing the total consumption of sweetener equation with direct and indirect consumption sugar equations, yet it resulted in a poor forecast.

PAGE 68

57 to the United States in order to iterate for solution from 2002 through 2015 using the mathematical programming software package GAMS. Intercepts of inverse linear demand and supply equations in quantity-price space {lUi, lUUi, IMi, and /MM; in equations [4.14] through [4.17], respectively) are calibrated with the actual values realized in 2001 (base year) from equations [C.7b], [C8b], [C9b] and [C. 10b] presented in Appendix C: lUi = us. 2001 (IU2*Q^us. 2001 + Shifter^ us. 2001 ) (4. 14) lUU] = P'us. 2001 (IUU2*Q^us.200i+ Shifter^ us. 2001 ) (4.15) IM J =P" MX. 2001 -(IM2*Q^^ MX. 2001 + Shifter'' MX. 200j) (4.16) IMM, = P^MX. 2001 (IMM2*Q^MX. 2001 + Shifter^ MX. 2001)(4. 17) By calibrating intercepts, both excess supply in Mexico (quantity presented as "A" on ESmx curve in Figure 4-1) and excess demand in the United States (quantity presented as "B" on EDus curve in Figure 4-1) are set to correspond to the actual volume realized in 2001. Slopes for both inverse demand and inverse supply curves in both countries are held constant, yet these curves shift over the forecast horizon according to the scenarios proposed in the following section. • • In order to represent Mexican sugar supply and export capacity adequately, the model is further calibrated by adjusting the average transportation cost from the rest of the world to the United States, based on the relationship expressed in equation [3.50] in chapter 3. To do so, the transportation cost is calibrated so that Mexico exports sugar over-quota at the minimum amount. This calibration procedure resulted in a rather high transportation cost from the rest of the world to the United States; however, it insinuates the irrational behavior of the Mexican sugar industry which has been suffering from

PAGE 69

58 financial stress, vividly illustrated by the mill expropriation by the government in 2001, and producing and exporting surplus sugar to the rest of the world at the same time. Simulated Scenarios Scenarios are formulated by considering alternative assumptions related to Mexico's sweetener market situations affected by continued gains in the productivity of the sugar industry, HFCS consumption, as well as a policy lever and U.S. sugar policy levers. Each scenario carries a combination of Mexico's sweetener market situation and U.S. sugar policy. To compare the impacts of scenarios, a baseline scenario is defined where status quo is maintained (Table 4-1). In the baseline scenario, it is assumed that shifts in sugar demand and supply in both countries continues at the average rates observed in recent years and that the U.S. government maintains price support and allocates quota among exporters in a flexible manner, abiding by the WTO minimum import requirement.^ Assumptions related to Mexico's sweetener market situations are summarized in Table 4-2. Four situations are proposed: higher sugar production, higher HFCS adoption, a combination of both and introduction of tax on HFCS as a Mexico's alternative policy lever. The rates of increase in production and HFCS adoption are defined relative to the baseline. Impacts of Mexico's tax on HFCS is based on the forecast by Haley and Suarez (2003) where Mexican HFCS consumption drops significantly in 2002 and 2003 due to a tax imposed on beverages that contain HFCS. The impact of this tax is assumed significant considering soft-drink manufacturers currently account for about one-third of domestic sugar consumption in Mexico (Buzzanell, 2002). In all situations, U.S. demand ^ Current allocation system is based on historical trade shares (Skully, 1998).

PAGE 70

59 and supply are held at "Baseline" level. The duration of harvest is held constant because it is partly affected by the weather. Also, the assumption where Mexico's HFCS adoption and tax on HFCS occur at the same time is not considered since the scenario becomes incompatible to simulate. The scenarios associated with "HFCS adoption" are consistent with an increasing trend since 1994 when NAFTA went into effect and speculation of expansion of HFCS production in Mexico. If Mexico continues to increase HFCS consumption, it may follow a similar path as was seen in the United States in early 1980s when soft-drink manufacturers decided to switch to HFCS from sugar. HFCS manufacturers, who are mostly the U.S. corporations, see this phenomenon as a business opportunity in Mexico. With HFCS capital-intensive facilities, existing HFCS plants in Mexico are operated by the firms based in the United States. The "HFCS Adoption" case assumes that HFCS will be adopted in a linear fashion until HFCS consumption is 50 percent of total indirect sweetener consumption in 2008 and its share remains constant for the rest of the forecast horizon. The 50 percent share of indirect consumption of sweetener is equivalent to about 27 percent of total consumption of sweeteners. In 2001 (base year), share of HFCS in indirect sweetener consumption and total sweetener consumption were 25.3 percent and 1 1.6 percent, respectively. Comparisons of HFCS and indirect sugar consumption forecast for "Baseline" scenario, "HFCS adoption", and "Tax on HFCS" situations are illustrated in Figure 4.2. Assumptions related to U.S. sugar policy levers are summarized in Table 4-3. Two kinds of policy levers are considered: one is to stabilize the demand price and the other is to allocate quotas to exporters. For the former, two possible sugar policies are used as

PAGE 71

60 alternatives to the current price support. In one policy, instead of directly supporting sugar price, the government indirectly supports a sugar price through buying up excess sugar in the market to assure that sugar price will not fall due to excess supply from overseas. In this assumption, the government buys up as much sugar as the net sugar import, i.e. total sugar import less minimum import requirement. In the bilateral trade model, the cost of buying up excess sugar is not included in the objective function, assuming that the government does not spend ex ante cost. Also for simplicity, sugar storage costs incurred by the government are ignored. The second policy option is to introduce sugar production controls in both the United States and Mexico. This option requires cooperation from the Mexican sugar industry: defection by either party will result in an unsuccessful outcome. In this way, both the United States and Mexico are assumed to control their sugar production according to forecasted demand, avoiding excess supply of sugar. In other words, the idea implies that the United States is willing to import sugar from Mexico as long as Mexico cooperates to reduce its sugar production to meet the sugar demand in both countries and also that Mexico can avoid excess surplus sugar which cannot be sold anywhere except in the world market. This sugar policy does not involve financial support from the government: the cost of the program is zero. U.S. policy levers related to quota allocations to exporters are treated with two different approaches: the U.S. government allocates import quotas in a flexible manner between Mexico and the rest of the world (status quo); and minimum quotas are maintained (the remainder of the minimum import requirement less that allocated to Mexico) to the rest of the world no matter how much Mexico exports to the United

PAGE 72

61 States. Both policies abide by the minimum import requirement under WTO; however, political feasibility is assumed to be quite different. Scenarios to simulate are formulated by combining specific Mexico's sweetener market situation and the U.S. sugar policy. A list of 16 scenarios is shown in Table 4-4. Finally, special simulations are prepared in order to further examine the effects of Mexico's production improvement and HFCS adoption on the Mexican sugar industry and welfare. In addition to the assumption regarding Mexico's production improvement (an additional 1 percent to the baseline) and HFCS adoption (share of indirect sweetener consumption by HFCS at 50 percent) summarized in Table 4-2, simulations are conducted by changing production improvement rate at additional 0.5 and 1.5 percent as well as HFCS adoption to achieve a market share of 30, 40, and 45 percent. Game Theory Analysis In order to assess policy recommendations using aggregated results from the various simulations, an analysis based upon game theory is introduced. The basis of the game used in the study is a non-zero-sum game with mixed strategies. Non-zero-sum means the sum of the pay-offs in each pair of strategies is not zero; in other words, one player's winning does not necessarily cause the other to lose. Mixed strategies means a player chooses a strategy to play with probability (Morris, 1994; Mas-Colell et al., 1995). Since the game is non-zero-sum, both cooperative and non-cooperative games are considered. While a cooperative game allows players to make binding agreements about how they will play or about sharing pay-offs, a non-cooperative game does not. In the latter case, the game is played by two parties: the U.S. and the Mexican governments who hold the strategies and make decisions on behalf of the economy as a whole. In the former case, the game goes through the process of considering the possible pay-offs to

PAGE 73

62 two coalitions formed among five parties (payees), i.e. three industries (the U.S. HFCS industry, the U.S. sugar industry and the Mexican sugar industry) and two countries; however, the final decisions are made by the governments who hold the strategies. Each country plays with multiple strategies that correspond to Mexico's market situations and the U.S. policy levers presented in the previous section (Table 4-5). In this game setting, a combination of strategies formulates a scenario. In the game, it is assumed that if the U.S. government introduces production control, Mexico always cooperates: a defection by either party is not considered. Also, the U.S. government chooses only flexible quota allocations to the rest of the world for simplicity. In the case of Mexico's strategies, HFCS adoption is treated as a strategy although it is neither a positive strategy nor controllable by the Mexican government. Pay-offs from each scenario are calculated for each payee. Pay-offs to the industries are expressed as present values of accumulated revenue between 2002 and 2015, assuming a three percent discount rate each year. The HFCS price is held constant at the average U.S. export price to Mexico realized between 1992 and 2001. Pay-offs to each country are expressed as present values of accumulated welfare, i.e. the sum of consumer and producer surplus. In doing so, changes in welfare are calculated only from the sugar market, assuming that sugar is a primal source of sweetener. Sugar cannot be substituted with HFCS for certain products due to the liquid form of HFCS. Also, sugar is preferred for certain products to HFCS due to flavor given to the final products. U.S. welfare is adjusted with tariff revenue from Mexico, the cost of the sugar program and the cost of buying up excess sugar. The cost of the sugar program is calculated by multiplying the price difference between the support price (loan rate for raw sugar, 18

PAGE 74

63 cents per pound) and the U.S. equilibrium price with the production quantity guaranteed at the loan rate. This cost is captured by comparing between scenarios with and without the price support, ceteris paribus. The cost of buying up excess sugar is calculated by multiplying the U.S. net sugar import, i.e. total sugar import less 1,256,000 MT of minimum import requirement with the U.S. equilibrium price. In the case of the policy to buy excess sugar policy, sugar storage costs incurred by the government are ignored. Note that since the values were converted in terms of U.S. dollars prior to simulations, the exchange rate realized in the base year (2001) was implicitly used for calculating payoffs. All the games proposed are solved through a two-players pay-off matrix (twodimensions) with the two governments are the decision-makers. When coalitions are formed, their individual pay-offs are pooled, assuming that the total pay-off is redistributed among them (Morris, 1994). In doing so, it is also assumed that industry revenue and each nation's welfare can be added together. After pay-offs are calculated for each party or coalition, these values are indexed as a relative gain or loss to the payoffs in baseline scenario to facilitate the decision process. Mixed-strategy games are first simplified by eliminating dominated strategies (strategies that played with zero probability) and then solved through maximizing the expected pay-off to each player from the game (Morris, 1994; Varian, 1992). Sources of Data Data for the Mexican sugar industry was obtained from the website of Comite de la Agroindustria Azucarera (COAAZUCAR, Sugar Agro-Industry Committee). The committee is in charge of monitoring sugar cane and sugar production at each mill as well as determining the cane price paid to farmers in the country. The latter task was

PAGE 75

64 taken over from the former government body after the privatization of the miUing sector. The committee carries an extensive data set regarding not only physical production and price but also detailed productivity and efficiency indicators such as sugar and fiber content in cane, mill downtime and sugar production loss during the process across 60 operating mills. Data for the U.S. sugar industry was obtained from the Sugar and Sweetener Situation and Outlook Yearbook and other publications by the Economic Research Service, USDA. Historic data for population were obtained form the website of the U.S. Bureau of the Census; and those for GDP, exchange rates and consumer price index were taken from OECD documents. Data sources used in the study are summarized in Tables 4-6 through 4-10. For demand analysis, aggregate time-series data at the national level are used in the regressions for both the United States and Mexico. Missing data were found in U.S. sugar consumption and related prices in 1991 and filled with average values of two adjacent years in order to maintain the data continuity. For the supply analysis, aggregate time-series data at industry level are used in the regression for both the United States and Mexico. No missing data were found in the data set. Nominal values of old Mexican pesos before devaluation found in the data are converted into current pesos. Local units are preserved during the estimation of elasticities in demand and supply analyses. In bilateral trade model, different local units are converted into common units such as metric tons and US dollars so that the model can achieve the equilibrium point in the system.

