|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 WELFARE ANALYSIS IN INTERNATIONAL SUGAR TRADE: THE CASE OF T HE EU ACP SUGAR PROTOCOL By SIBUSISO MOYO 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 2012
2 2012 S ibusiso Moyo
3 my family
4 ACKNOWLEDGMENTS I thank my dissertation committee, headed by Thomas H Spreen, co chaired by Zhifeng Gao. I am also grateful to Doug las Waldo, Andrew Schmitz and Edward Anthony Evans who served as me mbers of my dissertation committee for their wou ld also like to extend my sincerest graduate to the Food and Resource Economics Department (FRE) at the University of Florida for giving me the opportunity to enroll and proceed with my graduate studies Lastly many thanks to all faculty members, in FRE an d the UF Economics department that I consulted along the process for their generous contribution.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 10 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 INT RODUCTION ................................ ................................ ................................ .... 13 Background ................................ ................................ ................................ ............. 13 A Historical Overview ................................ ................................ ....................... 13 Why Study Sugar ................................ ................................ ............................. 15 W orld Sugar: Production, Domestic Consumption, Imports and Exports ......... 18 EU Sugar Policy and International Commitments ................................ ............. 21 Sugar protocol ................................ ................................ ............................ 22 SPS agreement ................................ ................................ .......................... 23 Everything but Arms initiative ................................ ................................ ..... 24 The Pre Reform EU Sugar Situation ................................ ................................ 26 The Pressure to Reform ................................ ................................ ................... 28 Reforms to the EU Sugar Policy ................................ ................................ ....... 30 The Research Problem ................................ ................................ ........................... 31 Overall Aim and Individual Research Objectives ................................ .................... 32 Outline of the Study ................................ ................................ ................................ 33 2 LITERATURE REVIEW ................................ ................................ .......................... 39 The Benefits of Trade ................................ ................................ ............................. 39 Sugar Trade Models ................................ ................................ ............................... 45 Effects of Trade Policies on the ACP EU Sugar Protocol ................................ ....... 51 3 METHODOLOGY ................................ ................................ ................................ ... 58 The Spatial E quilibrium Model ................................ ................................ ................ 58 Modeling the Price Floor in the EU ................................ ................................ ......... 62 4 EMPIRICAL ESTIMATION ................................ ................................ ...................... 66 Estimation of Demand Elasticities ................................ ................................ ........... 66 Data for Estimating Demand and Supply Elasticities ................................ ........ 69
6 Estimation of Supply Elasticities ................................ ................................ ....... 70 Estimation of Transportation Costs ................................ ................................ .. 73 The data ................................ ................................ ................................ ..... 75 Estimation ................................ ................................ ................................ .. 76 Results of the analysis ................................ ................................ ............... 78 Tariffs and subsidies ................................ ................................ .................. 80 5 RESULTS ................................ ................................ ................................ ............... 88 Scenario I: The World Sugar Model ................................ ................................ ........ 88 Scenario II: Impact of EU liberalization ................................ ................................ ... 90 Scenario III: Impact of EU and US liberalization ................................ ..................... 92 Scenario IV: Impact of US Only Liberalization ................................ ........................ 92 Scenario V: Impact of Free Trade Across All Countries ................................ .......... 93 Impact of Trade Liberalization on ACP Countries ................................ ................... 94 6 SUMMARY AND CONCLUSIONS ................................ ................................ ........ 107 Summary ................................ ................................ ................................ .............. 107 Research Steps ................................ ................................ ................................ .... 108 Main Findings ................................ ................................ ................................ ....... 109 Recommendations for Future Research ................................ ............................... 112 LIST OF REFERENCES ................................ ................................ ............................. 113 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 118
7 LIST OF TABLES Table page 1 1 Major sugar producers (1980 to 2010) ................................ .......... 34 1 2 ................................ ......... 34 1 3 ................................ ........... 35 1 4 ................................ ........... 35 1 5 Sugar productio .................... 36 1 6 ACP plus India export deliverables per marketing year (MT) ............................ 37 2 1 Current income transfer relative to various indicators (%) ................................ 57 4 1 Demand elasticities used in the model ................................ .............................. 81 4 2 Regression estimates for an acreage equation for Brazil ................................ .. 81 4 3 Supply elasticities for different countries in the model ................................ ....... 82 4 4 Comparison of ship sizes ................................ ................................ .................. 82 4 5 Deter minants of the size of load ................................ ................................ ........ 83 4 6 Determinants of the cost of shipping a ton of sugar ................................ ........... 83 4 7 A comparison of the actual shipping values and predicted values for certain routes ................................ ................................ ................................ ................. 84 4 8 Estimated transportation costs ($ per ton) between countries ........................... 85 4 9 Sugar policies for countries in world sugar model ................................ ............. 86 5 1 The baseline model ................................ ................................ ........................... 98 5 2 Accuracy of prediction of the model ................................ ................................ ... 99 5 3 The baseline model compared with years 2006 and 2010 .............................. 100 5 4 Impact of EU liberalization ................................ ................................ ............... 101 5 5 I mpact of simultaneous EU and USA liberalization ................................ .......... 102 5 6 Impact of US only liberalization ................................ ................................ ....... 103
8 5 7 Free world trade model ................................ ................................ .................... 104 5 8 Producer surplus comparison under different models ................................ ..... 105 5 9 ACP sugar protocol quotas pre reform and exports in 2006 ............................. 106
9 LIST OF FIGURES Figure page 1 1 EU sugar scheme pre reform ................................ ................................ ............. 38
10 LIST OF ABBREVIATIONS ACP African Caribbean and Pacific group of countries EBA Everything But Arms CMO Common Market Organization EC European Commission EPA Economic Partnership Agreement EU European Union FAS Foreign Agricultural Service of the United States Department of Agr iculture GAMS General Algebraic Modeling System GATT General Agreement on Tariffs and Trade GSP Generalized System Tariff HFCS High Fructose Corn Syrup LDC Least D evelopment Country MFA Multi Fiber Agreement MMT Million Metric Tons MSN Maximum Supply Needs UNCTAD United Nations Conference on Trade and Development USDA United States Department of Agriculture STABEX Stabilization of Export Earnings TRQ Tariff Rate Quota WTO World Trade Organization
11 Abstract of Dissertation Presented to the Gr aduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy WELFARE ANALYSIS IN INTERNATIONAL SUGAR TRADE: THE CASE OF T HE EU ACP SUGAR PROTOCOL By Sibusiso Moyo Decembe r 2 012 Chair: Thomas H. Spreen Co chair: Zhifeng Gao Major: Food and Resource Economics The European Union (EU) Common Agr icultural Policy for sugar has evolved since the 1960s and what has evolved is a costly supply management scheme which insulates domestic producers from international competition by means of a system of price supports and prohibitive import tariffs. There is evidence from sugar price data that the p olicies that were followed by the EU have resulted in domestic prices three times higher than world free market prices. Excess supply has been exported to the world markets using expensive market distorting subsidies. Part of the supply management scheme i nvolved granting duty free access for certain amounts of sugar from the African, Caribbean and Pacific countries (ACP), a block made up of mostly former British and French colonies. Following a complaint by Australia, Brazil and Thailand in 2002, claiming that the committed itself under the Uruguay Round Agreements, a World Trade Organization (WTO) panel ruled in favor of the three complainants. The EU was then obliged to bring it s domestic market regulation into conformity with its WTO obligations. To be compliant
12 with WTO regulations, in 2005 the European Agricultural Council agreed to a set of reforms that were to result in a 36% price cut. Given that most ACP countries have ag riculturally based economies, any agricultural sector that guarantees the economy a steady influx of foreign exchange is critical and needs to be cultivated. The study i s a world sugar trade model to understand the effects of EU sugar policy reform on worl d production and how it would affect sugar production in the ACP countries and the rest of the world, and what the effects of reforms are on world sugar prices, production and consumption. Results indicate that liberalization might be beneficial to some me mbers of the ACP countries, which is contrary to what other studies have suggested.
13 CHAPTER 1 INTRODUCTION Background A Historical O verview Sugar is a product of photosynthesis that occurs in all green plants as carbohydrates. Two crops are the main sources of commercial sucrose. They are sugar cane ( Saccharum officinarum ), a perennial grass mostly grown in tropical countries and sugar beet ( Beta vulgaris ) which is a beet root variety mostly grown in cold climates. The sugar cane plant, according to Clarke (1988), is the most efficient collector of solar energy in the plant kingdom, fixing 2% of available solar energy into sucrose. It is the a bility of cane and beet to store large amounts of carbohydrates in their stems and roots, respectively, that their juices are rich in pure sucrose. According to a European Commission (EC) Report (2004), s ugar contained in beet or cane is extracted by diss olving it in water and the resulting juice concentrated into sugar syrup that eventually crystallizes after saturation. Plant waste impurities retained on crystallization color the sugar brown, and it has a sweetening power less than that of white sugar. R efining involves eliminating these impurities to less than 0.5% to obtain perfectly white sugar. Raw beet sugar is not useable as such since the impurities give it a disagreeable taste, thus the industrial processing of beet is always to the white sugar st age of the marketed product. Raw cane sugar, on the other hand, can be ingested. The impurities give it a particular taste, and with it some nutritional value and a natural product image that is desirable to some consumers. World trade in cane sugar is pri marily at the raw sugar stage (EC Report, 2004).
14 There are historical indications that Europe has been engaged in sugar trade dating back to the era of Alexander the Great around 327 B.C. (Laszlo, and Rizzuto, 1990). Other trade examples include sugar bei ng brought in by Arab armies across the north coast of Africa possibly into Spain and Sicily around 700 A.D. In 1100, first samples of the Persian reed, as it was also known then, reached Britain, brought back by returning crusaders, but it was not until the early fourteenth century that sugar started arriving in sufficient quantities to be readily available to the public (Spence, 1997). The first regular trade in sugar to Britain began in 1319 when some Venetian traders started erratic shipments making su gar available only to the rich. Sugar beet cultivation and extraction of sugar from the plant is a much more recent development that occurred in the 18th century. A white colored beet called the r ages, but it was not until German scientist Andreas Sigismund Marggraf became inte properties. During his laboratory tests he succeeded in extracting sugar from thin slices of beets, using alcohol, and crystallizing it. Not much was done of it until a student of Marggraf, Franz Carl Achard, wrote a report detai ling how he had separated a significant amount of sugar. The report was presented to the King of Prussia, resulting in the first beet factory being built and the subsequent commercialization of the crop in 1801 (Laszlo, and Rizzuto, 1990; Spence, 1997). T he Napoleonic wars at the end of the 18th century disrupted what had become a developed flow of sugar from the Western hemisphere to Europe, which had
15 accounted for 12 million tons of sugar between 1690 and 1790. Specifically, the Napoleonic wars and the c ontinental blockade in 1806 prevented sugar from England and its colonies from entering the rest of Europe. Since most of the supply of sugar in Europe was imported cane sugar, shortages due to the war, led to a push for the development of a large scale be et sugar industry. Sugar beet production also grew as a result of market interventions that increased the landed price of sugar imported from colonies. For 300 years, European countries strictly regulated the trade in sugar to facilitate taxation of their sugar producing colonies (Ballinger, 1971). Since colonies were forced to export exclusively to mother countries, sugar tariffs created an indirect incentive for the beet industry to mushroom, and as the industry grew, it was in its best interest to ensure that taxes were maintained on sugar entering European and North American markets, hence maintaining market distortions for decades (Bo r rell and Duncan, 1993). Over the past four decades, sugar trade in Europe has been dominated by the Common Market Organ ization (CMO) which was established in 1968 and whose main goal was ensuring a fair income to community producers and self sufficiency of the community market and that imports were accommodated under a special system for African, Caribbean and Pacific (ACP ) nations Why Study Sugar There is widespread agreement that ever since sugar was introduced to the western world over 400 years ago, the sugar industry experiences the most government intervention among a set of soft commodities. Some governments have over time provided generous subsidies and increased the level of protection to keep the industry viable in view of its importance to the domestic economy, while others have nationalized
16 it in order to raise revenue for public services. Nonetheless, governm ent intervention and overall control of the industry has increased over the years, with the result that sugar is now a highly politicized commodity and sugar markets are one of the most distorted in the world (Abbot, 1990; Polopolus, 2002; Elobeid and Begh in, 2006). while cane sugar is produced primarily in developing countries. The result of this peculiar pattern is that the production of cane sugar is largely uncoordinated which tends to increase the unpredictability of output, by contrast, beet sugar production is subject to a series of policy directives, targets, and price formulae, all of which give the sugar industry stability and provide an incentive for expansion of output (Abbot, 1990). Marks and Maskus (1993) contend that sugar stands out because it has historically been one of the few agricultural commodities that could be produced in both temperate and tropical climates, with low income and high income countries potentia lly in direct competition with each other. Since the 1970s high fructose corn syrup (HFCS) has emerged as a popular substitute for liquid sugar in countries that grow or import corn. The substantial differences in the costs of production across countries a nd products have resulted in sugar being one of the most heavily protected farm commodities in the major developed countries of the northern hemisphere. The production of cane sugar is essentially a two stage process, the first of which is a labor intensi ve operation, while the second is highly capital intensive. Furthermore, the second stage has two main centers of specialization. One located in developing countries, produces and exports raw sugar which is a low value product. The other undertaken mainly in developed countries, produces and exports refined sugar which is
17 a higher value product. This creates serious discontinuities in the chain of production produ ct. The production of beet sugar, being a highly integrated process, does not share the same disadvantages that cane sugar faces. There has been a series of new developments which have profoundly altered the performance and outlook of the sugar industry ( Abbot, 1990). Most of them date from 1974 when the world price of sugar rose to record levels and set off an expansion of output which, in turn caused prices to collapse and set the conditions for successfully negotiating the 1997 International Sugar Agree ments. These modern developments are that (a) production continues to climb, while consumption remains constant or is falling sufficiency, which accelerated after 1974, has reduced the amount of sugar traded internationally and has led to the development of protected domestic markets. Protectionism has led to a situation that in several countries, the domestic price of sugar bears no relation to world prices. Some authors (Abbot, 1990; Polopolu s, 2002) for example, argue that the industry has in fact, surrounded itself with a super structure of protectionism which delays the process of structural adjustment. More seriously, it provided the conditions which enabled alternative sweeteners to devel op to the point where they now pose a serious threat to the long term future of sugar. It is the confluence of political and economic factors of rich and poor countries potentially at loggerheads with each other, the persistence of government policies suc h as tariffs, quotas and taxes on sugar, the potential welfare effects of possible policy