PAGE 76

65

PAGE 77

66 Table 4-1. Assumptions for Baseline (Status quo) Scenario Category Shifters Country Assumptions GDP United States and Will increase at the average real GDP annual growth rate realized between 1997 and 2001 (2.23 percent for the United States and 2.91 percent for Mexico). Market situations Population Mexico Will increase similarly as the forecast by the U.S. Bureau of the Census. (Annual growth rate: 0.88 percent for the United states and 1.15 percent for Mexico). Production cost Will decrease at the average annual reduction rate realized between 1997 and 2001 (1.10 percent for the United States and 1.80 percent for Mexico). Recovery rate United States Will increase at the average annual improvement rate realized between 1997 and 2001 (0.48 percent). Downtime Mexico Will decrease at the average annual reduction rate realized betweenl997 and 2001 (1.00 percent). Sugar loss Will decrease at the average annual reduction rate realized between nl997 and 2001 (2.00 percent). Duration of harvest Held constant HFCS consumption Will be consumed maintaining the same share in indirect consumption of sweetener realized in 2001 (25.3 percent) U.S. sugar Price support npi_ T T O A • i. The U.S. government maintains the pnce support (current policy) at 18 cents per pound (loan rate). policy Flexible allocation The U.S. government allocates quotas in a flexible manner between Mexico and the rest of the world.

PAGE 78

67 Table 4-2. Assumptions for Mexican Sweetener Market Situations Situations (code) Shifters Assumptions High Mexican Production (P) Production cost, Downtime, Sugar loss Will improve an additional 1 percent to the "Baseline". High MFCS Adoption (A) HFCS consumption It is assumed that HFCS consumption will increase in a linear fashion until it replaces 50 percentage or indirect consumption of sugar in 2008. After 2008, its share remains at 50 percent. High Mexican Poduction High HFCS adoption (PA) Production cost. Downtime, Sugar loss The same as "High Mexican Production" HFCS consumption The same as "HFCS Adoption" Tax on HFCS (T) HFCS consumption It is assumed that HFCS consumption for 2002 through 2004 will drop due to tax (Haley and Suarez, 2003) and remain at reduced consumption level (7 percent to the indirect consumption of sugar) for the rest of the forecast horizon.^ Other shifters are the same as the Baseline. A 20-percent tax on beverages that contain HFCS was introduced on January 1, 2002; suspended on March 5 by the president's decision; and then reimposed on July 16, 2002 with the decision by Mexico's Supreme Court of Justice (USDA, 2002a).

PAGE 79

68 6,000,000 5,000,000 H 4,000,000 g 3,000,000 e ^ 2,000,000 1,000,000 Year Indirect consumption of sugar (Baseline) — Indirect consumption of sugar (Tax on HFCS) -A— Indirect consumption of sugar (High HFCS adoption) 0 HFCS consumption (Baseline) -B— HFCS consumption (Tax on HFCS) -A — HFCS consumption (High HFCS adoption) Figure 4-2. Forecasted Indirect Sugar and HFCS Consumption in Mexico under Alternative Scenarios

PAGE 80

69 Table 4-3. Assumptions for U.S. Sugar Policies Categories Policies (code) Assumptions Stabilization of the demand price Price support (status quo) (S) rill T T A * A * A i * A The U.S. government mamtams the pnce support at 18 cents per pound (loan rate, 396.48 US$ per MT). Buying up excess sugar in the market (B) The U.S. government abandons the price support and buys up excess sugar in the market instead. Production control (C) The U.S. government abandons the price. Instead, the U.S. and Mexican governments collectively control the sugar production in such a way that the sum of quantities demanded in two countries is primarily met by the sum of the quantities supplied from two countries. Quota allocations Flexible allocation (status quo) (F) The U.S. government allocates quotas in a flexible manner between Mexico and the rest of the world. Minimum quota allocations for the rest of the world (M) The U.S. government reserves the minimum quotas (the remaining minimum import requirement less allocated to Mexico) for the rest of the world.

PAGE 81

70 Table 4-4. Listing of Examined Scenarios Scenario Mexican market U.S. sugar policies number (code) situations Stabihzation of the demand price Quota allocations 1 (baseline) Basehne (B) Price support Flexible 2 (P-S-F) High production (P) (S) allocations (F) 3 (A-S-F) HFCS adoption (A) 4 (HA-S-F) TT* 1 J Til 0 1 High prod.-HFCS adop. (PA) 5 (B-B-F) Basehne (B) Buying up excess 6 (P-B-F) T T' 1 J i High production (P) sugar in the market (B) 7 (A-B-F) HFCS adoption (A) 8 (PA-B-F) High prod.-HFCS adop. (PA) 9 (B-C-F) Baseline (B) Production control 10 (P-C-F) High production (P) (C) 11 (A-C-F) HFCS adoption (A) 12 (PA-C-F) High prod.-HFCS adop. (PA) 13 (PA-S-M) TT" 1 1 Til '^"1 1 High prod.-HFCS adop. Price support (S) Minimum quotas allocation to the 14 (^r A-D-iVij (PA) Buying up excess sugar in the market (B) rest of the world (M) 15 (PA-C-M) Production control (C) 16 (T-S-F) Tax on HFCS (T) Price support (S) Flexible allocations (F)

PAGE 82

71 Table 4-5. Strategies for the Sugar Trading Game Country Strategies United States Strategy 1 Maintains price support (status quo) Flexible quota allocations among Strategy 2 Abandons price support and buys up excess sugar in the market Mexico and the rest of the world Strategy 3 Introduces production control with Mexico Mexico Strategy 1 Maintains the current policy (status quo) Strategy 2 Higher sugar production Strategy 3 Higher HFCS adoption Strategy 4 Higher sugar production and higher HFCS adoption Strategy 5 Introduces tax on HFCS

PAGE 83

72 Table 4-6. Data Sources for U.S. Demand Data Unit Source Consumption of sugar (dependent variable) 1000 short tons Sugar Statistical Compendium by (Stock #91006, 1970-1990) and Sugar and Sweetener Situation and Outlook Yearbook (SSS-2002, 1992-2002) by Economic Research Service, U.S. Department of Agriculture Retail price of refined sugar cents / pound Gross Domestic Product (GDP) US$ Statistic database by Organization tor Economic Co-operation and Development (OECD) Population persons International Database by U.S. Bureau of the Census, U.S. Department of Commerce Data length: 1970-2002, quarter y Table 4-7. 1 Data Sources for Mexican Demand Data Unit Source Direct consumption of sugar (dependent variable) metric tons [MT] Azucar S.A. de C.V. Estadistica Azucareras (1970-1989), database by Financiera Nacional Azucarera, S.N.C de C.V. and Comite de la Agroindustria Azucarera (CO A AZUCAR) (1990-1999) Indirect consumption of sugar (dependent variable) Total consumption of sugar (dependent vanable) Retail price of standard sugar pesos / kg Gross Domestic Product (GDP) pesos Statistic database by Organization for Economic Co-operation and Development (OECD) Population persons International Database by U.S. Bureau of the Census, U.S. Department of Commerce Data length: 1970-1999

PAGE 84

73 Table 4-8. Data Sources for U.S. Supply Data Unit Source Production of sugar, total (dependent variable) 1000 short tons Sugar Statistical Compendium by (Stock #91006, 1970-1990) and Sugar and Sweetener Situation and Outlook Yearbook (SSS-2002, 1980-2002) by Economic Research Service, U.S. Department of Agriculture Production of cane sugar (dependent variable) Production of beet sugar (dependent variable) Retail price of refined sugar Wholesale price of refined cents / pound Sugar recovery rate, total percent Beet sugar recovery rate Cane sugar recovery rate Total farm production expenses million US$ Database by Economic Research Service, U.S. Department of Agriculture Data length: 1960-2002 Table 4-9. Data Sources for Mexico Supply Data Unit Source Production of sugar (dependent variable) MT Database by Comite de la Agroindustria Azucarera (COAAZUCAR) Wholesale price of standard sugar pesos / kg Cost of sugar production per ton of sugar produced pesos / ton of sugar Downtime observed at mills percent Sugar loss during the process percent Duration of the harvest days Data length: 1988-2000

PAGE 85

74 Table 4-10. Data Sources for Miscellaneous Data Source U.S. Consumer Price Index (CPI) (19821984=100) Database by the Bureau of Labor Statistics, U.S. Department of Labor Mexico Consumer Price Index (CPI) (1994=100) Database by the Banco de Mexico Exchange rate (US$Mexican pesos) Quota-tariff rates Agricultural Outlook (1999) by Economic Research Service, U.S. Department of Agriculture Sugar per unit transportation cost (bagged, seaborne freight rate) Personal communication with an exporter of sugar Loan rate for raw sugar in the U.S. Haley and Suarez (2002) Guarantee price for raw sugar in Mexico Database by Comite de la Agroindustria Azucarera (COAAZUCAR)

PAGE 86

CHAPTER 5 EMPIRICAL RESULTS AND INTERPRETATION This chapter presents and discusses the empirical results for each of three analyses: (1) demand and supply analysis for both the U.S. and Mexican sugar markets, (2) bilateral sugar trade analysis, and (3) game theory analysis. Demand and Supply Analyses Results for the U.S. demand and supply analysis are summarized in Table 5-1. In the demand equation, signs of estimates associated with each significant variable were as expected. Significant estimates at the 95 percent confidence level were associated with price, the dummy variables for quarter 1 and 3, and the dummy variable for HFCS availability. The estimated price elasticity of demand was inelastic. The significant estimate associated with the HFCS dummy variable implies that HFCS replaces sugar as a substitute in the market to some degree. In the supply equation for the United States, estimates associated with trend and production in the previous year (autoregressive term) were significant at the 95 percent confidence level for all three models, i.e. total, beet and cane sugar supply regression models. Estimates associated with sugar recovery rate were insignificant in all models. Estimates associated with price and cost were significant at the 95 confidence level for total and beet sugar supply regressions, but not for the cane sugar supply equation. Two possible reasons why cane sugar production does not respond to the refined sugar price but beet sugar production does are: (1) cane sugar has two steps in the refinery process while beet sugar has one and (2) sugarcane is a perennial crop while sugar beets is an 75