18 changes that make research on the economics of sugar production and trade interesting. World Sugar: Production, Domestic Consumption, Imports and Exports Sugar is one of the most autarkic of soft commodities in the world, whose production spanned 120 countries by 1995 (Hannah and Spence, 1996). It is this desire for self sufficiency that fueled the number of countries producing sugar especially after World War II. As a result world production e xpanded rapidly during the 1950 s where it grew 5.6% annually between 1950/51 and 1955/56. During the latter half of that decade the rate of growth fell to 4.35% per annum, bringing world output to 55.4 million tons in 1960/61 (Abb ot, 1990). Sugar production is spread all over the world, where three major producing areas are the Northern hemisphere beet producers (Eastern and Western Europe, Central Asia, North Africa and North America), Equatorial cane (Asia, Africa, North, Central and South America), Southern hemisphere cane, Oceania (Australia and Fiji), Southern Africa and South America (Brazil and Argentina). In most countries, sugar production is concentrated over three to five month campaigns timed to take advantage of maximu m sugar content. This characteristic creates a strong seasonality in sugar output, availability and exports can lead to world price volatility during the year (Hannah and Spence, 1996). Northern hemisphere beets are harvested largely within a three month p eriod from October to December; equatorial cane is processed mainly from November to April; and southern hemisphere cane from May to December. Consequently, from November to April world stocks of sugar build up rapidly and in a surplus year this can exert downward pressure on prices. Conversely, after April stocks begin to decline, reaching their lowest point in August or September,
19 and this can cause seasonal price rises in a normal year and rapid increases in prices in a deficit year (Hannah and Spence, 1 996). In Table 1 1 (at the end of C hapter 1) data is present ed on major sugar producing countries. The top ten sugar producers in 1980 accounted for 56.3% of world production. Their contribution to world sugar has continued to grow over the 30 year period starting in 1980, with the same countries accounting for about 73.7% of world production as of 2010 (FAS, USDA, 2011). Production statistics indicate that total world production grew by 80% between 1980 and 2010 from 84.6 million metric tons (MMT) to 152.2 MMT. Among top for Pakistan 2010 production statistics are 5.6 times those of 1980. Brazil, the biggest sugar producer by volume, multiplied its production fivefold, while China and India are the other two countries that posted significant production gains of 361% and 276% respectively. Production in the EU grew by 22%, while Cuba was the only top 10 country to have production decline. It declined by 85% over the 30 year time period. With res pect to consumption (Table 1 2), India tops the list with six million metric tons of sugar consumed in 1980, compared to 23.5 million in 2010, a growth rate of 252% over the 30 year period. The other main consumers are the European Union (a 50.5% growth r ate over the same time period), China (302.7%) and Brazil (93.5%). This is not surprising given that these are the most p opulous countries in the world. Consumption declined in the United States ( 3%), Japan ( 26.6%) and Russia ( 26.8%). It is unclear why sugar consumption fell over the past 30 years in these three countries, although the prevalence of high fructose corn syrup, a competitor especially in the US market, could be the reason. To summarize, the top
20 increased by 1 10.9% while global sugar consumption rose by 63.4% in the period 1980 to 2010 About a third of the world sugar produced is traded and this figure has been stable for the past 30 years, as shown in Table 1 3. What has not been stable though is the growth r ate in exports for some countries. Brazilian sugar exports grew tenfold between 1980 and 2010 followed by Guatemala and Thailand that grew nine fold. Colombia was the other country to post significant growth in exports, two and half times in 2010 what they were in 1980. The EU ( 55.2%), Cuba ( 92.6%) and India ( 97.9%) posted the largest percentage declines in exports. Despite sugar being produced in over 120 countries, ten of them account for about 80% of traded sugar in 2010 compared to 67.3% traded in 19 80. The level of domination in sugar imports is less pronounced than for exports (Table 1 4) Russia imports most sugar in the world followed by the EU, the United States, Indonesia and South Korea, comprising the top five. The European Union, despite bein g a sugar importer, also is a significant player in sugar exports, highlighting the unique footprint it has on world sugar markets. The majority of sugar shipped to the EU is imported under a special import quota referred to as preferential sugar based on the Cotonou Agreement that allows for sugar from some ACP nations and India to enter the European markets at zero duty (Schmitz, 2002). A significant number of developing countries are sugar producers, and they rely on the crop to boost their foreign curre ncy reserves. The African, Caribbean and Pacific (ACP) group of countries is one such cohort. It is dominated by Southern African nations, with Swaziland producing most sugar, followed by Mauritius and Zimbabwe,
21 with Kenya and Fiji completing the top five (Table 1 5). Despite their high degree of efficiency in production, sugar quantities are still relatively small and have historically been between 2.9% and 4.4% of world production over the past 30 years. Not all African countries are in the ACP. South Africa is the dominant producer in the African currently produces about a quarter of the African output. Production in Africa tends to be highly susceptible to weather changes for example, output in South Africa fell by 30% in 1983 due to drought conditions (Abbot, 1990). EU Sugar Policy and International Commitments Co operation between the European Union (at that time, the European Economic Community) and African, Caribbean and Pacific nations started in 1957 with the signing of the Treaty of Rome, which gave life to the European Common Market (EU website, 2012). The Treaty of Rome provided for the creation of European Development Funds (EDFs) whose aim was providing technic al and financial assistance to colonies and countries that had historical ti es to members of the community. The Lome Convention established how the European Community would co operate with 77 ACP countries. Lome I of 1975 defined the non reciprocal nature of preferences for most of the ACP countries to the European Economic Community (EEC). Under Lome I, a system (STABEX) was introduced to compensate ACP countries for shortfalls in export earnings due to fluctuations in prices or supply for commodities conc erned. Beyond that Lome II was signed in 1979 and mostly dealt with the mining industry. Lome III was signed in 1984 as a 10 year agreement, shifting focus to promotion of industrial development meant to promote self sufficiency and food security. Lome IV focused on the promotion of human rights, democracy and good governance, strengthening of the
22 position of women in society, protection of the environment, diversification of ACP countries, promotion of the private sector and increasing regional co operatio n. Trade in sugar between ACP Sugar producers and the EU was prior to the reform, which came into effect in 2006, based on the following agreements (a) the EU/ACP Sugar Protocol as annexed in the Cotonou Agreement, (b) the agreement on Special Preferentia l Sugar (SPS Agreement) and (c) the Everything But Arms Initiative (EBA) (Serrano, 2007). Sugar protocol It is documented (Serrano, 2007) that the Sugar Protocol opened the Community market to 1.6 million tons of cane sugar from 18 ACP countries, discriminating against countries which were not signatories. The agreement was first introduced in 1975 as part of the Lome I agreement and was continued through successive Lome agreements. With the expiration of the Lome IV agreement, the Sugar Protocol w as then annexed into the Cotonou Agreement of 2000. Article 1 of the EU/ACP Sugar Protocol emphasizes that the three pillars of the protocol were agreed quantities, guaranteed the Community un dertakes for an indefinite period to purchase and import at guaranteed prices, specific quantities of cane sugar, raw or white, which originates in the ACP states and which these states undertake to deliver it Article 3(2) of the EU/ACP Sugar Protocol e mphasizes that subject to Article 7, these quantities may not be reduced without the consent of the individual countries involved. Guaranteed prices are negotiated annually and refer to bulk sugar cost of insurance plus freight (CIF) to EU ports. Prices ha ve in practice, tended to be the same as those received by community sugar producers.
23 listed in Table 1 6 indicate w hat each country was to supply the EU upon joining the ACP Sugar Protocol. Article 7 states that, if during any del ivery period, a sugar exporting ACP state fails to deliver its agreed quantity in full for reasons of force majeure the Commission shall, at the request of the state concerned allow for necessary additional period for delivery. If this failed, then that qu antity could be reduced in respect of each subsequent delivery period by the undelivered quantity. It could also be decided that the undelivered quantity be reallocated among the other states. SPS a greement When Portugal and Spain joined the EU in 1986, t he ACP formulated a request to supply the raw sugar deficit of Portuguese sugar refineries, and in August 1992, the Commission drafted a proposal for a regulation on supplies to the Portuguese sugar refineries 1 This agreement first brought to light the id ea of maximum supply needs (MSNs), for the Community's refineries (Serrano, 2007). It also introduced the idea of a hierarchy of preference from domestic suppliers, to ACP under the Protocol, and lastly third country suppliers, for example, Cuba and Brazil who would be allowed to export to the EU with no preferences. The SPS agreement with ACP states was reached in June 1995, like the ACP/EU Sugar Protocol, it is a government to government agreement, but unlike the Protocol, it ran for a fixed duration and the ACP states were jointly liable to supply the quantities of sugar covered by the SPS agreement. The main aspect of the SPS Agreement was the concept of maximum supply needs, established with reference to 1 Sourced from http://www.acpsugar.org/old/prot ocols.htm#GuaranteedPrices
24 the EU refineries. So imports varied as per need for a particular year, thus in 1995/96 imports were 344,000 tons while in 2002/03 imports amounted to 217,000 tons (Serrano, 2007). Everything but Arms initiati ve There has generally been consensus that least developed countries (LDCs) should receive mo re favorable treatment in market access than other developing countries. This has led to a gradual liberalization of market access for products from these countries (Gotor and Tsigas, 2006). An EU trade website 2 reports that a United Nations Conference on Trade and Development (UNCTAD) in 1968 recommended the creation of a Generalized System Tariff Preferences (GSP) under which industrialized countries would grant autonomous trade preferences to all developing countries. A waiver was granted in 1971 from Ar ticle 1 of the GATT, which prohibits discrimination, to authorize developed countries to establish individual Generalized Schemes of Tariff Preferences. The first EU GSP scheme was implemented in 1971 but in order to update its scheme on a regular basis an d to adjust it to the changing environment of the multilateral trading system, the EU's GSP is implemented following ten year cycles. In February 2001, the Council adopted Regulation (EC) 416/2001, the so called Everything But Arms Regulation (EBA), granti ng duty free access to imports of all products from LDCs, except arms and ammunition, without any quantitative restrictions (with the exception of bananas, sugar and rice for a limited period). In doing so, the EU extended free access to all agricultural p roducts, doing away with tariffs. Provisions were made for free access for rice and sugar through a process of progressive tariff 2 http://ec.europa.eu/trade/wider agenda/development/generalised system of preferences/everything but arms/
25 elimination starting in 2006, and which were to result in full liberalization in 2009. This agreement extended duty and quota free access to all products originating in LDCs, except for arms and ammunition (Gotor and Tsigas, 2006). In an EU Council Regulation 3 416/2001 it is documented that at the Singapore ministerial conference in December 1996, member countries of the World T rade Organization (WTO) pledged to carry out an action plan to improve access to their markets for products originating in the least developed countries. It is further documented in the same regulation that in June 1997 the Council agreed on a platform cal ling for the Singapore conclusions to be implemented by granting least developed countries not party to the Lom Convention preferences equivalent to those enjoyed by signatories and, in the medium term, duty free access for essentially all least developed country products. When this initiative was adopted, it immediately extended duty and quota free access to a wide array of commodities including meat, dairy, fruits, vegetables, cereals, processed sugar and cocoa containing products and alcoholic beverages For fresh bananas, EU tariffs were to be gradually reduced to zero by the beginning of 2006. At the beginning of July 2001 and for the eight years that followed, the EU Commission opened duty free quotas for raw cane sugar for refining, initially amounti ng to 74185 tons white sugar equivalent and increasing by 15 per cent in each subsequent July to June marketing year and were pegged at 197335 MT in the 2008/09 marketing year. th erefore sugar from these countries now enters duty free S ugar from 18 ACP 3 http://eur lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2001:060:0043:0050:EN:PDF
26 countries also is duty free By granting duty free and quota free access to the EU market, it was expected that these countries would considerably expand their exports to the EU, wi th the commission estimating that imports could reach as high as three million tons a year (Bureau et al 2008). The Pre Reform EU Sugar Situation Pre reform the general philosophy and principles underlying the organization and marketing of sugar in t he community was the creation of a single market among members free of obstacles; creation of a common set of regulations, superseding national regulations, governing the production and marketing of both beet and cane sugar; ensuring that production was ba sed on a quota system; creation of a system of guaranteed minimum prices covering the basic quotas allocated to producers and planters; ensuring that producers shared in the cost of disposing of surplus exports; ensuring that isoglucose (HFCS), a competiti ve product was integrated into the organization of the market; and that imports were accommodated under a special system for African, Caribbean, and Pacific (ACP) sugar (Abbot, 1990). These principles evolved over time, mainly in response to political pres sure, the original policy objectives however remain intact, and the EU is self sufficient in sugar without the accumulation of growing surpluses, stabilizing production in those regions which are not well suited to growing beet sugar, increasing production efficiency in those regions which are, and keeping the cost of the sugar regime within manageable proportions. Sugar beet production in the EU is controlled by the imposition of country specific quotas which apply to white sugar, the final product, and thus have the effect of constraining farm lev
27 states on the basis of previous production. In addition to the basic quota granted, each enterprise could produce an additional quantity set at between 30% and 45% of the basic quota, according to market disposal potential. The basic quota and its supplement formed the maximum quota for each enterprise. The supplement evolved to become the allowed for changes that could occur due to the unpredictable nature of agricultural n both be sold on the EU market or exported, for which but must be exported to th production which was almost non existent in the early years of the CMO, has risen steadily with productivity gains exceeding 2.6 million tons or 20% of production under quota. When added to exportable s exports of a quantity equivalent to the ACP and India preferential imports (1.6 million market, with average exports o f around five million tons (EC Report, 2004). The principles of the EU sugar market regime are illustrated in Figure 1 1, and are adapted from the work of Frandsen et al (2003) and Milner et al (2004). The production of sugar is paid a guaranteed p rice and production of sugar is paid the price The prices of and sugar are linked to the intervention price by charging sugar produced a 2% levy, while sugar is levied a maximum of 37.5% (Milner et al 2004). Domestic
28 consumption is given by the intersection of the demand curve and the intervention price Excess production is exported to the world market at the price with the costs of exports equal to covered by export refunds financed by levies on and production equivalent to The Pressure to Reform For more than four decades, sugar trade in the EU had been regulated by the Common Market Organization (CMO ) which was established in 1968 and whose main goal was ensuring self sufficiency in the EU market (Abbot, 1990). What evolved over time though was a costly supply management scheme for the EU domestic sugar market which insulated domestic producers from i nternational competition by means of a system of price supports and prohibitive import tariffs (South Center, 2007). There is evidence from sugar price data (World Bank, 2011) that the policies that were followed by the EU over time resulted in domestic pr ices three times higher than world free market prices as well as production surpluses were exported to the world markets using expensive market distorting subsidies. Part of the supply management scheme involved granting duty free access for certain amount s of sugar from the African, Caribbean and Pacific countries (ACP), a block made up of mostly former British and French colonies. Many economists (Marks and Muskus, 1993; Borrell and Pearce, 1999) believe that the sugar protocol which allows 18 ACP countri es preferential access to the EU has allowed high cost producers to stay in business despite lack of competitiveness. Australia, Brazil and Thailand, all countries which did not have access to the protected EU market, filed a complaint, in September 2002, against the EU claiming that
29 committed itself under the Uruguay Round Agreements (Milner et al 2004). One of the main issues of the complaint was about 1.6 million tons of sugar that originates from the ACP and India that the EU re exported to the world markets at subsidized rates. A World Trade Organization (WTO) panel and the Appellate body ruled in favor of the three complainants, finding that the EU exceeded its subsidy commitments. The EU was then obliged to bring its domestic market regulation into conformity with its WTO obligations (South Center, 2007). The increasing costs of the Common Agricultural Policy (CAP) have also contributed to internal pressure to reform, while the expansion of the EU has made it necessary to revisit some of these agreements including the Sugar Protocol. To be compliant with WTO regulations, in 2005 the European Agricultural Council agreed to a set of reforms to the EU sugar sector that we re to result in an eventual 36% price cut, with beet growers compensated with direct income payments. In addition, they agreed on a quota buy back scheme, funded by those processors staying in the sugar sector, to compensate processors leaving the sector. The envisaged net result of reform was that by 2011, EU sugar production would be reduced by between 25% and 33% from roughly 20 million metric tons white sugar in 2005 to a figure between 13 and 15 million metric tons. The reforms were expected to leave t he EU sugar price at roughly double the world market prices compared to 2005 prices when it was three times world prices (Gain Report E35225, 2005). Other features essential to the proposed reform include phasing out of sugar intervention; eliminating over quota sugar exports; elimination of re exports of sugar imported under preferential terms.
30 The price reductions, however, contradict the interests of the ACP beneficiaries of the Sugar Protocol, since the guaranteed price under the protocol has tradition ally resembled the EU domestic price. In other words, ACP farmers under the agreement get paid at price levels that prevail in the EU, therefore lowering EU prices affects ACP farmers negatively. The South Center (2007) reports that the EU sugar market ref orm yielded even more consequences for the signatories of the Sugar Protocol, providing for the termination of preferences by October 2009. In September 2007, the EU denounced the sugar protocol, for two reasons, providing for the termination of preference s by October 2009. First, EU policy makers wanted to take pressure off the over supplied domestic market which had proven to be relatively resistant to initial reforms. The elimination of guaranteed imports hence would complement efforts to reduce domestic over supply. Second, there was increasing doubt whether the sugar protocol, if upheld for indefinite duration, would withstand legal challenges under WTO law, mostly because the Sugar Protocol preferences generally violate non discrimination obligations c ontained in the GATT 1994 (South Center, 2007). The EU had received a waiver which, expired in 2007, which allowed it to grant trade preferences under the Cotonou Agreement, of which the Sugar Protocol is an integral part. By 2008, the preferences were hig hly vulnerable to legal challenge from other WTO members. By denouncing the protocol, the EU was eliminating the possibility of further WTO complaints that could have been laid against it as early as 2008, the year after the waiver expired. Reforms to th e EU Sugar Policy According to Council Regulation (EC) No 318/2006 of February 20th 2006, the basic features of the proposal were that:
31 a 4 year phase in period beginning in 2006/07. in period. Sugar production quotas are not reduced except through a voluntary 4 year restructuring program where quota can be sold and retired. Payments for quota are Restructuring is financed by quota levies on producers and processors who do not Compensati on is available to farmers at an average of 64.2 percent of the price cut. The aid is included in the Single Farm Payment and is linked to payments for compliance with environmental and land management standards. Establishment of a prohibitive super levy to be applied to over quota production. The Research Problem ACP countries and the LDCs under the Everything but Arms agreements are about to lose rural income (agricultural production and export revenues, rural labor income) as the EU is further reforming its sugar policy. The loss of guaranteed high sugar prices could exacerbate rural poverty in ACP (and LDC) countries. The issue is then to investigate how to compensate them or what policies could be put in place to mitigate these potential losses. Attemp ting to evaluate the effects of policy changes that have global effects can be challenging. By its very nature, the Sugar Protocol provides an alternative marketing channel for sugar exporting ACP countries. For an average ACP country a share of 40% of dom estic production and 62% of the sugar exports are covered by the sugar quota, i.e. how much they are allowed to export into the EU, and the quota can be sold on the EU market guaranteed EU prices. This basic point leads to a revenue raising impact in typic al years, and in medium run to revenue stabilizing effect (Roland and Dietmar, 1995). It is thus unclear at this point how an EU policy
32 change would affect production in ACP countries and what the implications for the rest of the world are. Given that thes e countries have different sugar production efficiencies, some might benefit and others lose, the question is by how much. It is possible that if full liberalization were to occur in the EU sugar sector, the EU would cease to be a sugar producer especiall y of sugar beets, leading to a rise in the world sugar price. Because ACP farmers are currently paid prices that are two to three times higher than the world sugar price, abolishing the sugar protocol would probably lead to a welfare loss in ACP countries. The extent of the welfare loss is however unknown. There is therefore a need to understand the economic effects of EU reforms, while they are still in their early stages of implementation, in order to better inform policy making in the European Union, ACP countries and the rest of the sugar producing and trading nations. Overall Aim and Individual Research Objectives Given that most ACP countries have agriculturally based economies, any agricultural sector that guarantees the economy a steady influx of f oreign exchange is critical and needs to be cultivated. In parts of the world where poverty and livelihoods can be traced back to the farm. Thus understanding the possible trickle down effects of a policy changes made in Europe is important if poor countries are to win the war on poverty, and get on a path towards sustainable economic growth. Agriculture is also a significant employer in many of these countries, therefore un derstanding the potential effects of CAP policy changes on poor ACP farmers could help fill the void in literature related to liberalization of sugar markets which has tended to focus more on what happens in North America and Europe, with little to no cove rage
33 of Sub Saharan Africa. The study intends to build a world sugar trade model to understand the effects of EU sugar policy reform on world production and how it would affect sugar production in the ACP countries and the rest of the world, and what the e ffects of reforms are on world sugar prices. This study occurs in steps that initially that require a world sugar trade model be built to address the different scenarios that have arisen and need to be understood. Specifically what needs to be understood is the following: a. What are the effects of EU sugar policy reform on world production? How will this affect sugar production in the ACP countries and the rest of the world? Related to this is understanding what the effects of reforms are on world sugar prices. b. What are the welfare benefits, if any, of the EU sugar reforms on ACP countries and the rest of the world? Outline of the Study In addition to t he opening (C ha pter 1), the study has an additional four chapters that cover literature review ( Chapter 2), methodology (Chapter 3), results (C hapter 4) and conclusions ( C hapter 5). Literature review summarizes studies on the impacts of liberalization on sugar markets wi th specific emphasis on the global as well as ACP focus. The objective of a literature review, however, is to summarize and synthesize the arguments and ideas of others as a foundation of the arguments put forward in this dissertation. The methodology chap ter lay s out the theoretical basis of the models used, while the results and conclusion chapters will complete the dissertation.
34 Table 1 1. Major sugar producers Producing Country 1980 1990 2000 2005 2006 2007 2008 2009 2010 Brazil 7,027 7,793 20,100 28,175 26,850 31,450 31,600 31,850 36,400 India 5,170 12,575 20,219 14,170 21,140 30,780 28,630 15,960 19,460 EU 13,646 16,944 19,498 21,648 21,373 17,757 15,614 14,014 16,683 China 2,507 5,618 6,947 9,826 9,446 12,855 15,898 13,317 11,566 United States 5,205 6,070 8,194 7,146 6,713 7,662 7,396 6,833 7,118 Thailand 1,098 3,502 5,721 5,187 4,835 6,720 7,820 7,200 6,940 Mexico 2,763 3,605 4,979 6,149 5,604 5,633 5,852 5,260 4,900 Australia 2,963 3,797 5,448 5,388 5,297 5,212 4,939 4,814 4,700 Pakistan 609 1,987 2,595 2,937 2,597 3,615 4,163 3,512 3,420 Cuba 6,670 8,000 4,060 1,350 1,240 1,200 1,420 1,250 1,000 Top 10 47,658 69,891 97,761 101,976 105,095 122,884 123,332 104,010 112,187 World Total Production (WTP) 84,626 109,967 135,722 140,674 144,550 164,196 163,087 143,540 152,188 Top 10 as % of WTP 56.3 63.6 72.0 72.5 72.7 74.8 75.6 72.5 73.7 Table 1 2. Major sugar consumers Consuming Country 1980 1990 2000 2005 2006 2007 2008 2009 2010 India 6,667 11,535 17,296 20,385 19,870 22,425 23,500 24,200 23,500 EU 11,166 13,921 14,512 17,505 16,800 19,816 16,496 16,754 16,800 China 3,700 7,450 8,476 11,400 11,500 13,500 14,250 14,500 14,900 Brazil 6,098 6,800 9,100 10,600 10,630 10,800 11,400 11,650 11,800 United States 9,665 7,836 9,040 9,089 9,239 8,993 9,590 9,501 9,344 Russia 7,600 6,130 6,300 5,400 5,950 5,990 5,990 5,560 Mexico 3,125 4,038 4,445 5,199 5,406 5,133 4,728 5,065 4,600 Pakistan 818 2,270 3,300 3,750 3,850 3,950 4,100 4,175 4,200 Indonesia 1,739 2,340 3,200 3,550 3,850 4,300 4,400 4,500 4,400 Japan 3,268 2,827 2,142 2,238 2,220 2,300 2,350 2,375 2,400 Top 10 46,246 66,617 77,641 90,016 88,765 97,167 96,804 98,710 97,504 World Total Consumption (WTC) 91,035 106,824 127,372 142,396 141,824 150,411 151,413 153,504 153,265 Top 10 as % of WTC 50.8 62.4 61.0 63.2 62.6 64.6 63.9 64.3 63.6
35 Table 1 3 Major sugar exporters Exporting Country 1980 1990 2000 2005 2006 2007 2008 2009 2010 Brazil 2,333 1,500 11,300 18,020 17,090 20,850 19,500 21,550 24,300 EU 4,909 6,575 6,138 6,028 8,345 2,439 1,656 1,331 2,200 Thailand 569 2,611 4,147 3,115 2,242 4,705 4,914 5,295 5,000 Australia 2,318 2,927 4,123 4,447 4,208 3,860 3,700 3,522 3,600 Cuba 6,583 7,065 3,400 770 730 705 800 725 490 Guatemala 180 502 1,140 1,386 1,391 1,500 1,333 1,654 1,654 United Arab Emirates 0 0 850 1,600 1,665 1,600 1,715 1,725 1,750 South Africa 966 927 1,410 1,010 1,230 1,267 1,154 1,185 870 India 243 32 25 40 1,510 2,680 5,830 176 5 Colombia 288 426 959 1,231 988 942 661 585 730 Top 10 18,389 22,565 33,492 37,647 39,399 40,548 41,263 37,748 40,599 World Total Exports (WTE) 27,306 34,140 41,770 46,930 49,864 51,439 51,535 48,860 50,518 Top 10 as a % of WTE 67.3 66.1 80.2 80.2 79.0 78.8 80.1 77.3 80.4 Table 1 4. Major sugar importers (2000 Importing Country 2000 2005 2006 2007 2008 2009 2010 Russia 5,170 4,300 2,900 2,950 3,100 3,100 2,110 EU 1,786 2,549 2,630 3,530 2,948 3,173 3,450 United States 1 484 1,905 3,124 1,887 2,377 2,796 2,286 Indonesia 1,949 1,450 1,800 1,800 2,420 2,197 2,600 South Korea 1,514 1,652 1,669 1 518 1 648 1 550 1 600 United Arab Emirates 925 1 756 1,730 1,605 1,890 1,930 1,905 Japan 1 650 1 328 1 385 1,405 1,440 1,452 1,416 Malaysia 1 256 1 459 1,414 1,670 1,390 1,430 1,520 Canada 1 207 1 274 1 445 1 294 1,417 1,350 1,300 Saudi Arabia 765 1 155 1 260 1,280 1 625 1 575 1 350 Top 10 17,706 18,828 19,357 18,939 20,255 20,553 19,537 World Total Imports (WTI) 37,023 45,418 44,757 43,504 45,377 48,169 51,298 Top 10 as a % of WTI 47.8 41.5 43.2 43.5 44.6 42.7 38.1
36 Table 1 5 Sugar production in the ACP natio Country Name 1980 1990 1995 2000 2005 2006 2007 2008 2009 2010 Swaziland 254 504 495 550 598 653 634 653 650 658 Mauritius 730 602 532 396 580 550 535 460 480 505 Zimbabwe 314 502 524 583 525 445 436 350 297 259 Kenya 401 441 302 471 489 476 476 518 530 550 Fiji 473 461 535 392 330 275 230 230 190 200 Guyana 310 130 254 325 250 260 265 230 240 300 Malawi 113 175 210 225 215 220 275 310 330 330 Zambia 111 142 155 197 250 260 215 240 390 480 Tanzania 119 95 105 125 255 260 260 285 280 290 Jamaica 251 229 212 216 124 147 164 160 170 130 Uganda 5 30 75 130 200 195 200 240 290 320 Cote d'Ivoire 103 164 150 179 140 145 145 143 150 155 Belize 105 99 104 118 105 120 105 85 100 115 DR Congo 51 60 70 73 75 75 75 75 70 85 Madagascar 117 120 140 77 26 20 15 15 25 25 Trinidad and Tobago 114 121 90 115 35 25 33 0 0 0 Barbados 132 69 40 58 40 35 35 35 35 35 St. Kitts and Nevis 36 25 20 18 15 0 0 0 0 0 Suriname 12 1 1 1 1 0 1 1 1 1 ACP Sugar Total 3,751 3,970 4,014 4,249 4,253 4,161 4,099 4,030 4,228 4,438 % of World Total 4.4 3.6 3.4 3.1 3 2.9 2.5 2.5 2.9 2.9 USDA data
37 Table 1 6. ACP plus India export deliverables per marketing year (MT) ACP Country 1975 2003/04 Belize 39,400 40,394 Congo (Brazzaville) 10,000 10,186 Cote d'Ivoire -10,186 Fiji 163,600 165,348 Guyana 157,700 159,410 Jamaica 118,300 118,696 Kenya 5,000 0 Barbados 49,300 50,312 Madagascar 10,000 10,760 Malawi 20,000 20,824 Mauritius 487,200 491,031 Uganda 5,000 0 St. Kitts and Nevis 14,800 15,591 Surinam 4,000 0 Swaziland 116,000 117,845 Tanzania 10,000 10,186 Trinidad and Tobago 69,000 43,751 Zambia -0 Zimbabwe -30,225 India 25,000 10,000 Total 1,304,300 1,304,745 Source: EC Report (2004)
38 Figure 1 1. EU sugar scheme pre reform
39 CHAPTER 2 LITERATURE REVIEW Many studies have attempted to measure the gains in either industrial market economies or all world economies resulting from liberalizing trade of agricultural commodities. Because the major actors in this liberalization tend to be developed countries, muc h of the modeling effort has been focused on the effects of agricultural trade liberalization of developed country economies and whatever spill over effects there might be to developing economies. Developing countries are usually considered in these models as passive actors who follow the dictates of an international trading system. Agricultural policies in developing countries are complex and tend to serve many goals from tax revenue collection, income redistribution, securing political support or food sec urity issues (Hammer and Knudsen, 1990). The purpose of this chapter is a review of the literature on trade and its significance in the global economy, with an emphasis on the liberalization of sugar markets and an interest in understanding models of tra de in sugar that have been developed over time with applications to the EU ACP Sugar Protocol. T he B enefits of T rade In general there is agreement in the economics literature that gains from opening up economies to trade tend to outweigh gains from implem enting protectionist policies. Dornbusch presented the case for trade liberalization in his 1992 paper. Before giving an account of the case for liberalization, and why it is beneficial to developing countries, Dornbusch presented a short history of protec tion dating back to the Great Depression. In the 1930s a number of industrialized countries adopted restrictive trade policies and manipulated commodity prices to the disadvantage of developing countries that
40 primarily exported commodities and used them as a source of foreign exchange. It appears the developed world did this hoping that developing producing nations would eventually face declining terms of trade. Post World War II, the developed world moved towards market liberalization while developing co untries especially in Latin America pursued the policy of import substitution, i.e., developing local industries while protecting them from outside competition using tariffs, quotas and import licenses. In the late 1960s and 70s many countries recognized that protection using tariffs and quotas did keep imports out, but that the resulting decline in demand for foreign exchange led to the appreciation of the domestic currency, and hence a severe tax on exports of both traditional commodities and emerging i ndustrial goods. So because of the harm in exports that can be caused by unstable exchange rates, studies show that countries that adopted outward oriented policies, or those that at least neutralized anti export bias, performed better than countries who f ailed to understand and act on the adverse effects of restrictions on exports. Four main points are laid out in Dornbusch (1992) referred to as the gains to trade. First consumers are huge gainers because, post trade, their incomes can now buy more commod ities, resources are used more efficiently, since they no longer have to produce goods that can be imported at a lower price. Trade liberalization increases the variety of goods available in a market. It also raises productivity, by providing less expensiv e or higher quality intermediate goods. The argument is that in a restricted economy, only a narrow range of goods can be profitably produced and therefore a full range of technological possibilities, which rely on a potentially broader range of inputs,
41 ca nnot be exploited effectively. Instead access to a variety of foreign inputs at a lower cost shifts the economy wide production function outward, leading to higher national output and more goods being available for trade. Also free trade according to Dorn busch, leads to a more economically rational market structure. Gains from liberalization can result from scale economies and economies of scope that must rise in wider markets. Markets in protected economies are narrow and lack competitors from the rest of the world, and this has a tendency of fostering inefficiencies like oligopoly. Market structure can be different under free trade because protectionism can create market power for domestic firms that they otherwise would not have if the competition was mu ch stronger. Dollar and Kraay (2004) hypothesized on what could happen when developing countries liberalize trade and participate more in the global trading system. That is, if growth rates were to accelerate as a result of the moving from a closed to ope n economies, how income and equity issues would be impacted. They suggest that one of incomes and thus alleviate poverty is through opening up their markets to global competi tion. Their methodology identifies countries that opened up to globalization post 1980 based on their growth in trade relative to GDP in constant prices and based on their reduction in average tariff rates. Because of the unavailability of tariff data prio r to 1985, the authors use tariff reduction data for the period 1985 89 to the period 1995 97, and trade volume data from 1975 79 to 1995 97. To understand the experiences of globalizers, Dollar and Kraay compare this group to rich and non globalizing (cl osed) developing countries, reporting simple
42 average and population weighted average of trade volumes, tariffs and growth. The paper reports substantial increases in integration in the world economy among globalizers where the trade to GDP ratio went from 16% to 33% of GDP, while it grew from 29% to 50% of GDP for rich countries but trade actually fell as a share of GDP 60% to 49% for non globalizers. By comparing the rich, the globalizers and the non globalizers their results suggest that trade openness le ads to declining inequality between countries, and declining poverty within countries. Poor countries who participated more in international trade saw their growth rates accelerate which in turn lead to income growth domestically, while developing countrie s with closed economies fell behind. They also emphasize that per capita growth rates increased among the globalizing economies in the 1990s relative to the 1980s, ranging between 5% and 10% in successful economies, while growth in the rest of the developi ng world was stuck below 1%. Lastly concerning the consequences of rapid growth among globalizers on income equality across individuals they note that in the 1990s economies that opened up in the developing world grew faster than rich countries, creating an important trend toward equality among open economies. Work by Chenery et al (1986) suggests that periods of trade liberalization also tend to be periods where total factor productivity is unusually high. Trade is seen as Schumpeterian, creative destru ctionist, in the sense that taking bold moves that rid of barriers can promote a new growth environment. Such a discontinuity can involve introduction of a new good, introduction of a new method of production; the opening of new markets; the conquest of a new source of supply of raw materials and the carrying out of the new organization of any industry.