PAGE 87

76 annual crop. These results reflect these differences and imply that sugar beet production is more sensitive to price changes. Although cane sugar production is assumed to respond to raw sugar price, the coefficient of the price variable is not significant. Estimated price elasticities were both inelastic for total sugar supply and beet sugar supply as anticipated. The long-run supply price elasticities were calculated dividing the estimated price elasticities (short-run) by (1X) where X represents the estimates for production in the previous year (autoregressive term). Computation yielded all inelastic long-run price elasticities: 0.3875 and 0.6764 for total sugar and beet sugar, respectively. Results for the demand and supply analysis for Mexico are summarized in Table 52. In the demand equation, signs of significant estimates associated with each variable were consistent with a priori expectations. The only statistically significant estimate among the three price elasticities was direct consumption, and it was inelastic. The population variable accounted for most of the explanatory power of consumption in all models. Significant estimates associated with GDP in indirect sugar and total sweetener consumption indicate that consumers tend to consume more sugar through sugarcontaining products as their income increases. In the supply equation for Mexico, the signs of estimates associated with each variable corresponded with a priori expectations. The estimate associated with price was inelastic. While reduction in production cost and factory downtime indicated an increase in production, the length of sugarcane harvest duration was almost perfectly correlated to sugar output from the mills. The coefficient for the variable representing sugar loss during the process was not significantly related to sugar output implying that the degree of sugar loss was not as critical as other factors such as production cost and factory

PAGE 88

77 downtime for mills trying to improve their production efficiency. A positive and significant estimate associated with the trend variable indicates technology related to sugar production at sugar mills had been advancing during the time period covered in this study (1988-2000). V . Bilateral Sugar Trade Analysis The results from the bilateral sugar trade analysis are divided into two parts: U.S. sugar import forecasts and the comparison of pay-offs among industries and countries. Results and discussion are guided by comparing the impacts from changes in Mexican sweetener market situations and from changes in U.S. sugar policy. U.S. sugar import forecasts for the selected eight scenarios (scenario 1, 2, 3, 4, 8, 12, 13, and 16 in Table 4-4) are shown in Figures 5-1 through 5-8. Comparisons among scenarios 1, 2, 3, 4, and 16 illustrate the impacts from changes in the Mexican sweetener market situations; comparisons among scenarios 4, 8, and 12 illustrate the impacts from changes in U.S. sugar price stabilization policy; and comparison between scenarios 4 and 13 illustrates the impacts from changes in U.S. quota allocation policy. The results from the baseline scenario (Scenario 1) shows that if status quo surrounding the sugar industries in both countries is maintained over the forecast horizon, Mexico will not likely attain a net surplus sweetener producer status and hence will miss the opportunity to benefit from exporting sugar under a larger quota allocation (250,000 MT) under NAFTA (Figure 5-1). This is due to growing domestic sugar demand relative to domestic sugar production. For comparison, forecasts for Mexican sugar consumption and production in 2008 by this study are 5.4 million and 5.8 million MT, respectively and those by Koo and Taylor (2000) are 5.3 million and six million MT, respectively.

PAGE 89

78 When Mexico expands sugar production (Scenario 2), net surplus sweetener producer status will still not be attained (Figure 5-2). Yet, Mexico will generate enough surplus sugar to export over-quota (before 2008) and quota-free (after 2008), resulting in significant impacts on the U.S. market. In total, Mexican sugar will take up about onethird (Scenario 1) or more than half (Scenario 2) of the U.S. minimum import requirement at peak in 2008. The amount of export will decline in later years due to expanding domestic sugar consumption in Mexico. In Scenario 3, when Mexico adopts HFCS at a higher rate, Mexico's sugar export swells as a result of substitution between sugar and HFCS in the domestic market. This result includes direct impacts on the Mexican sweetener market as well as extended impacts on the U.S. sweetener market (Figure 5-3). Although Mexico will not attain net surplus sweetener producer status, over-quota export will reach over 1.2 million MT by 2007 and will remain over one million MT until 2014. This export quantity will take up almost the entire U.S. minimum import requirement and as a result, sugar export from the rest of the world will be marginalized. Similar results are drawn when increases in Mexico's production and HFCS adoption are combined (Scenario 4). A slightly larger scale of Mexico's exports than Scenario 3 is shown in Figure 5-4. The policy followed by the U.S. government in its allocation of its import quota has a large impact on sugar exports from both Mexico and the rest of the world. The aforementioned large-scale export of Mexican sugar is possible only if the U.S. imports the minimum amount of sugar and allocates sugar quotas in a flexible manner among exporters. This allocation method may cause friction with the other countries that export sugar to the United States since Mexican sugar has potential to take up a large portion of

PAGE 90

79 the U.S. quota. Thus, allocating such a large portion of the quota to one country may not be a feasible policy option in the United States. In Scenario 13, it is conjectured that the U.S. government maintains minimum quotas (the remainder of the minimum import requirement less allocated to Mexico) for the rest of the world no matter how much Mexico exports. As shown in Figure 5-5, Mexico's over-quota and quota-free export will be dampened because of Mexico's comparative disadvantage to the rest of the world, while the export from the rest of the world remains over 1.2 million MT over the entire forecast horizon. By contrast, a change in the U.S. government's policy on stabilization of the domestic price does not pose much impact on sugar exports to the U.S. market. With the same Mexican sweetener market situations (high production-high HFCS adoption), the U.S. policy options of price support, buying up excess sugar in the market, and production controls are compared (Figures 5-4, 5-6 and 5-7). The results indicate that all three scenarios bring about the similar trends in Mexican sugar export. Yet, U.S. production control with Mexico causes an overall increase in quota-free export after 2008. This is because the Mexican sugar price is maintained lower relative to the price support or buying up excess sugar scenarios (Figures 5-9, 5-10, and 5-11). The prices of sugar in both countries will converge in the integrated market without a U.S. price support (Figure 5-10) while the difference in prices are kept after 2008 with production control (Figure 5-11): the price difference corresponds to the transportation cost from Mexico to the United States. Mexico's tax on HFCS brings about very different results. In this scenario, Mexico is unable not only to export either under-quota, over-quota or quota-free but also

PAGE 91

80 to generate a sugar surplus after 2006 due to expanding domestic sugar consumption accelerated by higher sugar consumption by bulk users who chose sugar over HFCS. Mexico cannot produce enough sugar to meet domestic demand after 2008 and will import sugar from the United States, resulting in a higher domestic sugar price than that of the U.S. (Figure 5-12). This extreme case would occur only if the impact of tax on HFCS lingers over the forecast horizon as assumed in this study; however, the chance for Mexico to enjoy exporting sugar would be slim. In spite of fluctuating imports from Mexico and the rest of the world, U.S. domestic sugar consumption and production will remain relatively unchanged. Sugar demand and supply forecasts for both the United States and Mexico for baseline and high productionhigh HFCS adoption scenarios are shown in Figures 5-13 and 5-14, respectively. In either scenario, U.S. demand and supply are forecasted to remain approximately at nine million and 8.8 million MT over the forecast horizon, respectively. The results from the various simulations showed that the U.S. sugar price will gradually decline but will not dip below the support price level (396.48 US$ per MT) before 2008 if the United States accepts most of the imported sugar from Mexico rather than from the rest of the world by allocating quotas in a flexible manner to the exporters, no matter which U.S. demand price stabilization policy is in place (Figures 5-9, 5-10, and 5-11). This implies that the Mexican sugar price contributes to maintain a high sugar price in the integrated U.S.-Mexico sugar market; in other words, accommodating Mexican sugar can act as an alternative form of price support in the United States. On the other hand if the U.S. government maintains minimum quotas for the rest of the world no matter how much Mexico exports and abandons the price support, U.S. equilibrium price

PAGE 92

81 will fall below the support price (Figures 5-15 and 5-16). This is a result of cheaper sugar from the rest of the world flowing into the U.S. market. In reality, the U.S. sugar price will face downward pressure from importing world sugar as well as political pressure from the rest of the world, considering the likelihood that many sugar exporting countries will not easily give up their existing shares of the U.S. import quotas. Pay-offs to the industries and countries also portray interesting contrasts among scenarios. The results from three sets of selected scenarios are summarized in Tables 5-3, 5-4, and 5-5. In these tables, the present values of accumulated pay-offs are expressed in billion of dollars and those values are indexed relative to the baseline scenario inside the brackets. The impact of changes in Mexican sweetener market on pay-offs to the industries and the two nations' welfare is illustrated in Table 5-3. The listed five scenarios are based on the assumptions that the U.S. government maintains price support and allocates quotas between Mexico and the rest of the world in a flexible manner. It is clear that the U.S. HFCS industry will become better off if Mexico adopts HFCS: revenue for the U.S. HFCS industry increases by 78 percent; and that the industry will become worse off if Mexico introduces taxes on HFCS: revenue for the industry decreases by 72 percent. The U.S. sugar industry does not gain from either Mexico's increase in sugar production or HFCS adoption since either change in Mexican sweetener market generates a larger Mexican sugar surplus that is destined for the U.S. market. Tariff revenue to the U.S. government from Mexican sugar increases as Mexico increases sugar production or HFCS adoption. Large tariff revenue from high HFCS adoption scenario, which is unexpectedly larger than the high production-high HFCS adoption scenario, is due to

PAGE 93

82 forecasted larger exports in 2008 and 2009, which consequently brings a lower U.S. sugar price accompanied by a higher cost of the price support. Overall, U.S. welfare becomes worse off with changes in Mexican sweetener market for two reasons: (1) reduced producer surplus caused by a lower producer price as a result of increased Mexican export and (2) increased net costs even though increased tariff revenue is expected as mentioned above. The Mexican sugar industry gains from expanding sugar production, but the gains are dissipated when HFCS is adopted at higher rates, given the assumption that HFCS price held constant. In Mexico's HFCS tax scenario, the Mexican sugar industry gains not from exporting to the U.S. market but from domestic sales at higher prices. All the entities except for U.S. HFCS industry benefit from this policy; however, the policy lever may not be acceptable in the international trade environment. In fact, the Mexican government swung its decisions in the past: a 20-percent tax on beverages that contain HFCS was introduced on January 1, 2002; suspended on March 5 by the president's decision; and then reimposed on July 16, 2002 with the decision by Mexico's Supreme Court of Justice (USDA, 2002a). The impact of changes in the U.S. price stabilization poHcy is illustrated in Table 54. The compared three scenarios (Scenario 4, 8, and 12) are based on the assumption that Mexico increases sugar production as well as HFCS adoption and that the U.S. government allocates quotas between Mexico and the rest of the world in a flexible manner. Alternative sugar policies to the price support only bring about improvement in U.S. welfare; sugar industries in both countries and Mexico's welfare become worse off, posing a larger negative impact on the Mexican sugar industry and welfare than the sugar industry in its own country. When the U.S. government switches sugar policy from

PAGE 94

83 supporting price directly to buying up excess sugar in the marlcet, the cost of the program is expected to diminish, resulting in welfare improvement. Yet, if the U.S. government were to switch policies from the price support to buying up excess sugar, the timing to do so will be important so as to minimize the cost incurred by the government; the cost of buying up excess sugar will rise immediately after policies are switched while the cost of the price support will not because the U.S. sugar price will be maintained relatively high in the early stage of the forecast horizon. In practice, storage costs need to be considered. When the United States controls sugar production with Mexico, welfare improves by increased consumer surplus and tariff revenue combined with zero program cost incurred by the government. This scenario also demonstrates that U.S. welfare becomes better off at the expense of the sugar industries in both countries and Mexico's welfare. When the policy of buying up excess sugar and the production control policy are compared, the U.S. government could pursue the latter in light of the nation's welfare; however, the gain is very small compared to the loss bom by industries and Mexico. Both alternative policies do not satisfy pareto optimality and the overall loss outweighs the gain by the United States as a nation. The impact of changes in U.S. price stabilization policy is illustrated in Table 5-5 with the assumption that the U.S. government maintains the minimum quotas for the rest of the world no matter how much Mexico exports. When the U.S. government introduces alternative policies to price support, U.S. welfare gains but both the U.S. and Mexican sugar industries as well as Mexico's welfare lose to a larger degree. An extreme result comes from Scenario 13. In this scenario, the U.S. sugar price support becomes extremely costly if the U.S. government reserves the minimum quota for the rest of the

PAGE 95

84 world. As a result of cheap sugar from the rest of the world flowing into the U.S. market, the U.S. price will fall far below the support price. The loss bom by the Mexican sugar industry particularly stands out: nearly a half of its expected revenue disappears due to restricted access to the U.S. market. When fostering the sugar industry in its own country and Mexico as a neighboring trade partner, the U.S. government needs to accept sugar from Mexico more than from the rest of the world. This requires sensitive negotiations among exporters. Overall, the Mexican sugar industry and welfare are more prone to changes in the Mexican sweetener market and U.S. sugar policy than the U.S. sugar industry and U.S. welfare. Among tested changes, Mexico's HFCS adoption and U.S.'s quota allocation policy hold the greatest effects. Effects of Mexico's production improvement and HFCS adoption on the Mexican sugar industry and welfare are illustrated in Figures 5-17, 5-18, and 5-19. The rate of production improvement ranges from the average rate realized between 1997 and 2001 (baseline) to additional 0.5, I.O, and 1.5 percent to the average rate; the HFCS adoption ranges from 25 percent share of indirect sweetener consumption (baseline) to 30, 40, 45, and 50 percent (HFCS adoption situation). The effects on the Mexican sugar industry are shown in Figure 5-17. Expected accumulated revenue increases as sugar production improves an additional 1 and 1.5 percent above the average rate. An increase in HFCS adoption also brings about an increase in revenue until the share attained by HFCS climbs up to 40 percent. When HFCS share reaches 50 percent, revenue shrinks compared to the baseline. Increased revenue caused by HFCS adoption is due to higher domestic sugar price and decreased revenue is due to decreased quantity of sugar demanded. The effects

PAGE 96

85 on Mexico's welfare differ from those on the industry (Figure 5-18). Production improvement increases welfare, but not HFCS adoption, assuming that sugar is a primary and preferred source of sweetener. While the gains to the Mexican sugar industry from production improvement at 1.5 percent to the baseline (US$ 0.65 billion) can make up the loss to the industry itself from HFCS adoption, even at a 50 percent share (US$ 0.42 billion), gains to Mexico's welfare cannot. The net loss to the nation from HFCS adoption at 30 percent adoption share is forecasted US$ 5.12 billion (loss of US$ 5.41 billion from welfare and a gain of US$ 0.29 billion from the industry).' This loss is far greater than the sum of expected gains from production improvement at additional 1.5 percent, which is US$ 3.42 billion (gains of US$ 2.77 billion from welfare and US$ 0.65 billion from the industry). The Mexican government faces difficulties allowing faster HFCS adoption to happen in the domestic market (Figure 5-19). Game Theory Analysis Results from the game theory analysis are summarized in Tables 5-6 through 5-13. The analysis is based on the game setting played by the United States and Mexico with multiple strategies that correspond to Mexico's market situation (the Mexican government's strategy set) and the U.S. policy levers (the U.S. government's strategy set) in order to assess gainers and losers from trade. Calculated actual pay-offs to five payees (the U.S. HFCS industry, the U.S. sugar industry, Mexican sugar industry, U.S. costadjusted welfare, and Mexico's welfare) for each combination of U.S. and Mexico's strategies of the game are shown in Table 5-6. Values in the other tables are indexed relative to the baseline scenario. Three different forms of coalitions are considered: ' Assuming that industry's revenue and nation's welfare can be added together.

PAGE 97

86 coalition of countries, coalition of the sugar industries, and grand coalition that includes all five entities. Since the inclusion of the U.S. HFCS industry into a coalition is a determinant factor, both coalitions with and without the U.S. HFCS industry are also considered. Pay-offs to the U.S. HFCS industry and the government of Mexico's welfare fluctuate more than the other three payees (Table 5-6). Among the five strategies facing Mexico, introduction of tax on HFCS returns the best pay-offs to the Mexican sugar industry and welfare and thus the Mexican government would always choose this strategy. Yet, this policy lever may not be acceptable in the international trade environment. In the following games, Mexico's tax strategy is excluded. Indexed pay-offs are shown in Table 5-7. Values in corresponding cells are categorized in four ways: no change (100), gain (over 100, cells shaded in gray), slight loss (over 95 and under 100), and loss (less than 95, cells shaded with stripes). In this game, there is neither dominant strategy (a strategy that is chosen over other strategies) nor pure strategy (a single strategy chosen by the government that brings about improvement in the industry or industries as well as welfare in its own country: this strategy is played with probability of I for either U.S. or Mexican government). By inspecting the tendency of pay-offs, some interesting contrasts among payees rise to the surface without solving mathematically. In the United States, the government always prefers a production control strategy to the other two but the sugar industry is always better off with the price support strategy, no matter which strategy Mexico chooses. As Mexico increases sugar production or HFCS adoption, the U.S. sugar industry likely loses while the U.S. government and the HFCS industry never become worse off. The

PAGE 98

87 U.S. sugar industry will lobby against a production control strategy for fear that it may lose up to five percent (high production-high adoption market situation). On the contrary, the Mexican government's choice of strategy will be accepted by the industry in most cases, except for the case of where the U.S. government buys sugar in the baseline scenario. Since HFCS adoption harms both welfare and the industry, the Mexican government will continue to struggle to suppress HFCS adoption in its market. Results from the game played by two coalitions of countries (the United States and Mexico) are shown in Table 5-8. In this game, the U.S. HFCS industry is excluded from the game. As expected, the high production strategy becomes Mexico's pure strategy and thus the game is solved when the U.S. government chooses a strategy: production control policy (cells shaded with vertical stripes). This combination of strategies is coincidentally the best choice for the U.S. Coalition, but not for the Mexican coalition. The Mexican government would prefer the U.S. government to choose either the price support or buying up excess sugar strategies. Results from the game played by the sugar industry coalition and the government coalition are shown in Table 5-9. The high production strategy becomes a pure strategy for Mexico when a decision is made by the government coalition. The solution of the game is thus determined when the U.S. government chooses the strategy of buying up excess sugar. This result differs from the game played by the country coalitions mentioned above. By a government cooperating with the other government rather than with the industry in its own country, the Mexican government improves its pay-off by US $4.75 bilHon; the U.S. government loses by US $0.24 billion; and the Mexican sugar industry loses US $1.27 billion (Table 5-4). Since the loss bom by the U.S. government is

PAGE 99

88 relatively small, the choice by the Mexican government is made at the expense of its own sugar industry. On the contrary, the industry coalition would lobby for the U.S. production control strategy, assuming that the Mexican sugar industry promises to compensate the loss of US $0.01 billion (US $10 million) bom by the U.S. sugar industry with expected gain of US $1.27 billion (Table 5-6). With this contrast of the results between a country coalition (Table 5-8) and a government/industry coalition (Table 5-9), the U.S. government would likely form a coalition with its own industry while the Mexican government would try to form a coalition with the U.S. government. Results from the game played by the grand coalition that includes all players except the U.S. HFCS industry are shown in Table 5-10. The solution is the Mexican government high production strategy and the U.S. government buying excess sugar strategy. Although this solution does not satisfy pareto optimality for all four payees, the overall loss is minimal: the loss incurred by the U.S. sugar industry is US$ 0.31 billion (Table 5-6) and is theoretically compensated by the total gain of US$ 2.29 billion (US$ 0.36 billion from the U.S. government, US$ 0.13 billion from the Mexican sugar industry, and US$ 1.80 billion from Mexico's welfare. Table 5-6) This solution corresponds to the government/industry coalition seen in Table 5-9. In other words, if the U.S. government sees the benefit from pooling gains with Mexico rather than with its industry and if redistribution of gains is possible among governments and industries, no one loses from the game. Such an arrangement and agreement in practice would be expected to be difficult to reach and implement. When the U.S. HFCS industry is included in coalitions, the same solutions are reached: the combination of Mexican government high production strategy and U.S.

PAGE 100

89 government production control strategy for country coalitions (Table 5-11), Mexican government high production strategy and the U.S. government buying up excess sugar strategy for government/industry coalitions, and Mexican govemmet high production strategy and the U.S. government buying up excess sugar strategy for grand coalition. Yet, pooled gains become larger due to inclusion of gains from the U.S. HFCS industry, the pay-off matrices offer additional factors that need to be taken into the policy assessment process. Results from the game played by country coalitions are shown in Table 5-11. Although the solution is the same as the one without the U.S. HFCS industry included, pay-offs expected from Mexican government's other strategies become more attractive to the U.S. coalition. It is obvious that the U.S. coalition would receive a better pay-off if Mexico chooses the HFCS adoption strategy. This fact gives the U.S. coalition a strong incentive to influence Mexican government's choice of strategy. Intensified conflict of interests between the industry coalition and the government coalition is illustrated in Table 5-12. Without the HFCS industry in a coalition, the industry coalition prefers the U.S. production control strategy to the buying up excess sugar strategy for the sake of slight gains. When gains from the HFCS industry are pooled, the industry coalition will lobby for strategies that involve Mexico's HFCS adoption, no matter which strategy the U.S. government plays. This is possible only if the industry coalition promises to compensate the Mexican sugar industry for the loss. If redistribution of gains among industries is feasible, the Mexican sugar industry may prefer being compensated to expecting protection from the Mexican government's strategy. Ultimately, if the Mexican government chooses to overlook the impact on

PAGE 101

90 welfare and let its domestic sugar industry be taken care of by the U.S. sugar and HFCS industry, there will be little reason for the Mexican government to cooperate with the United States; however, Mexico's abandoning support to the sugar and its related industries may trigger political and social instability. Lastly, the results from the game played by the grand coalition are shown in Table 5-13. Although the solution of the game is the same as one without the U.S. HFCS industry, the pay-offs from Mexican government high production-high HFCS adoption strategy are almost as good. This result implies that adopting HFCS in the Mexican market is not necessarily unbeneficial as long as Mexico increases sugar production, under the assumption that gains to the grand coalition are redistributed among countries and industries. The challenge is how to put this into practice.

PAGE 102

91 Table 5-1. Summary of the U.S. Supply-Demand Analysis Country United States \. Dependent ^\ variables Independent variables ^\ Demand Supply Consumption of sugar Production of sugar Production of beet sugar Production of cane sugar Number of observations 132 43 43 43 Degree of freedom 124 36 36 36 Real retail price of refined sugar -0.2323 (-4.23) ** Real per capita GDP 1) 0.1378 (0.29) Population -0.6470 (-0.71) Dummy variable (Quarter=l) -0.0707 (-5.26) Dummy variable (Quarter=2) 0.0136 (1.18) Dummy variable (Quarterns) 0.0941 (10.14)** Dummy variable for availability of HFCS -0.1853 (-3.89) ** Trend 0.0067 (2.94) ** 0.0069 (3.20) ** 0.0091 (3.08) ** Real retail price of refined sugar in the 1 ) previous year 0.1471 (2.35) ** 0.2195 (2.19)** 0.0711 (1.44) Real total farm |)roduction expenses -0.2316 (-2.19)** -0.3986 (-2.74) ** -0.0830 (-1.34) Sugar recovery rate -0.2351 (-0.57) -0.5829 (-1.55) 0.1863 (0.58) Production in the previous year 0.6204 (5.26) ** 0.6755 (5.37) ** 0.3672 (3.24) ** Constant 19.8376 (1.52) 5.9033 (2.65) ** 7.8421 (3.22) ** 5.0365 (6.12) ** Total R' 0.8645 0.8950 0.7253 0.9489 Durbin-Watson '^^ 2.0563 (1) 2.5487 (2) 2.0495 (1) 2.0325 (2) 2.0968 (1) 2.0525 (2) 2.0399 (1) 1.9725 (2) and **: Significant at 90% and 95% confident level, respectively. 1) Deflated by CPl. 2) Values are after corrected by Yule-Walker method. ( ) corresponds to the order of lag assigned.