43 Anderson et al. (2006) considered how agricultural markets and value added would change, if over the decade that began in 2005 and ended in 2015, all mer chandise trade barriers and agricultural subsidies were simultaneously removed. The analysis tool used is a World Bank model called LINKAGE, in collaboration with Global Trade Analysis Project database (GTAP). LINKAGE is a computable general equilibrium mo del (CGE), which differs from other static CGE models in that it is recursive, and can be solved annually. According to Anderson et al. (2006), governments are the main distorters of market equilibrium primarily with border measures. By intervening in mark ets for foreign exchange, governments affect the price of tradeables relative to non tradeables while quantitative trade restrictions like quotas have the power to influence the relative prices of various tradeables. Anderson et al. (2006) argue that phasi ng out import taxes, and converting many non tariff trade barriers to tariffs over the years, has made measuring the extent of distortions to goods markets much easier because attention can focus on import tariffs and agricultural subsidies. The main gist of the paper was to understand how each region of the world welfare, agricultural markets, and farm incomes would change if all trade distortions were removed completely. The LINKAGE model accounted for key global multilateral commitments in its pre simula tion, namely the final stages of the Uruguay Round, phasing out of the Multifibre Arrangement (MFA) and the accession of China and Taiwan to the WTO, and the enlargement of the European Union (EU) from fifteen to twenty five nations. According to the LINKA GE model, removal of all trade barriers would lead to global gains of $287 billion per year by 2015. The biggest beneficiaries
44 would be high income countries who would net two thirds of the $287 billion, however as a share of national income, developing co untries would gain more, with an average increase of 0.8% compared to 0.6% for high income countries. When decomposed by sector the results indicate that a worldwide move to liberalize agriculture and food markets would contribute 63% of the total worldwi de gains, even though the share of agriculture in global GDP was 4% while merchandise trade contributes 9%. So basically liberalization would enhance trade while global value of output would remain unchanged. Therefore the global share of agricultural food and production exported rises from 9.5% to 13.2% which translates to $192 billion in increase in exports to developing countries. Even though Latin American countries would gain the most, low income countries would sell an extra $36 billion worth of goods per year, or a 52% increase. Numerous middle to high income countries are projected to lose farm jobs in the model, between the 2005 and 2015 period. For the most protected farm sectors the rate of farm employment decline would more than double if the wor ld were to move to completely free trade, requiring more members of farm households to seek off farm employment. Turning to specific agricultural commodities, rice and sugar are especially noteworthy. Their global shares of production exported treble. Dev eloping countries share of global output especially output rises in the case sugar, from 62% in the baseline to 80% under full liberalization. Cotton is a product that is important to several countries in sub Saharan Africa and South America (Brazil), and yet it is also produced under high subsidy in the United States. LINKAGE predicts that under full liberalization, the value of cotton production would drop by a third or $5 billion mostly in the USA and
45 the value of exports would decline by $3.6 billion. World totals would not change as the slack would be picked up by developing countries. The benefits from increased production would account $1.1 billion in net income for sub Saharan African countries and cotton exports of $1.9 billion per year in the abse nce of trade barriers and subsidies. This is equivalent to about a quarter in net gain in agricultural value added in sub Saharan Africa from full liberalization. The results also support the notion that returns to unskilled labor rise substantially in de veloping countries, and by more than wages of skilled workers which implies that full reform would likely improve equity and reduce poverty in developing countries given that the vast majority of poor earn income as unskilled workers. The bottom line, the refore, is that according to the latest GTAP database and the LINKAGE model, developing country agricultural production, employment, and real net income would increase by 2015 if all current distortions to world trade merchandise were phased out. Sugar T r ade M odels World sugar trade models have been around for decades to answer questions that have been relevant at different times. Bates and Schmitz (1969) developed a spatial equilibrium model to understand the long run price and trade effects of the United States 1960 embargo on Cuban sugar. Among the objectives of the model were estimation of the cost of transporting sugar using ocean going vessels, ascertain optimal trade patterns, predict prices and trade flows and calculate the long run price and trade effects of the US embargo on Cuban sugar. Gemmill (1976) described a world sugar economy using econometric techniques to understand production and modeling different policy scenarios. The purpose of this
46 study was to estimate supply and demand functions f or sugar for each of the major producing and consuming nations of the world, and use these functions to develop a model that would allow for the understanding of impacts of different policies. The model was conceived in both static and dynamic forms to giv e solutions both in long run equilibrium and in an annual, recursive mode. Gemmill (1976) utilized a spatial equilibrium model of the Samuelson (1952) type but rather than using quadratic Seale, because he found that is was much easier to adapt to a variety of trade policies than quadratic programming. Recently, Poonyth et al (2000) evaluated the impact of the World Trade Organization (WTO) restrictions (Uruguay Round agreement) on the EU sugar sector and the world sugar market. Using a non spatial static equilibrium model, they find that complete elimination of export subsidies by 2005 would require either a 10% reduction in production or the combination of 8% reduction in quotas and an 11% reduction in intervention prices. This emphasizes the point that the world market impacts of reductions in subsidized EU sugar exports depend on the manner in which those reductions are achieved. Relying on quota reductions alone results in smaller re ductions in total EU exports than if the intervention prices are reduced. They also find that higher world prices resulting from reduced EU exports would result in increased production of unsubsidized C sugar, the type of sugar that cannot be sold within t he EU, and thus has to be exported. Koo (2002) analyzed the impacts of alternative trade liberalization policies in the United States and European Union (EU) on the US sugar industry applying a model
47 developed by Benirschka et al (1996) which consisted of 17 sugar consuming and producing nations. The Benirschka et al (1996) econometric model estimates sugar production, consumption and carry over stock equations for all countries using time series data. Area and yield equations ar e used to determine supply. Area harvested is modeled as a function of lagged area, prices of sugar and alternative crops and government policies. Supply is given as product of area harvested and yield per hectare, while consumption is modeled on a per cap ita basis with price of sugar, income and time trend as the explanatory variables. Total consumption is computed by taking the product of per capita consumption and the population of the country. In Koo (2002) the market equilibrium condition requires that be zero, resulting with the aggregate excess demand equation solved for the equilibrium price. Three scenarios are considered by Koo (2002). In the first, US eliminates its TRQ and loan rates on sugar for 2001 to 2004, while other countries maintain their subsidies and import restricting programs. The results were that the world price of sugar would go up 32% in 2004 because increased US sugar imports raise demand for sugar in the world market. US wholesale prices would decrease 20.4% because imports increase US sugar supply. Sugar beet production would decline by 16.2% while sugar cane production would also decline by 11.1% in 2004, while consumption increases 5.5% in the same year. In the second scenario, the EU eliminates its import restrictions and subsidies on sugar for 2001 to 2004, while other countries maintain their subsidies and import restricting programs. The result of this according to Koo (2002) would be an increase in
48 the world price by 21.6% in 2004 The US wholesale price of refined sugar would increase by 6.5% mainly because the US restricts imports to stabilize price of sugar. US sugar cane and beet production increase, while consumption decreases. The third scenario had both the US and EU elimina ting import restrictions and subsidies for 2001 and 2004, while other countries maintain subsidies and import restricting programs. Under this scenario, the world price of sugar would increase by 68.2% from 9.68 cents per pound to 16.28 cents per pound in 2004 because the EU would have become a net importer, and so does the US, and as they both increased their imports global prices would increases due to increased global demand. US wholesale prices would respond by decreasing 4.7% while sugar beet and sugar cane production went down by 7.2% and 3.3% respectively. Consumption in the US would increase slightly by 2.1%. Elobeid and Beghin (2006) analyzed the impact of trade liberalization, removal of production subsidies and elimination of consumption distortio ns in world markets using a partial equilibrium international sugar model calibrated on 2002 market data and 2006 policies. Their model referred to as CARD 1 is a non spatial, partial equilibrium econometric world model consisting of 29 countries / regions with a rest of the world aggregate to close off the model. The characteristics of countries in the CARD model is that they are major sugar producing, exporting and importing countries for which only raw sugar production use and trade is specified. The mod el does not disaggregate raw sugar trade from refined sugar trade. Elobeid and Beghin (2006) chose not to model the ACP explicitly because of the consensus view that ACP countries would not be able to aranteed price. 1 Center for Agricultural an d Rural Development, Io wa State University.
49 Individual country sub models include behavioral equations for area harvested, yield, production of sugarcane and sugar beets on the supply side, and per capita consumption and ending stocks on the demand side. Equilibrium prices, quantiti es and net trade are determined by equating excess supply and excess demand across countries and regions. The Elobeid and Beghin (2006) paper considers three scenarios, the first of which starts with the removal of trade distortions affecting the sugar mar kets which are mostly tariffs, exports taxes/subsidies, tariff rate quotas (TRQs) and state trading. The second scenario considers the further removal of domestic production policies in addition to the trade liberalization of the first scenario, while the third set considers the additional removal of consumption distortions, which are the least frequent, along with previous reforms considered under scenarios one and two. The removal of trade distortions increases the world sugar price by 27% in the year 2011/12 while aggregate trade increases by a moderate 4% in the same year. Elobeid and Beghin (2006) found that in OECD countries where sugar producers are also protected b y domestic policies, removal of trade distortions has a small impact, even though consumption increases as sugar users face the world sugar price. By implementing the second scenario, a 48% price increase is induced in 2011/12 causing aggregate world sugar production and use to decrease by about 3% on average. Under this scenario, major sugar production relocation would take place away from highly protected OECD markets towards competitive producers in moderately protected developing countries chiefly Braz il and Cuba and to moderately protected OECD countries mostly Australia. The third scenario is full market liberalization, and it tended to be felt mostly within countries implementing the reform and less on a global scale.