PAGE 103

92 Table 5-2. Summary of the Mexican Supply-Demand Analysis Country Mexico Dependent variables Independem\^ variables Demand Supply Direct consumption of sugar Indirect consumption of sugar Total consumption of sweeteners Production of sugar Number of observations 30 30 30 13 Degree of freedom 26 25 25 6 Real retail pnce or standard sugar '* -0.0734 (-1.93)* -0.003535 (-0.07) -0.0215 (-0.98) Real per capita GDP 1) -0.2595 (-1.37) 0.9213 (2.96) ** 0.4363 (3.33) ** Population 1.1194 (8.51) ** 1.4175 (5.39) ** 1.2486 (10.92) ** Dummy variable for availability oi RbCS -0.1025 (-1.58) -0.0250 (-0.88) Real wholesale price of standard sugar in the previous year ' * 0.2152 (2.52) * Real production cost per ton of sugar " -0.3228 (-7.89) ** Downtime -0.4275 (-6.77) ** Loss or sugar dunng the process -0.2112 (-1.64) Duration of the harvest 1.0010 (9.13)** Trend 0.5435 (12.75) ** -3.4499 (-2.72) ** -20.5135 (-6.90) ** -11.9624 (-8.91)** 11.8453 (20.89) ** Total R^ 0.9344 0.9729 0.9905 0.7380 Durbin-Watson 1.9217(1) 1.9961 (2) 1.5933 (1) 1.5357 (2) 1.1648 (3) 1.8715(1) 1.5834 (2) 2.9128 and **: Significant at 90% and 95% confident level, respectively. 1) Deflated by CPI. 2) Values are after corrected by Yule-Walker method. ( ) corresponds to the order of lag assigned. 3) Indicates neither positive nor negative correlation.

PAGE 104

93 o o o s > 1,600 1.400 1,200 1,000 800 600 H 400 200 0 J] Jl Year S Mexico under-quota export I Mexico over-quota export Mexico quota-free export 0The rest of the world export to the U.S. Figure 5-1. U.S. Sugar Import Forecast (Scenario 1 "Baseline") 1,600 1 1,400 oooooooo — — — — — — oooooooooooooo Year @ Mexico under-quota export Mexico quota-free export Mexico over-quota export The rest of the world export to the U.S. Figure 5-2. U.S. Sugar Import Forecast (Scenario 2 "P-S-F')

PAGE 105

94 1,600 T „ 1,400 H ^ 1,200 § 1,000 o CNC<-/TtU-)\Ot-~00CNO — (NC^. TflO OOOQOOOQ — — — — — — OOOOOOOOOOOOOO CSCNCN(N(N(Nr-4CN{NtN(NCNr-)(N Year Q Mexico under-quota export Mexico quota-free export Mexico over-quota export H The rest of the world export to the U.S. Figure 5-3. U.S. Sugar Import Forecast (Scenario 3 "A-S-F") o o o S "o 1,600 1,400 1,200 1,000 800 600 400 200 0 CN 't o o o o o o
PAGE 106

95 ] Mexico under-quota export I Mexico over-quota export Year Mexico quota-free export O The rest of the world export to the U.S. Figure 5-5. U.S. Sugar Import Forecast (Scenario 13 "PA-S-M") 1,600 1,400 1,200 o 1,000 o ^ 800 a > 600 400 200 a. 00 Gv O O O — ' o o o tS (N (N o O o O tN Year ^ Mexico under-quota export Mexico over-quota export o ts Mexico quota-free export The rest of the world export to the U.S. Figure 5-6. U.S. Sugar Import Forecast (Scenario 8 "PA-B-F")

PAGE 107

96 o o o 1,800 1,600 1,400 1,200 1,000 800 600 400 200 0 — j 8 o o o o >n vo t~oo ov o o o o o o o o o o n CS IN fS CN o O O O o o Year I Mexico under-quota export I Mexico over-quota export Mexico quota-free export 0The rest of the world export to the U.S. Figure 5-7. U.S. Sugar Import Forecast (Scenario 12 "PA-C-F") H o o o > 2,500 2,000 1,500 1,000 500 ro r00 OV O CN >o O O o O o o o O O O o O o o o O o O O O O O CM CN (N (N Year M Mexico under-quota export Mexico quota-free export Mexico over-quota export ^ The rest of the world export to the U.S. Figure 5-8. U.S. Sugar Import Forecast (Scenario 16 "T-S-F")

PAGE 108

97 450 Figure 5-9. Forecasted Equilibrium Sugar Prices in the U.S. and Mexican Markets (Scenario 4 "PA-S-F") 450 Figure 5-10. Forecasted Equilibrium Sugar Prices in the U.S. and Mexican Markets (Scenario 8 "PA-B-F")

PAGE 109

98 U.S. price — A — Mexico price Figure 5-11. Forecasted Equilibrium Sugar Prices in the U.S. and Mexican Markets (Scenario 12 "PA-C-F") U.S. price — 4t — Mexico price Figure 5-12. Forecasted Equilibrium Sugar Prices in the U.S. and Mexican Markets (Scenario 16 "T-S-F")

PAGE 110

99 10, 9 o o o o > 000 000 ^ 8,000 000 000 000 000 S 3,000 .000 ,000 0 -Bs o Year U.& demand Mexico demand — H — U.S. Supply Mexico Supply Figure 5-13. Forecasted Sugar Demand and Supply for both the United States and Mexico (Scenario 1 "Baseline") Year U.S. demand • Mexico demand — B — U.S. Supply Mexico Supply Figure 5-14. Forecasted Sugar Demand and Supply for both the United States and Mexico (Scenario 4 "P-S-F")

PAGE 111

100 cs o o o o o o cs o o vo o o r4 o o 1 1 00 Ov O o o — o o o (N (S fS o o o o o Year U.S. price • Mexico price Figure 5-15. Forecasted Equilibrium Sugar Prices in the United States and Mexico (Scenario 14 "PA-B-M") 450 ^ 100 50 0 -1 1 1 1 1 1 1 1 1 1 1 1 1 1 — §0000000 — — — — — 0000000000000 (NCSO
PAGE 112

101 09 o o 00 X) B 'S 3 U O > (A C U U ^ ?^ O 1) o u ^ 2 3 ^ > + is hi 3 Si « c r5 T3 > 00 i3 U « c Ecu O 'u C u o C/2 ^ o o 2 O — , o O 1 00 o 2 o 2 d ^ o in w 00 CO O O O O CO ^ o °9 d -H O (N — ' O On Tt r<-i O m o I 00 — ' O C\ —1 o d !:: O 0\ 00 CO 0\ ON 00 ro On U-) o\ ro 0\ On 00 O .5 CT lU « 3 OQ o 'S c _; 00 (-1 Sso On ON — ' O 00 O ' d ^ ro d ro r~On On 00 rn On IT) On 00 ^ ro 00 oo r<-) ON in ON ro "-^ ro On oi ON °o CN On cn ON TlO 00 00 — ' ON — "1 iri O On ON . (N O d ro O ^ 00 CN O ON ON IT) On ro ^ 00 NO On 00 rn ON V-) On NO iri O On ON ~; 00 o t-.2 ^ c3 CO c , u < GO r o C u o PL, I I < NO O (N ~: NO ON O 1^ ro in ON <» o6 in —1 < -a c C II ° H -3 o 00 U S O o n 3 a, tr (U cs 00 -N 2 II o — 2 a " o. c .2 J= 3 o o x: II < 'c _o O U a, 00 D II U i-T ?3 3 1/3 [/3 (/3
PAGE 113

102 o i3 b 3 1^ o o O o pie 2; ^ CO q ON oo On So d St •a u 'S s 9 :2 o -H O ON m ON ^ O 00 CO C w Q XI X O I O I o u 3 u ^ 5 00 ^ 00 —I ON — ; O On 3 « C O > ^ 1 <^ 2 CO ST' O 2 d w 0\ . CM O O o ^ 2 u ,03 ^ NO 00 o o 03 3 OO h 3 > (L) o o -a NO O ON ON [1^ 3 O 3h c u S o NO O NO CN O NO O ON ON d On (N O d o O 00 o ON oo ON ^ IT) On 00 NO On 00 On ON o d o "1 00 o r~ 00 O ' o5 00 ^:^ ^ 8 O ^ ^ 8 c u 00 o 3 CO 3 PL, 00 o 1 2 1 'i 00 *c ca c 1 c < (U CJ 00 00 O ' •c u 00 r

PAGE 114

103 ^ico Welfare 74.443 (100) 69.095 (92.82) 69.882 (93.87) 64.393 (86.50) Me; Sugar industry revenue 21.737 (100) 12.159 (55.94) 14.532 (66.86) 13.983 (64.33) Net cost (B-C-D) 0.116 (100) 5.979 (5,168.80) 1.820 (1,573.52) -0.200 (-172.54) Cost of buying up excess sugar (D) o ^ 8 O 1 Cost of price support (C) 0.208 (100) 6.112 (2,931.51) O 1 O 1 ed States Tariff revenue from Mexican sugar (B) 0.093 (100) 0.133 (142.87) 0.180 (194.04) 0.200 (215.06) Unit 1 Adjusted welfare (A+B-C-D) 354.370 (100) 338.136 (95.42) 354.819 (100.13) 357.277 (100.82) Welfare (A) 354.486 (100) 344.115 (97.07) 356.639 (100.61) 357.077 (100.73) Sugar industry revenue 35.713 (100) 33.480 (93.75) 27.886 (78.08) 26.516 (74.25) HFCS industry revenue 5.684 (100) 10.132 (178.27) Scenarios Baseline (status quo) Scenario 13 ("PA S M") Scenario 14 ("PA B M") Scenario 15 ("PA C M")

PAGE 115

104

PAGE 116

105

PAGE 117

106

PAGE 118

107 Mexico welfare 72.29 72.29 66.76 on option Mexican sugar 21.31 21.31 20.52 h product HFCS ad U.S. adjusted welfare 354.08 354.85 355.44 Hig High U.S. 35.37 35.21 33.93 U.S. HFCS 10.13 10.13 10.13 Mexico welfare 65.726 65.739 65.739 option Mexican sugar 21.322 21.355 21.355 HFCS ad( U.S. adjusted welfare 353.97 354.89 355.12 strategies High U.S. sugar 35.31 35.09 35.09 U.S. HFCS 10.13 10.13 10.13 Mexico's Mexico welfare 76.24 76.24 71.49 § Mexican sugar 21.87 21.87 23.14 gh produci U.S. adjusted welfare 353.90 354.73 354.97 X U.S. 35.56 35.40 35.39 U.S. HFCS 5.68 5.68 5.68 Mexico welfare 74.44 69.77 69.77 n the current policy [status quo) Mexican sugar 21.74 22.15 22.15 U.S. adjusted wciiore 354.37 354.79 354.92 Maintaii U.S. sugar 35.71 35.40 35.40 U.S. HrL.o 5.68 5.68 5.68 Maintain price support (status quo) Buying up excess sugar in the market Production control with Mexico U.S. strategies V2 X lo ts i2 . on 5 u = 5-1 . CO CO U 0\ IT) 00 00 CO o in NO ON >n t^ t-"n m in 00 NO o o CX 3 Q. O" CO aj — ' H