50 Nolte et al (2011) states in t heir paper that the current Common Market Organization (CMO) for sugar is expiring in the year 2014/15 and that there is a possibility that quotas in sugar production and marketing might be abolished. If this scenario were to happen, they confer that produ cers within the EU would still be protected by tariffs, however they foresee increased competition between EU and member states and re allocation of sugar production to more competitive regions possibly discouraging future imports from the ACP, LDC and oth er nations. The paper uses a Takayama and Judge (1971) type spatial equilibrium model to analyze how the abolition of sugar quotas after 2014/15 would impact production, prices and EU imports. According to Nolte et al (2011) the spatial equilibrium frame work, even though an adequate model for this type of analysis, is known to perform poorly in reproducing trade matrices. To overcome this problem they calibrate the model by attaching a non linear costs term to each trade flow. Their model is formulated as a mixed complementarity p roblem, which means that their problem does not have an objective function to be optimized. In their critique of the spatial equilibrium framework Nolte et al (2011) state that such models behave like tools for normative analysi s despite the fact that optimization in economics is a tool for positive economic analysis. The linear programming formulation is restricted to trade flows where is the number of exporting and importing regions, thus the trade flow restriction do es not allow the replication of observed trade patterns of products that show trade flows that exceed the number of constraints. According to them the model does not allow for cross hauling. Further on, they state that because of the failure of the spatial equilibrium framework to account
51 fully for observed trade patterns, they adopt a hypothesis of non constant and non uniform transaction costs which allows them to attach a non linear cost term to each trade flow. One of the main results of Nolte et al ( 2011) is that under all world market price scenarios, the abolition of the quota lead to an increase in production in the EU and correspondingly to a decrease in preferential imports. The higher the world market prices, the more pronounced was this tendenc y. Their model predicts that if the world market prices were sufficiently high, preferential imports were entirely displaced and the EU turns to exporting to the world market again. Effects of T rade P olicies on the ACP EU Sugar Protocol On September 28 2 007 the Council of the European Union issued a press statement terminating the sugar protocol (ACP EC Cotonou partnership agreement) also ceasing imports from India under the same agreement with effect from October 1 2009. This agreement had since 1975 ena bled certain African, Caribbean and Pacific (ACP) states and India the ability to supply sugar to the EU market on preferential terms. This was to be replaced by a policy adopted in 2001, but taking effect in 2009, the reforms were implemented, starting with 36% reduction of the guaranteed price for EU producers. Also being negotiate d were Economic Partnership Agreements (EPAs) with ACP states which was intended to create conditions for ACPs to use trade as a tool for development, allowing them to access the EU markets with terms similar to those
52 Anticipating the se changes considerable research was done to understand the Analysis of the EC of development theory that fi nancial aid can be better targeted to indicators of development then trade tied aid, however there seems to be wide agreement between Convention worked well. According to them, the Sugar Protocol could be seen as an international commodity policy since it affected prices, trade and hence, export earnings guaranteed prices, specific quantities of can e sugar, raw or white, which originate from Protocol affects prices, trade, and export earnings and economics welfare of sugar exporting ACP countries. To provide an economic evaluation of the performance of the Sugar Protocol over the period 1975 91 Herrmann and Weiss (1995) distinguish between two kinds of economic benefits the transfer benefit, which i s defined here as the welfare gain arising from a higher sugar ex port price under the sugar protocol compared with a hypothetical non protocol situation; the risk benefit, which is a welfare gain for the risk averse planners in the sugar exporting countries arising from the stabilizing impact of the sugar protocol on ex port earnings. The total welfare gain as a consequence of a policy is defined as the sum of the transfer and risk benefit (Herrmann and Weiss, 1995) where the transfer benefit is the increase in income due to a policy and the risk benefit captures the wel fare change which is attributed to a change in the income risk following a policy. They argue that
53 sugar export revenues (R) and the instability of sugar export revenues (I R ) enter the welfare function of a user country of the sugar protocol to give a nati onal welfare function where the impact of a policy instrument on a national welfare can be calculated as (2 1) where is the coefficient of relative risk aversion and is the coefficient of variation of The first term on the RHS of Equation 2 1 is the welfare change in national export second term on the RHS captures the national benefits from reducing costly fluctuat ions of export earnings. It is the risk benefit of the Sugar Protocol. Herrmann and Weiss (1995) concluded that during 1975 to 1991 total sugar export earnings for all countries amounted to 14.4 billion ECU 2 compared to 10.9 billion ECU if the protocol was non existent. The 3.5 billion ECU was the accumulated transfer benefit due to the Sugar Protocol and was interpreted as the welfare gain for the ACP countries plus India. Sugar export earnings of user countries had been increased by 32% in preferentia l situation compared to the non preferential scenario. Without the protocol Herrmann and Weiss (1995) found that instability of sugar export earnings ranged from 36.5% for Suriname to 241.6% for Kenya. With the exception of Suriname, the study found that M auritius had a 70.6% reduction in instability, Jamaica followed with a 69.5% reduction, Trinidad and Tobago 67% and the export revenues of other countries are stabilized by 30% or more. Only Zimbabwe, Kenya and India had 2 The European Currency Unit was a basket of the currencies of the European Community member states, used as the unit of account of the European Community before being replaced by the euro on 1 January 1999, at parity. The ECU itself replaced the European Unit of Account, also at parity, on 13 March 1979.
54 stabilization effects below 10% mai nly because of their relatively low share of sugar exports to the EU. The median stabilization impact of the Sugar Protocol across user countries, excluding Suriname, amounted to 33.6%, i.e., reducing instability in many countries by more than a third. B orrell and Hubbard (2000) quantified the cost of EU protection and suggest that to quantify the effects of the Common Agricultural Policy (CAP) on the EU and the rest of the world requires imagining how the economies of many nations would change if the CAP on sugar were to be abolished, replacing it with free trade. To measure the aggregate effects of the CAP they adopted and modified the Global Trade Analysis Project (GTAP) database and standard model developed by Hertel (1997). To simulate what would happ en if the CAP were abolished, EU barriers to trade and direct subsidies were eliminated from the model. This serves to remove all the CAP support to EU farmers and lowers the prices they receive and EU consumers pay. The model then simulates how EU produce rs and consumers would respond to such changes. Borell and Hubbard (2000) found that the effects of the CAP were that, firstly, it was the largest source of distorting subsidies. Specifically producers of over 40% of world production receive prices that ar e 50% to nearly 400% higher than the world price. But on a value basis, it is the subsidies in the EU that are largest. Quota restricted access to the EU market provided export subsidies of $560 million a year to over 20 countries, with Mauritius being the biggest recipient with nearly $200 million in subsidies a year. Secondly, world prices would rise up to 38% with trade liberalization. Their model predicted a fall in prices by around 65% in Japan, 40% in Western Europe, and 25% in the United States, Mexi co, Indonesia and Eastern Europe. Lower prices in these
55 countries would induce increases in consumption and decrease in production of sugar, which would raise import demand and increase world prices by 30% to 38%. In response, efficient low cost producers would increase production. While it has been argued that the liberalization of agricultural trade offers potentially large benefits for developing countries [e.g. Dornbusch (1992), Dollar and Kraay (2004), Anderson and Tyers (1991)], there are a substanti al number of developing countries which may lose out, namely those that currently enjoy preferential trading arrangements, either bilateral or multilateral, with developed countries (McDonald, 1996). During the period 1975 92 gross income transfers are est imated to have been worth US$4.45billion at 1987 prices, or, on average 0.69% of GDP in ACP countries. There are, however, substantial variations across countries in magnitude and relative importance of quota rents. For Belize, Fiji, Guyana, Mauritius, St Kitts and Swaziland they represent a large proportion of GDP and would undoubtedly suffer from substantial reduction in quota rents. With reforms estimates for the year 2000 indicated appreciable reduction in income transfers. The ACP countries as a whole suffer a loss of some US$123million (at 1987 prices) in trading revenues. Most of the income is borne by Caribbean countries, nearly US$41million, and Fiji and Mauritius nearly US$62million. The relatively small populations in these countries mean the loss es will occur in those ACP countries least able to diversify and those most dependent in sugar for foreign exchange earnings. Given the pressure that the World Trade Organization (WTO) exerted on the EU, Milner et al (2004) explored different ways in wh ich reform could affect the transfer of welfare to ACP countries. They concluded that the importance of transfers for each of
56 the protocol countries would depend on the size and degree of diversification of the For example as shown in T able 2 1, for some countries it is important (almost 10%) in Barbados, Belize, Fiji and Swaziland, and yet extremely important (greater than 10%) for Guyana and Mauritius. The transfer per capita figures show a very marked variation depending on the alloca tion of the transfers across countries and on the size of countries in population terms. For some countries the transfer are negligible (less than $1/person) and yet in some countries they are substantial being over $50 per person as in Swaziland, Fiji, Be lize and Guyana, and in the case of Mauritius, it is equivalent to $150 per capita. A review of literature on the impact of the EU Sugar Policy Reforms on the world sugar market yields mixed results. Most studies appear to show that liberalization would i n general lead to the EU becoming a net sugar importer, and with a probable shifting of production to countries with lower production costs. What is not clear is the extent of the impact on welfare and prices. All studies are in agreement though that no ma tter how small the prices changes are, they are at least all positive. To carry out and understand the present effects of the sugar reforms on ACP countries, this research will have to establish what the transfers as a result of the protocol were for each of the participating countries in the years prior to the reforms and afterwards, with the differences used as part of the indicators of the policy impacts.
57 Table 2 1. Current income transfer relative to various indicators (%) Transfer as a percentage of Total sugar exports Total exports Value of sugar production at world prices Value of sugar production at EU prices GDP Transfer per capita ($) Barbados 59.4 7.1 169.6 64.0 0.6 60.2 Belize 43.5 8.9 68.4 25.8 1.9 59.9 Congo, Rep 6.7 0.0 7.7 2.9 0.0 0.2 Cote d'Ivoire 24.9 0.1 11.1 4.2 0.0 0.2 Fiji 48.8 8.6 78.3 29.6 2.9 59.2 Guyana 53.6 12.7 112.4 42.4 8.7 79.4 Jamaica 58.1 3.8 118.8 44.8 0.6 17.2 Kenya 59.3 0.1 1.7 0.6 0.0 0.0 Madagascar 51.8 0.5 51.7 19.5 0.1 0.3 Malawi 45.4 3.9 31.3 11.8 0.7 1.2 Mauritius 57.2 11.9 138.6 52.3 4.0 150.6 St Kitts and Nevis --0.0 0.0 --Suriname -0.0 0.0 0.0 --Swaziland 35.4 7.0 52.2 19.7 4.3 51.3 Tanzania 59.6 0.6 20.4 7.7 0.0 0.1 Trinidad and Tobago 53.4 0.3 86.3 32.6 0.2 11.3 Uganda 0.0 0.0 0.0 --Zambia 15.8 0.6 12.6 4.8 0.1 0.5 Zimbabwe 26.3 1.1 19.1 7.2 0.2 1.6 Protocol Total 47.3 2.1 61.5 23.2 0.6 3.0 Source: Milner et al (2004)
58 CHAPTER 3 METHODOLOGY The methodology section develops the mathematical programming model in the spatial equilibrium framework that is the basis of the sugar trade model. The S patial E quilibrium M odel Takayama and Judge (1971) developed a spatial price equilibrium model, build ing on the work of Cournot (1838), Enke (1951) and Samuelson (1952), where prices, production, consumption and geographical flows for a single commodity are determined, when linear functions are acceptable approximations to the regional demand and supply f unctions. Let denote the geographical regions which compose the discrete but divisible production and consumption locations, where We assume that each country has both production and consumption regions. There are transportation costs involved, since the countries are separated physically, let these unit costs be expressed as for all and Let us assume that for each country, demand and supply quantities are given by the following linear functions of price: (3 1) where is the quantity demanded in the region and is the demand price in the region and and ; and (3 2) where is the quantity shipped in the region and is the supply price in the region and For each region, assume that the quantity actually consumed is less than or equal to the quantity shipped into the region from all the supply regions. Thus, where for all and and is the quantity shipped from the
59 region to the region. The actual supply quantity is assumed to be greater than or equal to the effective supply from region to all regions, meaning that the quantity shipped between the regions is less than the supply available, i.e., where for all and Given this framework, the objective is to develop a mathematical programming model which yields a competitive spatial equilibrium price and allocation solution. A simple welfare function that only accounts for producer an d consumer surplus can be specified as follows (3 3) which in the case of linear demand equations, the problem can be re specified explicitly as (3 4) and the right hand side can be furthe r simplified into Equation 3 5 which gives a quadratic programming problem. (3 5) The Takayama and Judge model (1971) allows for modifications to be made to accommodate government action. Consider that the importing country or regions through legislation levy import tariffs and exporting countries or regions imposed export subsidies. We represent the cost of transporting a unit of good from to by and a tariff imposed by the country or region on a commodity from the country or region
60 by We represent a subsidy paid by the country or region to her own exporters by exporting the commodity to the importing country or region by Since subsidies are paid to only to the exported part of total supply, while tariffs are imposed on quantity actually imported, the social cost to the consumer is (3 6) We can set up an optimization problem with Net Welfare (NW) as the objective function, with supply and demand balance constraints. Specifically the supply and demand balance constraints require that (a) Outgoing shipments from region 1 do not ex ceed total world supply for all (3 7) (b) Incoming shipments into region 2 do not exceed total world supply for all (3 8) Given the demand and supply equations, we can setup an optimization problem with net welfare (NW) as the objective function, with supply and demand balance constraints. The problem can be presented as follows Max (3 9) subject to for all 1 Note that terms country and region are used interchangeably in this model. 2 Also note that is the supply region, while is the demand region.
61 for all for all and The nature of the solution to this problem can be understood by investigating parts of the Kuhn Tucker conditions. By taking the first order conditions, that is differentiating the Lagrangian with respect to and Let the Lagrangian of the problem be expressed as follows (3 10) where then the derived first order conditions are The first order conditions imply that the shadow price in region j equals the demand price in that region if is positive. Similarly equals the supply price in that region if is positive. Since and this implies that and From equation (3.1.10c) we can substitute and for shadow prices and to get where is to be referred to This can be re arranged such that (3 11) This means that the demand price in region is less than or equal to the supply price in region plus the net tariff of shippi ng from region to A distinguishing
62 characteristic of a spatial price equilibrium problem is that the objective of the researcher is to compute supply and demand prices, and determine trade flows satisfying the equilibrium condition that the demand pri ce is equal to the supply price plus the cost of transportation if there is trade between the pair of supply and demand markets. If the demand price is less than the supply price plus the transportation cost, then there will be no trade. Some countries imp ose an ad valorem tariff on imports such that exporters face a price where is the ad valorem tariff imposed by region j Note that when then This implies that the net welfare function, E quation 3 9 can be modified to (3 12) Even though sugar is produced from sugar beets and cane, the final product is assumed to be homogeneous. The outputs of the model are (a) the net price in each region or country, (b) the quantity of exports and imports for each region or country (c) which regions export, import or do neither, and (d) the volume and direction of trade between a pair of countries. Modeling t he Price F loor in t he EU Thore (1986) extended the spatial equilibri um model by Takayama and Judge (1971) to deal with the case of disequilibrium caused by rigid prices and / or price controls. He extends his approach to the spatial equilibrium model and shows that it can be solved by mathematical programming approaches. A s in Takayama and Judge (1971), let us consider a single good traded in several geographical regions,
63 The demand for the good in region is written and let the demand price function be positive, differentiable and decreasing on The supply in region is written and let the supply price function be positive, differentiable and increasing on as (3 13) which is a concave function. Let denote the quantity transported from region to Unit transport costs are Thore (1986) then forms a mathematical programming problem (3 14) s ubject to for Unlike in the Takayama and Judge (1971) model the Thore (1986) combines the demand and supply balance constraints into a single constraint, which states that sugar demand in each region cannot exceed the local supply and net shipments into and out of the region. Following the work of Thore (1986), let us now assume that the price is rigid (3 15) where and are the price floor and price ceiling in the region respectively. In order to handle the case of disequilibrium, the following concave programming model is proposed (3 16) s ubject to for and
64 The sum of the first three terms on the left hand side of Equation 3 16 denotes demand minus supply in region after accounting for inter regional trade. Note further that the optimal solution to Equation 3 16 must satisfy that both these cannot both be positive at the same time, because if that were the case, it would be possible to increase the value optimal solution by subtracting some positive constant from both of them. Hence according to Thore (1986) denotes a possible excess supply in region and denotes a possible excess demand in reg ion where we can mathematically represent these two as and which means (for the later equation) that excess supply ( ) in region is the difference between what is supplied in region and all the other sugar that is brought in from different regions. Ignoring the price ceiling, since there are none in the world sugar trade model, the Lagrangian for this problem can be set up as follows (3 17) where then the first order conditions derived are defined as follows As in the Takayama and Judge (1971) model, from the first order conditions we find that equilibrium prices and are equivalent to the shadow prices. There is
65 now, however an additional condition to be considered which is used to model policy intervention in the form of pric e control. Adding a price floor, Equation 3 17, results in the condition that but so The equilibrium solution from this model therefore allows us to accommodate the fact that the EU has as one of its price intervention policies, a price floor. This model allows for any country that produces sugar to impose a price floor, however since on ly the EU has such a pricing policy in this model, the price floor vector is denoted by (3 18) which only allows the EU price floor to enter the objective function of the quadratic programming model, thereby accounting for one of the significant policies responsible for maintaining EU sugar prices two to three times above the Caribbean world price. Having developed the theoretical framework for the model, the next step is the estimation of parameters for demand, supply and transportation cost equations that are necessary to build a mathematical programming model.
66 CHAPTER 4 EMPIRICAL ESTIMAT ION The main intention of this research is to quantify the extent of the losses that could be suffered by poor countries that previously had guaranteed access to the EU market, but now face uncertainties as a result of a new set of policies adopted by the EU. To understand the impact of the change in EU policies, the main sugar producing and consuming countries in the world are incorporated into the model. To reach this objective, a world sugar trade model is developed to determine the direction of flows o f sugar from production to consumption regions, and to determine equilibrium prices. The process involves estimating domestic supply and demand equations for all countries in the model. The demand and supply elasticities and transport costs are then used t o set up a baseline model which computes net social welfare, defined as the sum of producer and consumer surplus less transportation costs, using 2009 as the base year. The social welfare function is optimized using the Generalized Algebraic Modeling Syste m (GAMS), a mathematical programming language. This chapter explains the steps followed in estimating the demand and supply equations and the transportation costs. Estimation o f Demand Elasticities The application of econometrics to the theory of demand r equires, in addition to data, a specification of an econometric model. For a consumer, demand functions can be generalized for goods as (4 1) The equations indicate the quantity demanded of each of the goods as a function of all prices and income (Intriligator, 1978). Econometric studies of demand include both single demand equations and systems of demand equations. In the case of
67 a single demand equation, the idea is to select one equation from Equation 4 1, and estimate its parameters such that if the first equation is taken, for example, and a stochastic term, added to account for omitted variables, misspecification and error s in measuring variables A single demand equation study would estimate (4 2) In this study a single demand equation for sugar for each of the countries in the model is estimated. In order to estimate a single demand equation a variety of functional forms have been utilized and perhaps the simplest functional form is the linear one, which can be written as (4 3) where the prices and income are treated as exogenous expl anatory variables (Intriligator, 1978). Gi ven an estimated linear demand, E quation 4 3 the implied elasticity of demand, evaluated at the mean value of price and quantity is given by (4 4) where and are mean values. Gemmill (1980) examined the demand for sugar using pooled data from 73 nations and estimated flexible functional forms for demand equations. In that paper, the quantity of sugar consumed per capita, in the country in the year may be expected to be a function of own price at retail price per capita income prices of subsititutes and tastes thus specifying: (4 5) The Gemmill (1980) study however then drops the price of substitutes from the specification because the main substitutes to sugar, sweeteners and High Fructose
68 Corn Syrup (HFCS), tend to be collinearly related to sugar and are only of importance to a few c ountries such as the United States. Substitution between sugar and other carbohydrates was found to be unimportant by Viton and Pignalosa (1961). That paper finds that pooled international data on sugar suggests that the more complicated forms of specifica tion may not be much superior to the simple semi logarithmic 1 form. It therefore appears that the simplest procedure for estimating demand elasticities consistent with economic theory is to estimate where is the per capita quantity of sugar demand ed, is the wholesale price of sugar relative to other goods, and is the income variables usually measured by per capita Gross Domestic Product. The focus is that the single equation demand model be appropriate, where an appropriate model is defined a s one which generates unbiased (or at least consistent) and efficient parameter estimates ( Thursby and Thursby, 1984). Simple sugar demand equations were estimated following the form presented in Equation 4 3. Consumption at time t was the dependent var iable, with three explanatory variables, price, income measures by per capita Gross Domestic Product, and a time trend variable. (4 6) Results for the estimation for the individual countries are listed in the first column of Table 4 1 and are also compared to elasticities found from other studies. There is a huge variation in ela sticity estimates from running Equation (4 6) and what other stu dies report as shown in Table 4 1. To understand why pr ice estimates for agricultural supply response differ, Askari and Cummings (1977) highlight a list of possible reasons why 1 The sem i logarithmic functional form, which applies logs to the dependent variable only, is commonly used in econometrics because its coefficients represent useful concepts that are easily interpreted. An example of an semi log model is
69 elasticity estimates may deviate from each other. Probably the most important factor confronting a researcher is what price to use. A ccording to Askari and Cummings (1977), prices most frequently cited in estimating supply response include the price of the crop actually received by farmers, the ratio of the price of the crop to some consumer price index, the ratio of the price of the cr op received by farmers to some index of the farmers inputs and the ratio of the price of crop received by farmers to some index of prices of competitive crops. Basically any two studies that employ variations of each of these prices are bound to come up wi th different results, and yet none of these prices might be the right one to use. The fundamental issue though, according to Askari and Cummings (1977) is farmer might wa nt to produce more of a particular commodity. If output is changing because a farmer wants to keep his own consumption of the crop the same, in the face of rising input costs, then using the ratio of crop price to some index of input prices is advisable. I f producers are motivated by the need to buy more goods then crop prices deflated by some index of consumer prices could be reasonable. The same type of thinking applies on the consumption side when deciding what type of price to apply in demand estimation and how different forms of variables are used in the estimation. Data for Estimating Demand a nd Supply Elasticities Estimation of demand elasticities for at least 23 countries in the model can be challenging due to data requirements. In some cases, rel iable price and income data could not be found as in the case of Cuba, Pakistan and South Korea. Whenever such challenges were encountered the elasticities were not estimated, but existing literature was used to establish reasonable estimates of elasticiti es. One such source was the
70 Food and Agricultural Policy Research Institute (FAPRI) a joint project of University of Missouri and Iowa State University, maintains a downloadable elasticity database. In the world sugar model, individual country demand is modeled as a function of income and domestic prices in the country concerned. Sugar consumption information for each country was sourced from the US Department of Agriculture ( www.fas.usda.gov/psdonline/ ) Income (GDP) and exchange rate information for most countries was obtained from the Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania, also known as the Penn World Tables (Heston, et al., 2011). Domes tic sugar prices were obtained from individual country statistical databases. In many of the countries being modeled, the departments of agriculture or the central statistical offices maintain wholesale domestic sugar prices, and they were used in estimati ng demand. The World Bank was used as a source of the domestic prices are if the trade distortions prevalent in a country are known. To estimate supply production, acre age and farm gate price data was obtained from the Food and Agricultural Organization. Estimation o f Supply Elasticities Unlike the straightforward estimation of demand elasticities, there are some methodological difficulties in the estimation of supply re sponse. The main difficulty is that neither planned output nor anticipated price is observable because weather and other environmental factors can make observed output deviate from planned output. On prices, the farmer knows only the current and historical prices. Most time series studies for crops use acreage as a proxy for output because acreage is thought to be more
71 subject to farmers control than output, while most researchers assume that that farmers anticipate prices from their knowledge of current an d past information (Rao, 1980). According to Nerlove (1956), the elasticity of acreage is probably only a lower price, but rather to the price they expect, and this expected price depends only to a not observable, it can be hypothesized that they change from year to year by some fraction of the difference between the actual and the anticipated price in year Since the goal is estimating the supply elasticity, Nerlove (1956) presents a pathway by which an acreage equation can be used to derive at least the lower bound supply elasticities for each of the countries being mod eled. Assuming that acreage is a linear function of expected price and an error term (4 7) and expected price can be defined by the following equation where (4 8) Price elasticities can be computed from the coefficient of price in equation in an acreage equation, however, since Equation 4 7 has unobservable expected prices it cannot be used to estimate elastici ties directly. By substituting Equation 4 8 into Equatio n 4 7 Nerlove (1956) shows that acreage can be described by the following equation (4 9) whose variables are all observable, and is the acreage equation used in studies such as Koo (2002) and Elobeid and Beghin (2006) to estimate supply. Nerlove (1956)
72 showed tha t the parameter estimates from Equation 4 9 can b e related to those of Equation 4 8 suc h that (4 10) Therefore defining which is the elasticity estimate, given that the estimated equation is in log form. (4 11) Having derived an elasticity estimate, the supply equation for a country can then be traced using mean sugar quantities and prices. In the case of Brazil, the above approach can be applied to determine the supply elasticity as the e xample in Table 4 2 show s. Given that and then Also this implies that is the elasticity of supply as measured by the response of acreage to expected price of sugar. The same approach is used to determine the elasticit y of supply for the rest of the countries in the model. Supply elasticities for countries in the model are est imated and presented in Table 4 3. F or those countries where it was hard to source reasonable data to allow an econometric estimation, other studi es are used for elasticity information. shown in the first column of Table 4 3, tended to yield higher elasticity values particularly for the African group of countries, India, Indonesia and Pakistan than other s tudies. This could be due to the reasons cited earlier in relation to the suitability of the type of price data used in the estimation of supply response. Sensitivity analysis will be employed where halving, doubling and
73 tripling elasticity estimates will be implemented then document how the model responds to such changes. Estimation o f Transportation Costs Transport costs between countries can pose a formidable barrier to trade, similar in effects to tariffs and institutional constraints (Binkley and Harr er, 1981). While transport costs are thought to reflect unalterable geographic factors like distance between traders, and hence not policy relevant, there is evidence, however, that many factors can influence rates, and rates need not be proportional to di stance (Binkley and Harrer, 1981). In their econometric model Binkley and Harrer (1981) emphasize that a advantage in shipping as well as in production. Their study provides evidence that distance maybe of relatively small consequence for large vessels, suggesting that a critical factor in shipping advantage is the nature of the port systems at origin and in destinations. Port facilities available, such as channel depth, deter mine the types of ships that can be handled efficiently as they affect the size of the ships that can be handled and how quickly loading and off loading can be done. They also investigate the effect of trading volume on routes and find that location of por ts with respect to major trading routes matters. Implying that a port located along a major route may lower shipping rates than one less favorably located. The Binkley and Harrer (1981) model takes the following form (4 12) where RATE is the rate (dollars per long ton), DIST (voyage distance in thousands of miles), SIZE is shipment size (thousands of tons), TERMS are loading and unloading terms, FLAG is the registry of ship (US or foreign). QUART is the quarter in which the
74 shipme nt occurred, VOL refers to volume of grain for the route in question and PORT refers to the origin and destination ports for the shipment. Martin and Clement (1982) carried out an analysis of factors affecting freight rates at the port. They specifically looked at port specific factors using a case study of the lower Columbia River Port area. Their study, similar to Binkley and Harrer (1981), focused on grains, but both are significant because of the nature of international ocean freight market. Both pape rs emphasize that ocean freight market largely consists of tramp ships and ocean liners. The main difference between the two being that ocean liners tend to belong to shipping conferences which determine routes, work on fixed rates, transporting packaged f inished goods, refrigerated products, and containers for example. The tramp shipping market consists of hundreds of private shipping firms, based all over the world, operating thousands of ships which are available for hire on spot demand basis. They prima rily transport dry bulk commodities like grains and sugar, and do not maintain fixed routes, but instead respond to market forces. Due to the fact that tramp shipping tends to be less disorganized, charter contract negotiations between the ship owner and p otential shipper are represented by brokers who communicate through the Baltic Shipping Exchange in London and other major stock exchanges of the world (Martin and Clement, 1982). The model proposed by Martin and Clement (1982) is a slight modification of the work that was done by Binkley and Harrer a year earlier. It takes the form (4 13) where DIST is distance, TON is shipment size in long tons, LD is lay days, DWT is ship size in dead weight tons, AGE is a ge of ship, DTH is the maximum depth at the
75 destination port, P is size of port, T is terms of shipping, FUEL is bunker fuel price in dollars per barrel, F is flag of registry, and EX refers to US west coast grain exports. Martin and Clement (1982) argue that the Binkley model may not be well specified because it has a coefficient of determination of 46%, and they also suggest that the model appears unstable due to a heavy reliance on indicator variables. Nonetheless the results of these studies are simila r. The data To estimate the per unit cost of freight a dataset for the year 2001 was purchased from Maritime Research Inc. of New Jersey. The dataset had information on the week in which a ship was chartered to transport sugar from a production to a demand region. Spe cified in the dataset are port of origin, destination, the rate charged for each shipment per ton, the terms of shipment and who the shipping (charter) company was. The data describes the main sources and destinations of sugar traded globally in 2001. Spec ifically Brazil and Cuba accounted for 50% of global sugar charters in 2001. Data is not always available to describe the exact country where the refers to Western Europe, including Russia. Because data on sugar charters from Maritime Research Inc. is available until 2001, the original intention was to use oil and ship charter indices to extrapolate data to 2009. For that reason the Baltic Dry Index (BDI) and the Crude Oil Light Sweet NYMEX are utilized. The BDI is a number issued daily by the London based Baltic Exchange. Not restricted to Baltic Sea countries, the index tracks worldwide international shipping prices of various dry bulk cargoes. The index provides "an asse ssment of the price of moving the major raw materials by sea. Taking in 26 shipping routes measured on a
76 time charter and voyage basis, the index covers Handymax, Panamax, and Capesize dry bulk carriers carrying a range of commodities including coal, iron ore and grain. Economists and stock market investors read it because the index measures the demand for shipping capacity versus the supply of dry bulk carriers. Estimation The cost of shipping a ton of sugar is estimated as a function of the following variables, distance, size of the load, the loading rate (time it takes to load a ship with the sugar cargo), discharge rate. Also included are the Baltic Dry Index (BDI) and the oil index ( OILSW.IDX). The data purchased did not have information on the size of the load, which captures information on the size of shipments between countries in the World Sugar Model (WSM). Binkley and Harrer, 1981 indicate that this is an important determinant of cost of shipping. It has been shown in econometrics that OLS esti mators are biased and inconsistent for the case of omitted explanatory variables. It has further been shown the instrumental variable technique or two stage least squares (TSLS) can be used to obtain consistent estimators in the presence of omitted variabl es (Wooldridge, 2003). Consider a simple model (4 14) where we do not have data for LOAD SIZE when sugar is transported between the 23 countries in the model. Since neglecting LOAD SIZE resu lts in biased and inconsistent estimators, in this case the problem is solved by first running a regression that predicts LOAD SIZE. In the second stage, where the cost of transporting a ton of sugar is estimated, then predicted load size is used where there otherwise would have been an omitted variable. Assuming that it takes the following functional form
77 (4 15) To understand the size of load from different parts of the world, the following va riables need to be taken into consideration. First the different ship sizes that operate in the ocean freight industry are a factor. In terms of how much cargo in tons they carry, there are four types of ship sizes Handysize (10,000 to 35,000), Handymax (3 5,000 to 59,000), Panamax (60,000 to 80,000) and Capesize (100,000 and over). Also in deciding on the size of the load, distance, the loading and discharge abilities of ports should be fa ctors. Given this information, Table 4 5 provides the parameter estim ates of a model that fits the following regression (4 16) and are dummy variables for Handymax and Capsize respectively, and they are being compared to the Handysize. The model only has two dummy variables instead of three because there were no ships that fit the Panamax category in the sample data set. Having predi cted the size of the load, then SIZE is used in the next stage of the estimation to predict the freight costs per ton. The regression model for rate is specified as follows (4 17) where LR is the loading rate, DR is the discharge rate, BDI is the Baltic Dry Index, OIL refers to the Crude Oil Light Sweet Index traded on the New York Mercantile Exchange, and LT is the loading time while DT is the discharge time. Because of the seasonal nat ure of sugar harvesting, a variable that captures the time of the year the shipping is done is also included.