PAGE 120

109

PAGE 121

110 'ob o o .id o 2 ^ c o Q. O T3 ea on U CO o 3 o 00 s u B c ^ c '3 o c C3 00 1/5 3 U c/l — X c a 1) on T5 — « o i! . 00 on U T3 C C3 ^ 2 00 3 U pa _ u .s ^ 00 C o _ o i3 o X

PAGE 122

111 c o c o o a V3 00 jz S .2? o c fc— C3 O tl ^ t/) -a . -s 1 -S a ^ ^ S ^ . on on U On as ON q 00 On 60 o o c g D. O T3 a U •T3 C 60 3 3 c . -S 1 -S 2i ^ 3 53 X =: as as 00 ON OO ON . C/3 00 U o "o a. 1) o 3 u a .E ^ 2 '3 2 T3 C 60 cn 3 3 C o o 1) « 1) ON as as as 00 •S .2, ,=5 . 00 00 U o o '3 ^ 1> o 3 ^-^ ^ Q. 3 1/5 cr => v= .E ^ S3 '5 S5 _ o ^ c 3 " Da U 3 (U -s o _ w o cj h 2 " o X

PAGE 123

112

PAGE 124

113

PAGE 125

114

PAGE 126

CHAPTER 6 CONCLUSIONS AND IMPLICATIONS FOR POLICY Conclusions and Implications for Policy Sugar is a basic commodity with a long history of utihzation and trade between regions where excess demand and excess supply exists. This trade in the sugar market has provided a classic example for study in the field of agricultural economics. Although its economic value has diminished, as seen in the trend of the declining world prices and a smaller percentage of income spent on sugar in developed countries, its bargaining power continues as a powerful force in the political arena today. Emergence of a new substitute commodity, high fructose com syrup (HFCS) produced predominantly by the United States, has brought about changes in the climate of the sweetener markets, particularly those tied to U.S. markets. These changes are expected to continue. Recent trends in the international trade environment to move towards freer and borderless trade have also accelerated changes in the market climate for sugar trade. When examining sugar trade issues between the United States and Mexico, Mexico's adoption of HFCS and the provisions of North American Free Trade Agreement (NAFTA) related to sugar and HFCS play critical roles for shaping the sweetener market balance. HFCS, already with a large share in the U.S. sweetener market, has gained sizeable market power, as well as power in the political arena. The HFCS industry supports the U.S. sugar program under the umbrella of the American Sugar Alliance (ASA). Despite the efforts of critics, the U.S. sugar program remains intact due largely to successful lobbying efforts by ASA. On the other hand, the growth 115

PAGE 127

116 of HFCS was not welcomed in Mexico. Mexico, a traditional net sugar exporter is now facing pressures from importing a sugar substitute. Although NAFTA promised Mexico favored opportunities to export sugar to the U.S. market, implementation of NAFTA also resulted in opening the door for HFCS consumption in Mexico. The consequences of this threatening trend of increased HFCS consumption was inscribed on the sweetener history when Mexico lost the HFCS dumping case with the United States in 2001. This study attempted to answer five specific questions: (1) What was the impact of changes in trade regime on the U.S. and Mexican sweetener market since NAFTA was implemented in 1994?; (2) How much sugar surplus can Mexico generate, how much sugar will cross the border both underand over-quota, and what will happen after 2008 when all the restrictions are eliminated on Mexican sugar?; (3) What will be the impact of changes in the Mexican market situation and how much influence will HFCS adoption cause in both the U.S. and Mexican sweetener markets?; (4) What will be the impact of changes in U.S. sugar policy on both the United States and Mexico?; and (5) Is there alternative sugar policy for the United States to current price support?. In the following, answers for each question are summarized. Impact of Changes in Trade Regime NAFTA brought about mixed impacts on the United States and Mexico and some are different from what was expected. The Mexican sugar industry has benefited little in the past ten years under the NAFTA regime. By comparison, NAFTA did not bring about drastic change to the U.S. sugar market: expanded exports from Mexico have failed to materialize. Rather, attention was poured into issues of HFCS and its inmiediate impact on Mexico's sweetener market. As seen in the trade dispute over HFCS, the Mexican government struggled to suppress HFCS adoption in its market.

PAGE 128

117 Mexico's Export Potential The opportunity for Mexico to export sugar to reach the 250,000 MT expanded quota (which would be equivalent to approximately 20 percent of the U.S. minimum import requirement) seems unlikely to be enjoyed due to Mexico's restricted production compared to its sweetener consumption. Yet, Mexico possesses a large potential to export over-quota as well as quota-free when all restrictions are lifted. The magnitude of overquota and quota-free export depends on the expansion rate of Mexico's sugar production and HFCS adoption. Specifically, if Mexico adopts HFCS at an increasing rate, a considerable amount of surplus sugar destined to export will be generated, which consequently poses a direct impact on the Mexican sugar industry and an extended impact on the U.S. market. The Impact of Changes in Mexican Market Situation Mexico's sugar production, HFCS adoption, and a tax on HFCS are examined as factors that change the Mexican sweetener market situation. The impact from these three factors on Mexico is eminent. The Mexican sugar industry and welfare become better off from higher sugar production which can be achieved through reduction of production costs and mill downtime; however, this positive impact dissipates once HFCS is adopted at higher rates. The negative impact from adopting HFCS on Mexico originates with the imbalance between domestic sugar production and sweetener consumption, thus making Mexico unable to attain a net sweetener surplus producer. Mexico then faces a large sugar surplus that cannot be sold anywhere but the U.S. market with tariff (over-quota) or the world market. Simulations showed that the sugar industry can tolerate a higher rate of HFCS adoption if the gains from high production exceed the losses from adopting HFCS, whereas the nation's welfare cannot. This implies that the Mexican government faces

PAGE 129

118 difficulty in allowing faster HFCS adoption to happen and thus will continue to struggle to suppress HFCS adoption in its market. The impact from these three factors on the United States is also significant; however, U.S. domestic sugar production and consumption will remain relatively unchanged, although various simulations showed fluctuating U.S. import from Mexico and the rest of the world caused by changes in the Mexican sweetener situation. One notable result is that when Mexico's high production and high HFCS adoption is combined, Mexico's sugar export takes up the entire U.S. minimum import requirement, leaving sugar export from the rest of the world marginalized. This scenario results in not only reduced revenue to the U.S. sugar industry as well as reduced U.S. welfare, but also triggers political pressures from the other sugar exporters whose shares in the U.S. market are at stake. Mexico's HFCS adoption turns out to be beneficial only to the U.S. HFCS industry, but Mexico's tax on HFCS benefits all except for the U.S. HFCS industry. In the latter case, Mexico will not be able to generate sugar surplus and thus export either underquota, over-quota or quota-free; however, gains in producer surplus outweigh the losses in consumer surplus in both countries due to high prices of domestic sugar. This policy lever may not be acceptable in today's trade environment and in fact the decision by the Mexican government has changed back and forth in the past. The Impact of Changes in U.S. Sugar PoHcy Quota allocation and price stabilization policies are examined as possible changes in the U.S. sugar policy. The impact of a quota allocation policy poses a large impact on both Mexico and the U.S. markets. For Mexico, the aforementioned large scale of sugar export is possible only if the U.S. government allocates quotas among exporters in a

PAGE 130

119 flexible manner; if not Mexico's export will be dampened because of its comparative disadvantage to the rest of the world. For the United States, accommodating Mexican sugar can serve as an alternative form of price support because of the higher price of Mexican sugar compared to that from the rest of the world; otherwise the U.S. price will likely dip below the U.S. support price level caused by a large quantity of sugar from the rest of the world flowing into the U.S. market at a lower price. In this sense, sugar trade agreements were mutually beneficial for both the United States and Mexico. Two alternative price stabilization policies are examined in the study: buying up excess sugar in the market and production controls. Simulation results show some improvement in both U.S. sugar industry revenues and cost-adjusted U.S. welfare as a result of reduction in sugar program costs as well as an increase in tariff revenue from Mexican sugar, yet overall positive effects are minimal. It is noteworthy, however, that maintaining the price support may become extremely costly when combined with the situation in where the U.S. government promises to accept a large amount of sugar from the rest of the world. If the U.S. government switched policies from the price support to buying up excess sugar, the timing to do so would be important so as to minimize the cost incurred by the government: the cost of buying up excess sugar will rise immediately after policies are switched whereas the cost of the price support will not because the U.S. sugar price will be maintained relatively high in the early stage of the forecast horizon. When 'buying up excess sugar policy' and 'production control policy' are compared, losses in sugar industry revenues in both countries and in Mexico's welfare are larger for the latter policy, given the similar magnitude of gains in U.S. welfare. In general, these alternative policies still would be more effective than any price-led sugar policy because

PAGE 131

120 the inelastic supply price elasticities in both countries imply that sugar production is not responsive to price changes. Contrasted with the possible improvement in U.S. sugar industry and welfare, Mexico does not benefit from changes in U.S. price stabilization policy. Alternative Sugar Policy by the United States An alternative U.S. sugar policy to the current policy of price supports is sought from the aggregate simulation results using game theory. Cooperative games among industries and governments (the U.S. HFCS industry, the U.S. sugar industry, the Mexican sugar industry, cost-adjusted U.S. welfare, and Mexico's welfare) are also examined by assuming that both nation's welfare is transferable and that total pay-offs are redistributed among coalition members. As expected, the Mexican government and sugar industry will always and harmoniously choose a high production strategy and avoid any strategies that involve HFCS adoption. By comparison, the U.S. sugar industry does not agree with the choice made by the U.S. government: the U.S. sugar industry prefers price support to alternative policies. When coalitions are formed, one case for country coalitions (United States vs. Mexico) and the other for the government/industry coalitions (governments vs. industries), the games reach different solutions: the former game results in U.S. sugar production control strategy and the latter game results in U.S. government follows the strategy of buying up excess sugar. This contrast indicates that the U.S. government prefers to form a coalition with its own industry while the Mexican government prefers to form a coalition with the U.S. government. For the Mexican government, Mexico's welfare improves by having U.S. cooperation at the expense of its own sugar industry. This implies that Mexico faces a choice either of improving the nation's well-being or of

PAGE 132

121 protecting the sugar industry consisting of a large number of growers and related workers. When the U.S. HFCS industry is included in a coalition, a conflict of interest between the governments and industries becomes intensified. When pay-off to the U.S. HFCS industry is pooled into a coalition, the industry coalition (the U.S. HFCS industry, the U.S. sugar and Mexican sugar industries) will lobby for the strategy that involves Mexico's HFCS adoption. Although the game has reached a different solution through government's decision making, this is true because the industries recognize a better payoff to the coalition. This contrast implies that the U.S. HFCS and the U.S. sugar industries have a strong incentive to influence Mexico's choice of strategy. Furthermore, since the Mexican sugar industry theoretically becomes better off from being compensated by the industry coalition, it may be tempted to allow HFCS adoption rather than expecting protection from its own government. The results from games with a grand coalition indicate that there is no solution that satisfies pareto optimality. If redistribution of pooled revenues from industries and welfare is feasible, there are solutions that minimize the total losses for all entities. In summary, in no case did the game reach a solution that included the current U.S. policy of price support as the best policy; alternative policies examined in the study are better sugar policies to pursue. Also, the stance of the U.S. HFCS industry on sugar policy can be influential. If the U.S. HFCS industry is included in the decision-making process, a sugar policy that differs from that chosen by the governments can be lobbied for with enticing side agreements within the coalition.