78 Results of the analysis Five variations of model based on Equation 4 17 were estimated, and they differ from each other mostly because some vari ables are added and others dropped based on their explanatory power. Model 1 (in Table 4 6) uses the explanatory variables distance, size of load, loading rate, discharge rate, the BDI and OIL indices to predict cost of shipment. In model 2 the BDI is drop ped due to a lack of explanatory power, and model 5 chosen as the best to estimate the cost of shipping sugar between countries per ton. Model 5 is chosen over model 4 because according to model 4 there is a negative relationship between OIL prices and cost of shipping, which is not expected. Instead it is expected that the cost of shipping is positively correlated to oil prices. The estimated model indicates that at t he 5% level of significance the variables that are significant in explaining the variation in freight cost per ton are distance, size of load, loading rate, discharge rate, squared distance and the squared size. The rest of the variables are not significan t. As with Binkely and Harrer (1981), while the coefficient for distance is positive (0.00197), that for the square of distance is negative ( 1.97E 07) which is indicative of a declining effect as distance increases. The estimated coefficients on the shipm ent size and size squared are 0.0009 and 1.09E 08, which according to Binkley and Harrer (1981) suggest the presence of scale economies over a certain range. The loading and discharge rates measure the degree of mechanization at the ports. There is a sign ificant negative relationship between the degree of port mechanization and the cost of shipping. In other words, it costs more to ship a unit of sugar the more time it spends at a loading or discharge port. The Baltic Dry Index is not a significant explana tory variable at the 5% level, while the oil index is significant at the
79 10%, but does not have the right sign. As expected the loading and discharge times, in models 3 and 4, are positively related to the cost of shipping. Given that the data used was fo r the year 2001, however its estimates are used to determine the cost of shipping for the year 2009. To make the 2001 estimated applicable to 2009, two options were considered. The first was to assume they remain unchanged. The second was to compute a fact or that measured how much the BDI and OIL indices have changed between 2001 and 2009, then multiply the regression estimates by the scaling factor. Multiplying the regression estimates from Equation 4 16 by the scaling factor (2.5) results in cost estimates that are higher in some cases almost double the rates reported by F.O. Licht 2 a leading authority on information on soft commodities, agriculture, food policy, markets and trade. For this reason, th e first option that assumed unchanging estimates over the decade is adopted. Commercial shipping companies were approached for shipping cost data when this model was being built. The information that they provided did not cover all the routes in the model hence the process of econometrically estimating shipping costs per ton was undertaken. However, the information they provided was useful in testing how accurate the prediction model was. Table 4 7 presents information for six routes from Brazil, Cuba, Gu atemala, India, South Africa and Thailand, were actual values are compared to what the model predicted. Based on this information it appears that the model under predicted the cost of shipping from Brazil by 15% on average. It appears to have been inaccur ate by about 18 and 38 percent for Guatemala and India, while being modestly accurate, with an error under 10% for sugar originating from South 2 http://www.agra net.com/portal2/home.jsp
80 Africa and Thailand. Overall it appears the model over predicts on average the cost of shipping by 7 percent. T he results of th e estimated transportation costs are presented in the form of a matrix and are posted in Table 4 8 Tariffs and s ubsidies Major sugar traders like the EU and the United States have complicated systems that allow sugar from a select group o f countries at preferential tariff rates. There are import quotas that the EU imposes on ACP countries and India. There are import quotas imposed by the USA as part of the TRQ system it uses to control sugar trade. imports from non preferential countries, while the USA imposes a tax of 1.4606 cents/KG for cane and 3.6606 cents/KG for beet. There is a long list of import taxes and export subsidies that other countries impose. D etail ed i ndividual country t rade policie s are presented in Table 4 9
81 Table 4 1. Demand e lasticities used in the model Author Estimated Various sources FAPRI Africa Group 1.066 0.015 7 Australia 0.002 Brazil 0.009 0.246 3 0.224 3 0.08 Canada 0.240 4 0.06 Caribbean Group 0.100 China 0.258; 0.228 Colombia 0.478 7 0.09 Cuba 0.040 5 0.160 5 0.09 European Union 0.584 0.120 6 0.320 6 Guatemala 0.016 7 0.11 India 0.063 0.14 Indonesia 0.085 Japan 0.810 0.344 3 0.260; 0.002 7 Malaysia 0.100 8 0.450 8 0.06 Mexico 0.019 Middle East 0.12 Pacific Group 0.040 9 Pakistan 0.08 Russia 0.19 South Africa 0.083 0.381 3 0.370 3 0.13 South Korea 0.389 3 0.256 3 0.09 Thailand 0.675 3 0.358 3 0.12 United States 0.140 10 ; 0.042 7 Table 4 2. Regression estimates for an acreage equation for Brazil Dependent variable Independent Variables Parameter Estimates Intercept 5.1165 (2.6624) Producer Price (US $/ton) 0.1262 (0.0537) Area allocated to production 0.6412 (0.1812) Trend 0.0144 (0.0057) Regression 0.9690 Durbin H test 0.4687 (0.3196) Source: wn computation. 3 Gemmill (1979). 4 Tweeten et al (1997). 5 Dye and Sicotte (2004). 6 Poonyth et al (2000). 7 Devadoss et al (1995). 8 http://www.fao.org/DOCREP/005/X0513E/x0513e05.htm 9 Hafi et al (1993). 10 Schmitz et al (2006). CAFTA and US Sugar.
82 Table 4 3. Supply e lasticities for different countries in the model Author Estimated Various sources FAPRI (Beet) FAPRI (Cane) Africa Group 1.742 0.017 7 Australia 0.066 7 0.14 Brazil 0.352 0.20 Canada 2.932 0.17 Caribbean Group 0.084 China 0.631; 0.196 0.13 0.09 Colombia 0.018 7 0.11 Cuba 0.08 European Union 0.180 0.60 0.31 Guatemala 0.008 7 0.10 India 1.854 0.978 7 0.21 Indonesia 2.068 0.320 7 0.08 Japan 0.336 7 0.22 0.18 Malaysia 0.480 0.09 Mexico 0.082 0.891 7 0.20 Middle East 0.06 0.06 Pacific Group 0.070 11 Pakistan 3.021 0.589 0.04 0.07 Russia 0.103 0.10 South Africa 0.461 0.047 7 0.12 South Korea 0.144 7 Thailand 0.162 0.17 United States 0.420 0.271 Table 4 4. Comparison of ship sizes Ship Classification Dead Weight Tons % of World Fleet % of Dry Bulk Traffic Capesize 100,000+ 10% 62% Panamax 60,000 80,000 19% 20% Supramax 45,000 59,000 37% 18% w/ Handysize Handysize 15,000 35,000 34% 18% w/ Supramax Source: Wikipedia 11 Reddy (2009).
83 Table 4 5. Determinant s of the size of load Estimate Standard Error Intercept 4,350** 1,494 Handymax 14,552*** 2,537 Capesize 64,534*** 6,083 Distance 3.3767*** 0.5779 Loading Rate 0.6554*** 0.1128 Discharge Rate 1.9747*** 0.3286 Square of distance 0.0002** 0.0001 Adjusted R Squared 0.5199 No of observations 337 Durbin Watson 2.005 **,*** indicate significance at the 95% and 99% level respectively Table 4 6. Determinant s of the cost of shipping a ton of sugar Model 1 Model 2 Model 3 Model 4 Model 5 Intercept 31.0491*** 28.4475*** 21.0283*** 29.5033*** 27.7389*** ( 2.7398 ) ( 2.7176) ( 2.7854 ) ( 2.8689 ) ( 1.3465) Distance 0.0020*** 0.0033*** 0.0042*** 0.0044*** 0.00429*** ( 0.0002 ) ( 0.0005 ) ( 0.0006 ) ( 0.0005 ) ( 0.0005) Size of load 0.0003*** 0.0003*** 0.0006*** 0.0009*** 0.0009*** ( 0.0000 ) ( 0.0000 ) ( 0.0000 ) ( 0.0001) ( 0.0001) Loading Rate 0.0005*** 0.0005*** 0.0003429*** 0.0004*** ( 0.0001 ) ( 0.0001 ) ( 0.0001 ) ( 0.0001) Discharge Rate 0.0022*** 0.0021*** 0.00137*** 0.00166*** ( 0.0003 ) ( 0.0003 ) ( 0.0004 ) ( 0.0003) Baltic Dry Index 0.0007 0.00231 ( 0.0022 ) ( 0.0020) OIL (NYMEX) 0.21191 0.1690 0.1163 0.2601* ( 0.1724 ) ( 0.0913 ) ( 0.0912 ) ( 0.1561 ) Squared distance 1.22E 07** 1.80E 07*** 1.97E 07*** 1.92E 7*** (0.0000) (0.0000) (0.000) ( 4.67E 8) Loading time 12 0.1706*** 0.07443 ( 0.0540 ) ( 0.0611) Discharge time 13 0.17538*** 0.04103 ( 0.0453 ) ( 0.0518 ) Squared size 1.09E 08*** 1.20E 8*** (0.0000) ( 1.53E 9) Adjusted R Squared 0.5207 0.5301 0.5276 0.6141 0.6072 D Watson 2.000 1.995 2.054 2.057 2.033 No of observations 298 298 298 298 298 *, **, *** indicate significance at the 90%, 95% and 99% level respectively. 12 Loading time is defined as the size of load over loading rate. 13 Discharge time is defined as size of load over the discharge rate.
84 Table 4 7. A comparison of the actual shipping values and predicted values for certain routes Africa group Australia Brazil Canada Caribbean group China Colombia Cuba EU Guatemala India Indonesia Japan Malaysia Mexico Pacific group Pakistan Russia South Africa South Korea Thailand UA E USA Actual Values Brazil 65 42 60 66 69 66 44 59 42 68 45 C uba 41 38 G uatemala 46 35 42 42 42 35 42 I ndia 24 18 18 21 25 South Africa 49 40 44 48 44 35 47 T hailand 25 18 27 18 42 25 Predicted Values Brazil 47 42 50 48 48 54 44 50 45 46 43 Cuba 49 36 43 Guat emala 48 43 48 47 55 47 46 Indi a 38 27 29 14 38 Sout h Africa 50 50 41 53 48 46 50 Thai land 24 17 30 21 45 26 Percent difference between actual and predicted Brazil 28 1 16 27 31 17 0 15 6 32 3 C uba 11 13 G uatemala 4 23 15 13 30 33 10 I ndia 57 49 63 35 54 S outh Africa 2 26 8 10 9 31 7 T hailand 3 3 12 15 8 4 Source: wn computation.
85 Table 4 8 Estimated transportation costs ($ per ton ) between countries African Group Australia Brazil Canada Caribbean Group China Colombia Cuba EU Guatemala India Indonesia Japan Malaysia Mexico Pacific G roup Pakistan Russia S. Africa S. Korea Thailand Middle East USA Africa n G roup 7 42 43 50 48 44 48 49 44 49 26 25 47 39 49 48 27 35 21 44 39 28 52 Australia 52 85 141 61 141 52 141 141 59 141 132 39 53 50 141 113 56 59 132 52 126 59 64 Brazil 42 51 10 42 34 47 38 39 42 43 50 48 48 54 44 50 50 45 38 46 51 51 43 Canada 46 250 250 10 250 47 250 250 30 250 250 47 50 55 250 250 50 39 250 47 250 51 18 Caribbean Group 46 50 33 26 9 48 15 17 35 25 50 47 50 54 25 48 50 41 45 48 48 51 26 China 42 250 250 50 250 8 250 250 48 250 250 26 17 32 250 250 43 47 250 11 250 45 52 Colombia 47 52 39 26 17 49 11 15 38 21 52 47 50 54 23 49 51 44 49 48 51 52 25 Cuba 47 51 39 23 18 49 14 10 36 22 51 47 50 54 19 49 51 43 49 48 49 51 21 EU 43 47 41 32 35 47 37 36 8 42 44 48 48 53 40 47 46 32 46 47 48 46 37 Guatemala 48 52 45 33 27 48 21 23 43 11 52 48 47 55 29 46 51 47 52 46 50 51 32 India 24 40 47 50 47 38 48 48 44 47 7 27 42 29 48 45 14 35 33 38 30 19 52 Indonesia 24 250 250 51 250 26 250 250 48 250 250 7 32 19 250 250 33 44 250 28 250 36 52 Japan 44 250 250 51 250 16 250 250 47 250 250 30 9 36 250 250 45 48 250 13 250 47 53 Malaysia 32 250 250 51 250 26 250 250 48 250 250 13 31 13 250 250 30 42 250 27 250 34 52 Mexico 48 52 45 28 27 49 23 20 41 28 52 45 50 52 11 51 51 46 52 48 48 52 26 Pacific Group 49 40 61 53 61 39 59 60 49 56 60 38 38 47 61 19 51 48 61 38 54 53 54 Pakistan 24 250 250 50 250 40 250 250 44 250 250 30 44 33 250 250 10 34 250 40 250 16 52 Russia 34 47 43 42 41 48 42 42 33 45 35 44 49 49 45 46 37 7 41 47 45 38 46 South Africa 23 71 65 54 74 50 76 76 50 78 63 41 53 48 78 78 40 46 32 50 72 41 56 South Korea 43 250 250 50 250 12 250 250 48 250 250 28 14 34 250 250 44 47 250 7 250 46 53 Thailand 37 36 48 50 46 24 47 46 49 46 31 17 30 21 45 40 35 45 42 26 8 39 50 Middle East 24 250 250 50 250 42 250 250 44 250 250 33 46 37 250 250 16 35 250 42 250 10 52 USA 47 250 250 17 250 47 250 250 33 250 250 47 50 55 250 250 51 41 250 47 250 51 11 Source: wn computation. The level of transport cost that is set to prevent shipping into a particular j region is $2 50.
86 Table 4 9 Sugar policies for countries in world sugar model 21 Tariff 22 Subsidy Africa Madagascar 5%; Mauritius 65% customs duty; Australia Ended administered price arrangements in 1989 and removed import tariffs in 1997. Free Brazil Imposes a 17.5% tariff on imports from non MERCOSUL countries (Brazil has zero imports). There is a subsidy (BRR 5.07/MT) targeting high cost growers in Northeast region. 16% CET Canada Imposes a tariff on refined imports from $22.05/ton to $30.86/ton depending on the polarization of sugar. Developing countries pay zero duty on raw sugar, and Australia and Cuba, from where the bulk of the raw sugar is imported, are exempt from duty. Free Caribbean China China charged as of 2004 an import duty rate of 50% to MFN and an in quota rate of 15% for both raw beet and cane sugar. The general tariff was 125% in 2004. This is the latest information available from ICTB website. 125% Colombia Sugar imports from th e Andean community are allowed duty free. The basic duty on raw and refined sugar imports from the non Andean Community is 20%. Export subsidies of 2.5% of the f.o.b. value for centrifugal and panela sugar is received by Colombian exporters. This is not pr ovided for exports to the United States. Colombia sets guaranteed sugar prices close to the world price. 20% 2.5% Cuba Imposed in 2008 an ad valorem tariff of 40% for raw sugar beet and cane imports. The rate is 30% for MFN. 40% European Union Export refunds are paid to exporters to cover the gap between the EU price and the world price when sugar is sold from intervention stocks. Production quotas are used to limit the sugar eligible for support. The surplus of A and B production above domestic consum ption is exported with subsidy. C quota sugar must be exported at world prices. Sugar imported from ACP is re exported with subsidy. The with in quota rate is EUR 98/ton and out of quota rate is EUR339/ton. Everything But Arms is limited by quotas until 20 09 when duty free access is allowed. EUR339/ton World Price less EU Price in 2009 Guatemala Raw beet or cane sugar imported are levied a 20% duty. 20% India Imposes an import duty of 60% plus INR 850/ton countervailing duty on raw sugar. National minimum support price for sugarcane (INR 620/ton in 2001/02) are augmented by state governments by another 20% to 50%. There is a transport subsidy to encourage exports (INR 140/ton in 2001/02). 60% Indonesia Imposes a tariff rate of 20% on raw cane suga incomes, the government also sets a sugar floor price (IDR 2,600/kg in 2001/02). 25% Japan Imposes a prohibitive duty on refined sugar of JPY 21.5/kg with an additional surcharge of JPY 53.88/kg. But imports of raw sugar are duty free. Free 21 M ost of the information on this is extracted from Elobeid and Beghin (2005) and International Customs Tariffs Bureau (http://w ww.bitd.org/) 22 Refers to tariffs on raw sugar for processing in mills to WSE.