PAGE 133

122 Limitation of the Study and Suggestions for Future Research HFCS production and consumption are not quantitatively estimated and integrated into the trade model. This limits the effects on the price and quantity of both sugar and HFCS from substitution between the two commodities. In the trade model, the rest of the world is treated as a single homogenous region; however, a country or region such as Brazil or the European Union that embraces significant sugar demand and supply could have an individual impact on the U.S. market. Including changes in sugar stocks in each region as well as flows of sugar-containing goods across borders into the bilateral model is expected to enhance the quality and depth of the results drawn from the model.

PAGE 134

APPENDIX A MAJOR EVENTS IN THE SUGAR INDUSTRY HISTORY IN MEXICO AND THE U.S. Mexico United States 1910 The Mexican Revolution 1917 Farmers ownership of Ejido land started (Article 27, Mexican Constitution) 1972 HFCS-42 production started. 1977 HFCS-55 production started. 1980 Coca Cola and PepsiCo replaced 50% of sugar use with HFCS-55. 1988 Government started privatizing sugar mills and dismantling AZUCAR S.A. 1989 Crystalline fructose production started. 1990 A minimum import quota of 1.256 million MT (raw value) of sugar was established in agreement with the implementation of GATT (October). 1990 Farm bill All major brands of soft drinks utilized 100% HFCS as the nutritive sweetener ingredient. 1991 COAAZUCAR was formed (amendments in Decreto Cafiero). Sugarcane growers started to be paid by sucrose content instead of by the weight of sugar cane (54 % of the wholesale price of standard sugar based on KARBE system). 1992 Agrarian reform (Article 27) allowed to sell/rent Ejido land. 1994 NAFTA agreement (January) Devaluation (December) Sugarcane growers started to be paid 57% of the wholesale price of standard sugar based on KARBE system. 123

PAGE 135

124 Mexico United States 1995 Inflation rate hit 57% while the price of sugar increased 25.7%. 1996 The government announced increases in import duties on HFCS-42, HFCS55, and crystalline fructose to 12.5, above the then-current rate of 10.5 percent. (December). 1996 farm bill 1997 SECOFI initiated anti-dumping investigation (February) and imposed temporary tanff on two grades of HFCS (June). SECIFI published the formula to determine the wholesale price of standard sugar (March). 1995 • CRA (Com Refiners' Association) asked for the review of Mexico's antidumping actions under chapter of NAFTA. WTO established panel for Mexico's HFCS dumping case (January). 1 QOO 2000 WTO panel ruled against Mexico's dumping case (January). Vicente Fox was elected as president, ending 71 years of authoritarian oneparty rule in Mexico (December). 2001 WTO Appellate Body turned down the Mexico's appeal of HFCS dumping case. Mexican government expropriated 27 sugar mills (September). 2002 A 20-percent tax on beverages that contain HFCS was introduced on January; suspended on March 5; and reimposed on July 16. National Sugar Policy for 2002 2006 (February) Source: Created from Polopolus and Alvarez, 1991; Greene, 1998; Garcia Chaves et al., 2002; Buzzanell, 2002; and various issues by USDA.

PAGE 136

APPENDIX B CORN STATISTICS Figure B-I shows the recent com production and consumption for selected countries. Currently, the United States is the largest producer as well as consumer of com, followed by China. However, in the 2002/2003 season, China became a net producer. Trends show that production in China has been increasing for the last three seasons while that of U.S. has been decreasing. Figure B-2 shows the transition of food and industrial com use in the United States during the past two decades. The largest industrial use for com is for fuel alcohol, representing approximately 41 percent of total use in 2002. HFCS is the second largest use, representing approximately 24 percent of total use. HFCS use has been level in the last five years, while fuel alcohol use exhibits rapid growth. Figure B-3 shows transition of the U.S. com price (No.2 Yellow) in Chicago market during the past two decades. The price shows some fluctuation around 2.5 US dollars per bushel. Figure B-4 shows U.S. exports of products made from com in 2002 expressed in million US dollars. Com gluten feed, com gluten meal, and com oil are the dominant products made from com. When three kinds of fructose (fructose solids containing more than 50% fructose. Chemically pure fructose, and fructose syrup with 50%+ fructose) are aggregated, the total value becomes fourth largest, following com oil. If HFCS is already saturated in the United States, it will have better marketing opportunities overseas; however, the business size still will be small relative to that for the leading products. 125

PAGE 137

126 300,000 250,000 H S 200,000 o o o ^ 150,000 B 3 100,000 © > 50,000 0 1999/2000 2000/2001 2001/2002 Year 2002/2003 China production -U.S. consumption U.S. production -Chiifc consumption European Union production Brazil consumption Figure B-1. Recent Com Production and Consumption for Selected Countries Source: Com Refiners Association, 2004 2,500 Figure B-2. Food and Industrial Com Use in the U.S., 1980-2002 Source: Com Refiners Association, 2004

PAGE 138

127 Ji 4.5 Year Figure B-3. Com Price (No.2 Yellow) in Chicago Market, 1981-1998 Source: USDA, 2004 Corn gluten feed Corn gluten meal Corn oil, crude Corn oil, fully refined Modified starches derived from corn starch Glucose syrup not containing fructose or containing in the dry state less than 20% fructose Corn meal Corn starch Glucose (dextrose) Fructose syrup with 50%+ fructose Chemically pure fructose Fructose solids containing more than 50%* fructose 0 50 100 150 200 250 300 350 Value [million US$] Wrd00v71.exe Figure B-4. Exports of Products Made from Com in 2002 Source: Com Refiners Association, 2004

PAGE 139

APPENDIX C DERIVATION OF INVERSE LINEAR EQUATIONS Assuming that markets in both countries are competitive and that the demand and supply equations linear in quantity, the demand and supply functions are expressed in the form ofP^ = a + b*Q^ (b<0) and = c + d*Q^ (d>0), respectively. By definition, the supply price flexibility {Tfp)\s given as: rfp = d*(Q'/P'). (C.I) Since the supply price flexibility is the reciprocal of price elasticity (Efp), the slope of the inverse supply equation is derived from equation [C.I]: d= rfp*(P^/^)=(l/Efp)*(P^/Q^). (C.2) Similarly, the slope of the inverse demand function of the form = a + b*^ is given: b=Tfp *(I^/Q^) =(l/E^p)*{P^/Q^) (C.3) where rf pis the price flexibility for the inverse linear demand function and e'^p is the price elasticity for the linear demand function. In the case of other variables in supply and demand functions, the slopes in the inverse linear functions are also expressed with estimated elasticities. Suppose that the supply function is expressed with cost variables as = c + d*Q^+ e*COST. The slope associated with cost is expressed by definition: e = AP^/ACOST . (C.4) Estimated elasticity associated with cost is: 128

PAGE 140

129 E^cosT^ (AQ^/ACOST)*(COST/Q^). (C.5) With supply price elasticity (Efp), equation [C.4] is expressed as e = i-EfcosT /E^p) *(P^/COST). (C.6) Final forms of inverse linear function for sugar demand and supply for the U.S. market derived from corresponding equations [4.1] and [4.5] are expressed as: , = IU, + IU2 *Q''us. t + Wi *GDP us, , + IU4 *POP us, t (C.7a) or P^us, , = IUi + IU2*Q^us. , + Shifter'' us. t (C.7b) P^us, t = lUUi + IUU2 *^us, , + IUU3 *COST us, t + IUU4 *RCV,, (C.8a) or P^us. , = lUUi + IUU2*Q\s. , + Shifter' us. t (C.8b) Note that the inverse linear demand function is re-specified on an annual base rather than quarterly in order to balance the market with the supply equation and that the inverse linear supply function is re-specified with the current price without lagged quantity for simplicity. Coefficients for corresponding variables are presented in the Table C-1. Table C-1. Coefficients for Inverse Linear Functions -U.S.Inverse linear demand function Inverse linear supply function W2 (l/E''p,usnP^us/Q''us) IUU2 (l/E'p.us)*iP'us/Q'us) IU3 {-E GDP, us /E p,us)* (P^us/GDPus) IUU3 {-E^cosT. us / E^p.us)* (P'us/COSTus) IU4 (-E POP. us/ E p^ us)* (P^ us/POP us) IUU4 {-E^RCv/E^p. us)*(P'us/RCV) In the case of Mexico, the equation for annual total sugar is expressed with estimates of direct consumption of sugar as preliminary estimation showed statistically insignificant estimates associated with price variables in total and indirect consumption of sugar. By assuming that indirect consumption demand is totally inelastic (i.e., a vertical demand curve in quantity-price space), total sugar demand is expressed with a

PAGE 141

curve that has the same slope as direct consumption of sugar with the intercept shifted by adding the vertical indirect consumption demand curve. In this way, the demand equation is expressed in terms of total consumption of sugar (Q^*^, ) with significant estimates (M2 in equation [4.2]), incorporating the effect of change in indirect consumption of sugar as a consequence of the change in HFCS consumption. P^Mx. , = IM, + IM2*Q'^MX.t + IM3*GDPmx , + IM4*P0Pmx. t /Ms*^'^ , (C.9a) or P^MXt = IMi + IM2 *Q'^mx t + Shifter^ MX, , (C.9b) P^Mx,t = IMMi + IMM2*Q^Mx I + IMMs*COSTmx t + IMM4*DT, + IMM5*SUGL0SS, + IMM6*DURTN, (C. 10a) or P'mx, , = MM, + IMM2 *Q'mx t + Shifter' mx t (C.lOb) and the coefficients are presented in Table C-2. Table C-2. Coefficients for Inverse Linear Functions -MexicoInverse linear demand function Inverse linear supply function IM2 (l/E''^P,MXnP^Mx/Q"Mx) IMM2 (l/EfpMxYiP'Mx/^Mx) IM3 {-e gdp, mx /e p mx )* (P'^mx/GDPmx) IMM3 i-E^cosT, mx/ E^p,mx)* (P'mx/COSTmx) IM4 {-E POP, MX /E p,mx)* (P^mx/POPmx) IMM4 {-E'dt/E'p_mx)*(P'mx/DT) IMM5 (-E'suGLOss / E^p. mx) * (P^Mx/SUGLOSS) IMMs {-E'durtn/ E^p, mx)* (P'mx/DURTN)

PAGE 142

LIST OF REFERENCES Alvarez, Jose and Leo Polopolus. "The Florida Sugar Industry." EDIS document SC042, Food and Resource Economics Department, University of Florida, Gainesville, FL, June 2002a. Alvarez, Jose and Leo Polopolus. "The Sugar Program: Description and Debate." EDIS document SC020, Food and Resource Economics Department, University of Florida, Gainesville, FL, June 2002b. Azucar S.A. de C.V. Estadii'tica Azucareras. "Ventas de azucar el pais por clase, destino y tipo de operacion 1970-1990." Mexico, DF, 1990. Borrel, Brent. "The Mexican Sugar Industry." International Economics Department, Policy, Research and External Affairs Working Paper Series No.596. Washington, DC: Worid Bank, February 1991. Buzzanell, Peter. "The U.S.-Mexico High Fructose Com Syrup (MFCS) Trade Dispute." In Schmitz, Andrew, Thomas H. Spreen, William A. Messina, Jr., and Charles B. Moss, Sugar and Related Sweetener Markets: International Perspectives (53-64). New York: GABI Publishing, 2002. Buzzanell, Peter, and Ron Lord. "Mexico: Sugar and Com Sweetener, an Update." Sugar and Sweetener S&O V 20(2), June 1995. Comite de la Agroindustria Azucarera (COAAZUCAR). "Desarrollo Operative Campo-Fabrica 1996/2002." Internet site: http://www.sagarpa.gob.mx/Coaazucar/menu2/nacional.htm (Accessed May 2003a). Comite de la Agroindustria Azucarera (COAAZUCAR). "Rangos de Superficie y Numero de Caneros a Nivel Nacional Zafra 2000/2001." Intemet site: http://www.sagarpa.gob.mx/Coaazucar/menu6/ann09.htm (Accessed May 2003b). Comite de la Agroindustria Azucarera (COAAZUCAR). "Superficie de Riego y Temporal y su Produccion de Cafia Zafra 2000/2001." Intemet site: http://www.sagarpa.gob.mx/Coaazucar/menu6/sprt03.htm (Accessed May 2003c). Comite de la Agroindustria Azucarera (COAAZUCAR). "Avance del Tipo de Cosecha, Numero de Cortadores, Cosechadoras Integrales y Alzadoras al 17 de Febrero del 2001 Zafra 2000/2001." Intemet site: http://www.sagarpa.gob.mx/Coaazucar/menu4/ult_estim_prod.htm (Accessed May 2003d). 131

PAGE 143

132 Comite de la Agroindustria Azucarera (COAAZUCAR). "Acciones de Cosecha de Cafia que Reflejan Calidad Zafra 2000/2001." Internet site: http://www.sagarpa.gob.nix/Coaazucar/menu4/ult_estim_prod.htm (Accessed May 2003e). Comite de la Agroindustria Azucarera (COAAZUCAR). "Resultados Economicos del Campo Cafieros del Campo de las Zafras 1987/2002." Internet site: http://www.sagarpa.gob.mx/Coaazucar/menu3/indexl.htm (Accessed May 2003f). Comite de la Agroindustria Azucarera (COAAZUCAR). "Relacion Historica de los Precios de la Caiia de Azucar como Promedio Nacional." Internet site: http://www.sagarpa.gob.mx/Coaazucar/menu7/com_precio.htm (Accessed May 2003g). Congressional Research Service, Library of Congress, Agriculture: A Glossary of Terms, Programs, and Laws, 2nd Edition, Washington, 1999. Com Refiners Association. "Food and Industrial Com Use 1980 to Present." Intemet site: http://www.com.org/web/foodseed.htm (accessed May 27, 2004). Diario Oficial de la Federacion. Mexico, DP. Secretaria de Medio Ambiente y Recursos Naturales. March 26, 1997. Farm Foundation. 'Trade Dispute in an Unsettled Industry: Mexican Sugar." Intemet site: http://farmfoundation.org/flags/schwedel.pdf (Accessed August 2003). Food and Agriculture Organization of the United Nations Statistics (FAO). Intemet site: http://apps.fao.org/page/collections?subset=agriculture (Accessed Febmary 2003). Garcia Chaves, Luis R., Thomas H. Spreen and Gretchen Greene. "Structural Reform and Implications for Mexico's Sweetener Market." In Schmitz, Andrew, Thomas H. Spreen, William A. Messina, Jr., and Charles B. Moss, Sugar and Related Sweetener Markets: International Perspectives (81-100). New York: GABI Publishing, 2002. Garcia Chaves, Luis R., Gretchen Greene, Thomas H. Spreen, Daisuke Sano and Chris O. Andrew. Transitions in the Mexican Sugar Industry: An Analysis of the Production and Marketing System. Lake Alfred: Florida Science Source, 2004. General Accounting Office (GAO). Sugar Program: Supporting Prices Has Increased Users' Costs While Benefiting Producers. RCED-00-126, Washington DC, 2000. Greene, Gretchen. "Transitions in the Mexican Sugar Industry." Ph.D. dissertation. University of Florida, 1998. Haley, S. and N.R. Suarez. Sugar and Sweetener Outlook. Washington, DC: USDA, January 2003.

PAGE 144

133 Haley, S. and N.R. Suarez. "U.S. Sugar Policy and Prospects for the U.S. Sugar Industry." In Schmitz, Andrew, Thomas H. Spreen, William A. Messina, Jr., and Charles B. Moss, Sugar and Related Sweetener Markets: International Perspectives (53-64). New York: GABI Publishing, 2002. Institute Nacional de Estadistica, Geografia e Informatica (INEGI). "Social and demographic statistics." Internet site: http://www.inegi.gob.mx/difusion/ingles/fiesoc.html (Accessed July 2003). Koo, Won W. and Richard D. Taylor. "2000 Outlook of the U.S. and World Sugar Markets." Agricultural Economics Report No.444, Northern Plains Trade Research Center, North Dakota State University, Fargo, ND, July 2000. Lopez, Rigoberto A. "Economic Surplus in the U.S. Sugar Market." Northeastern Journal of Agricultural Economics. 19(1): 28-36, April 1990. Mas-Colell, Andreu, Michael D. Whinston and Jerry R. Green. Microeconomic Theory, New York: Oxford University Press, 1995. McCoy, Terry L. "Latin American Sweetener Markets: Economic Reform and Regional Integration." In Schmitz, Andrew, Thomas H. Spreen, William A. Messina, Jr., and Charles B. Moss, Sugar and Related Sweetener Markets: International Perspectives (81-100). New York: GABI Publishing, 2002. Morris, Peter. Introduction to Game Theory, New York: SpringerVerlag, 1994. Moss, Charles B. and Andrew Schmitz. "Coalition Structures and U.S. Sugar Policy." In Schmitz, Andrew, Thomas H. Spreen, William A. Messina, Jr., and Charles B. Moss, Sugar and Related Sweetener Markets: International Perspectives (53-64). New York: GABI Publishing, 2002. Offenbach, Lisa A. Effects of Sugar and Ethanol Related Policies on the Market for High Fructose Com Syrup. Ph.D. Thesis. Department of Agricultural Economics. Kansas State University, Manhattan, KS, 1995. Organisation for Economic Co-operation and Development (OECD). "Statistics Portal." Internet site: http://www.oecd.org/statsportal/ (Accessed November 2003). Petrolia, Daniel R. and P. Lynn Kennedy. "A Partial-Equilibrium Simulation of Increasing the U.S. Tariff-Rate Sugar Quota for Cuba and Mexico." Selected paper presented at the American Agriculture Economics Association meeting. Long Beach, CA, July 2002. Polopolus, L and J. Alvarez. Marketing Sugar and Other Sweeteners. New York: Elsevier Science Publishing Company Inc., 1991.

PAGE 145

134 Secretana de Agricultura, Ganaderia, Desarrollo Rural, Pesca y Alimentacion (SAGARPA), Sistema Integral de Informacion Agroalimentaria y Pesquera (SIAP). "Avance Comparativo Siembras y Cosechas: Perennes: Situacion al 31 de Mayo 2002 y 2003." Internet site: http://www.siap.sagarpa.gob.mx/ar_comdeagr.html (Accessed July 2003). Skully, David W. "Auctioning Tariff Quotas for U.S. sugar Imports." Special article presented in Sugar and Sweetener SSS-223, Economic Research Service, USDA, Washington, DC, May 1998. Spreen, Thomas H., Mechel Paggi, Anouk Flambert, and Waldir Femandes, Jr.. "An Analysis of the EU Banana Trade Regime." Selected poster presented at the American Agriculture Economics Association meeting, Tampa, FL, August 2000. Electronic abstract at http://agecon.lib.umn.edU/aasa.html/#aaeaOO. Takayama, T and G. G. Judge. "Equilibrium Among Spatially Separated Markets: A Reformulation." Econometrica 32: 510-24, 1964. U.S. Department of Agriculture, Economic Research Service. "Sugar Statistical Compendium Stock #91006." Washington, DC, October 1991. U.S. Department of Agriculture, Economic Research Service. "Com Sweetener Statistics Stock #94002." Washington, DC, September 1993. U.S. Department of Agriculture, Economic Research Service. "Agricultural Outlook." Washington, DC, March 1997. U.S. Department of Agriculture, Economic Research Service. "Agricultural Outlook." Washington, DC, September 1999. U.S. Department of Agriculture, Economic Research Service. Sugar and Sweetener Situation and Outlook Yearbook. Washington, DC, various issues, 2001a. U.S. Department of Agriculture, Foreign Agricultural Service. "Mexico Expropriated 27 Sugar Mills." GAIN report #MX1 161, Washington, DC, September 2001b. U.S. Department of Agriculture, Foreign Agricultural Service. "The North American Trade Agreement." Internet site: http://www.fas.usda.gov/info/factsheets/nafta.html (Accessed July 2001c). U.S. Department of Agriculture, Economic Research Service. Sugar and Sweetener Situation and Outlook Yearbook. Washington, DC, various issues, 2002a.

PAGE 146

135 U.S. Department of Agriculture, Foreign Agricultural Service. "New National Sugar Policy." GAIN report #MX2031, Washington, DC, February 2002b. U.S. Department of Agriculture, Economic Research Service. Sugar and Sweetener Situation and Outlook Yearbook. Washington, DC, various issues, 2003a. U.S. Department of Agriculture, Foreign Agricultural Service. "Sugar: World Market and Trade." Internet site: http://www.fas.usda.gov/htp/sugar/2003/may (Accessed July 2003b). U.S. Department of Agriculture, Economic Research Service. Feed Outlook Rreport. Washington, DC, 2004. U.S. Department of Commerce, U.S. Census Bureau. Internet site: http://www.census.gov/ipc/www/idbsprd.html (Accessed February 2003). U.S. Department of Labor, Bureau of Labor Statistics. "Consumer Price Index." Internet site: http://www.bls.gov/cpi/home.htm (Accessed November 2003). Varian, Hal R. Microeconomic Analysis, ^'^ edition. New York: W. W. Norton & Company, 1992.

PAGE 147

BIOGRAPHICAL SKETCH Daisuke Sano was bom on January 27, 1967, in Japan. He received the Bachelor of Science in 1989 and the Master of Science in 1991 in applied biochemistry from the University of Tsukuba in Japan. He served the Ministry of Agriculture, Forestry and Fisheries of Japan as Technical Officer from 1992 to 2000; meanwhile he received the Master of Science in food and resource economics from the University of Florida in 1999 funded by a scholarship from Japan International Cooperation Agency (JICA). He received the Doctor of Philosophy in food and resource economics from the University of Florida in 2004. 136

PAGE 148

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. Thomas H. Spreen, Chair 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. Lisa A. House, Cochair Associate Professor of Food and Resource Economics I certify that 1 have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quahty, as a dissertation for the degree of Doctor of Philosophy. Chris O. Andrew 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. „ Terty L. M^Coy Professor of Political Science 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. Kenneth L. Buhr Assistant Professor of Agronomy

PAGE 149

This dissertation was submitted to the Graduate Faculty of the College of Agricultural and Life Sciences and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August 2004 Dean, College of Agncu Sciences Dean, Graduate School