87 Table 4 9 Continued. Tariff 23 Subsidy Malaysia Controls sugar imports through quota restrictions by licenses. The country imposes a 5% ad valorem rate on sugar imports as well as a specific tax of MYR 426.7/ton. Wholesale and retail sugar prices are controlled (MYR 1,345/ton for the wholesale price and MYR 1.4/kg for the retail price). 5% Middle East Iran imposes a tariff rate of 19% on raw sugar imports. Saudi Arabia free. Syria 1%; Lebanon 5%; Jordan free. Mexico Imposes a duty of $0.3166/kg on U.S. sugar imports and $0.3958/kg on third country imports. Every year the government announces the reference price for standard sugar, which is used to calculate the price paid to sugarcane growers. Growers are given 57% of the wholesale reference price of a ton of standard sugar (MX pesos 4,561.08/ton in 2001/02). Pacific Pakistan Imposes a 15% import tariff on raw and refined sugar. 15% Russia Seasonal tariffs are added during periods of peak domestic production to protect producers and support prices. The in quota tariff rate was 5% but no less than EUR 0.015/kg and the over quota rate was s et at 40% for raw and white sugar but no less than EUR 0.12/kg for raw sugar and EUR 0.14/kg for white sugar. The over quota seasonal rate was 50% but not less than EUR 0.15 /kg for raw sugar and EUR 0.18/kg for white sugar. South Africa Imposes duties based on the difference between the world price and a set reference price. The duty was ZAR 784/ton in 2001 and ZAR 1312/ton in July 2002. South Africa provides import access of sugar to Swaziland, Mozambique, Zambia, and Zimbabwe. Free South Korea Impos es a 3% tariff on raw sugar and a temporary 50% tariff on refined sugar. The wholesale sugar price is controlled by the government. 3% Thailand Maintains high internal sugar prices using quotas and import tariffs. The country has a 65% in quota tariff ra te and a 99% out of quota tariff rate. 99% USA Has an MFN import duty of 0.625/lb (raw value) but most quota suppliers are exempt. The above TRQ rate is 15.36¢/lb for raw sugar and 16.21¢/lb for refined sugar (TRQ was 1.361 million tons in 2001 and 1.289 million tons in 2002 and it was 1,117,195 tons i n 2009). Under NAFTA, Mexico has duty free access to the U.S. of up to 25,000 MTRV until 2008 when all imports from Mexico are duty free. The TRQ for countries in the model are: Africa 88115; Australia 87402; Brazil 152691; Caribbean 57802; Colombia 25273; Guatemala 50546; India 8424; Mexico 7258; Pacific 9477; South Africa 24220; Thailand 14743. Cane: 1.4606 c/KG Beet: 3.6606 c/KG 23 Refers to tariffs on raw sugar for processing in mills to WSE.
88 CHAPTER 5 RESULTS The results reported in this chapter are in line with events that occurred in the European Union (EU) with regard to the timeline they followed in liberalizing the EU sugar industry. The study shows how the rest of the world (especially ACP countries) is i mpacted under different EU policies. The study describes how changes in the [WSE] in metric tons) affected its trading partners. Five scenarios are presented. First, the 2009 baseline results are compared with those generated using 2006 values, the year in which reforms in the EU sugar policy were first introduced. Initially, the pric e floor is varied alone to understand its singular impact (i.e., whether the price floor alone was an effective policy tool). Second, the European Union is allowed to completely liberalize supply prices by removing all forms of sugar tariff and non tariff barriers. Third, after the European Union and the United States are allowed to simultaneously liberalize their trade policies, the impact on the rest of the world is quantified. Fourth, the study analyzes the impact of a US only liberalization and fifth al l countries are allowed to liberalize their supply prices by removing all forms of protection, and then the result is compared to the baseline model. Scenario I: The W orld S ugar M odel To recap, the most prevalent policy instrument in most regions that the model incorporates is the ad valorem tariff. Ad valorem tariffs are based on value of sugar as compared to the per unit taxes which are based on fixed quantities. In addition, other countries have other forms of protection, like in the United States, w h ere there is a tariff
89 rate quota (TRQ), which restricts imports to a select group of countries, with an over quota tariff pegged at 1.4606 US cents/KG for sugarcane and 3.6606 US cents/KG for sugar beets. In the European Union, alongside the price floor, which is a production support tool, there are quotas that restrict imports to a handful of countries, mostly ACP sugarcane, along with an export subsidy, equivalent to th e difference between world prices and the EU price floor. The export subsidy is pegged at $300/ton for the baseline model. Baseline model results for 2009 are pres ented for 23 regions in Table 5 1. They indicate that on the demand side, the European Union had the most expensive sugar and the highest supply price in the world, which was indicative of the high price floor in 2009. Also, markets clear in 11 of the 23 regions. With respect to production, the model predicted that India produced the most sugar, followed by Brazil the European Union, China, the United States, and Thailand, respectively. Regarding consumption, the model predicted as expected that India was the largest consumer in 2009, followed by the European Union, China, Brazil, the United Stat es, and Russia. EU policy dictated that only A & B sugars could be sold within its borders, implying that C sugar had to be exported and any imports re exported. Therefore, consumption in the model is capped at 17.4 million tons, as that was the amount equ ivalent to A & B production. Producer surplus was used to measure welfare, showing benefits accruing at $ 942 million for African farmers, $ 236 million for Caribbean farmers, and about $ 62 million for Fiji (the only country in the Pacific Group) farmers in 2009.
90 The accuracy of the model was evaluated (Table 5 2) by comparing the actual production and consumption values that happened in the base year with what the model predicted. It appears that on average the model was much more accurate in predict ing prod uction, which it under predicted by 3.3 percent on average, while consumption was under predicted on average by 12.1 percent. T he results presented in Table 5 3 show production response to changes in the price floor using the 2009 baseline model for the 2 006 to 2010 prices, which the effects of lowering the price floor alone are negligible. Supply prices change slightly in the positive direction, except for the United S tates where prices remain unchanged. Production changes are also negligible (less than one percent) for most countries in the model. Scenario II: Impact of EU liberaliz ation In scenario I, we considered what happened to the global sugar markets if the EU p rice floor alone was lowered. Scenario II considers the impact of completely liberalizing the European Union while trade distortions in other countries remain unchanged. According to the model, EU liberalization drives the supply price in Europe down by 5 3 %, to $3 68.18 per ton (Table 5 4 ). The Caribbean countries and the United States are the two other regions where prices fall, by 3 0.6 % and 1.8 %, respectively, as a result of the EU action. In the rest of the world, supply prices rise, ranging from 8.5% in the Middle East to about 31.2 % in Brazil. It is noteworthy that prices go up by 13.2% in Africa and by 26.4% in the Pacific (Fiji), two regions that are part of the EU/ACP sugar protocol. The implication is that African and Pacific farmers would benefit fr om EU liberalization while those in the Caribbean would not.
91 The model predicts that allowing only the European Union to liberalize results in global production increasing by an average of 5.8 % per year. However, in the EU, production would fall by 9.4%, while US production would decline by just under 1%. The Caribbean records a 3.1% drop in this scenario. The effects on consumption are mostly felt in Europe, where liberalization results in consumption going up 61 .2 %. In the baseline model, we imposed a bi nding consumption constraint on Europe to limit it to 17.4 million tons of A plus B sugar. Lifting this constraint has a huge impact on consumer surplus. With liberalization, the European Union would cease being an exporter to become a large net importer, increasing EU imports of sugar by about 80 9 %. Prior to the 2006 reforms, the European Union imported under 2 million tons of sugar per year, mostly from ACP countries and India. Liberalization, however, allows more imported sugar from Brazil, Australia, a nd Thailand, countries that previously did not have preferential access to the EU market. With regards to the welfare effects, there is interest in understanding how policy changes would affect ACP countries. It appears that while liberalization would resu lt in the producer surplus increasing by 2 2.9 % and 3 0.6 % for Africa and the Pacific (Fiji), respectively, it would decrease by a significant 3 2.6 % for the Caribbean countries. This could be an indication that Caribbean sugar revenues are highly dependent o n EU policies, while the same cannot be said about Africa or Fiji. African countries in the model produced about 3. 3 million tons of sugar under the baseline model, of which only 0.8 million tons was traded internationally, mainly as part of the EU/ACP su gar protocol. The protocol absorbed about 25% of the African sugar; hence, it is understandable that no matter how important the EU market
92 is to African exports, the amounts exported are not enough to cause African farmers to suffer as a result of liberali zation, unlike with the Caribbean farmers. Lastly, the increase in global prices is felt by consumers across the globe as consumer surplus drops in most countries, except the European Union, the Caribbean, and United States where it increased by 159 8 %, 6. 6%, and 0.3%, respectively. Scenario III: Impact of EU and US liberalization When both the European Union and the United States are assume to co mpletely liberalize (Table 5 5 ), supply prices decline by 50% in the European Union, by 2 5 .8% in the United S ta tes, and by 30.5 % in the Caribbean. In contrast, supply prices increase in the rest of the world, ranging from 10 .4 % in China to 30.8 % in Brazil. It is interesting to note that production declines by close percentage levels in the European Union (9. 4 %) and the United States (8. 4 %). Therefore, when these two major producers and consumers of sugar liberalize, global s ugar production increases by 5.8 %, with much of that accounted for by India (1 8.8 %), Malaysia (14. 4 %), and South Africa (11. 7 %). Percentage p roduction changes for other countries are in the single digits. Once more, liberalization results in a 61 % increase in sugar consumption in the European Union, but only a 1.9 % increase in the United States. Allowing these two countries to liberalize their sugar supply prices also makes them net importers, which benefits US and EU consumers. Scenario IV : Impact o f US Onl y Liberalization If the United States liberalizes its sugar markets while the rest of the world remains unchanged global sugar prices woul d increase for 20 of the 23 regions in the model by about two to three percent for most countries as reported in Table 5 6. Supply prices would remain unc hanged in China and drop by 4.4% in Colombia, while the
93 United States would face a 38.9% decline in su pply prices as a result of the liberalization policy. The effects on production are relatively small, less than one percent in most countries, while in the United States produc tion would decline by almost 13% United States suga r imports would increase by 115% as a result of liberalization, while producer surplus drops by 43.6% and consumer surplus ticks up by almost six percent. It appears from the model that a United States only liberalization has a modest impact on ACP countries with their producer surpl us increasing by between 1.35 % in the Caribbean, 2.19% in the African region, and 2.91% in the Pacific region respectively. Most of the gains to United States liberalization would be realized by American consumers who would pay lower sugar prices, and glob al exporters who would find a bigger market to sell to in the United States. Scenario V: Impact o f Free Trade Across All Countries The effect of global liberalization on supply prices is that prices go up in 20 of the 2 3 regions of the model (Table 5 7 ). T he three regions experiencing declines are the Caribbean, the European Union, and the United States, where prices fall by 2 6. 7%, 47. 6 %, and 2 2.2 %, respectively. The net effect is that liberalization drives up supply prices by 2 2 9 % on average. Under free t rade liberalization spurs global production by 8. 8 %. Specifically production would go up by 4. 8 % in Africa, 11. 4 % in the Pacific, and 2 5 4 % in India. Regions that experienced a drop in supply prices, the Caribbean, the European Union, and the United State s, also experience production declines of 2.6%, 8.9%, and 7. 2 %, respectively. The average consumption increases across all countries are driven mostly by increases in the European Union (60 %), Thailand (1 6 %), and China (14.8%), respectively. Most countries experience slight consumption declines, but
94 they are not large enough to offset the consumption increases of the European Union, Thailand, and China. In regards to trade, the most notable aspect of liberalization is that EU imports of sugar would increas e from less than 2 million tons per year to about 13.4 million per year to satisfy the increase in demand in the European Union. A similar trend is experienced in the United States, where imports increase by 6 5.9 %, compared to the baseline. Based on producer surplus, while liberalization would be good for Africa, whose surplus would increase by 3 0.7 %, to $ 1,232 million, and for the Pacific, where it would increase by 46.1% to $ 91. 3 million, it would be bad for the Caribbean, whose welfare (producer s urplus) would decline by 2 7.8 %, to $ 170.7 million. Impact o f Trade Liberalization o n ACP Countries During the ACP sugar protocol years, each member country was allocated a quota which allowed its sugar duty free access to the European Union. ACP countries were allocated 701 thousand, 428 thousand, and 165 thousand metric tons for the African, Caribbean, and Pacifi c groups, respectively. Table 5 8 reports the benefits of liberalization, as measured by the producer surplus, for three regions of the ACP. It a ppears liberalization would be beneficial for ACP countries at least when looked at in the aggregate using producer surplus. Prior to liberalization, each country got paid based on the proportion of the quota amounts they supplied to the European Union. T o illustrate th is point, column (3) of Table 5 9 indicates that the Congo had a 1.5% share of the quota, so it got paid 1.5% of the ACP protocol revenues, assuming it fulfilled its quota for that year. With liberalization, quotas ceased to exist and produc er gains for each country were determined by other factors, such as each country compared with other countries in
95 producing sugar in the post liberalization period. By liberalizing the world markets, sugar producers in less developed countries would have t o compete with Brazil, Thailand, and Australia who tend to be low cost producers. To arrive at a more accurate country specific prediction of the impacts of liberalization, supply and demand elasticities for each ACP country are needed. Since this informat ion was challenging to estimate, the study used regional groupings to circumvent the data challenges. Despite these challenges, some predictions can still be made about how individual countries are likely to fare in the post li beralization period. In Table 5 9 the fourth column presents total supply available in each country in 2006. The fifth column shows total exports by each of these countries in the same year. The sixth column presents exports as a share of total supply and the seventh column shows the proportion of exports that went to the European Union in 2006. In the African region, Mauritius exported 71%, followed by Swaziland (16%), the Congo (10%), and the rest of the countries (6% or less each) of their total sugar supply to the European Union. Based on these figures, it is highly likely that Mauritius at 71% would experience more significant revenue drop than would the Congo at 10% as a result of liberalization. The model predicts that the Caribbean countries would experience negative producer surplus changes as liberalization occurs. Their exports shares to the European Union are on average higher than most ACP countries, so a 36% drop in support prices would affect them negatively. On the other hand, Fiji, which exported 37% of its production to the European Union, has a positive gain in producer surplus after liberalization. The possible explanation for this is that Fiji has proximal sugar markets (e.g., China and Japan) that are capable of absorbing its production.
96 There is a general consens us that EU and US subsidies depress global sugar prices to the detriment of farmers in less developed countries. According to the International Sugar Organization (ISO, 2006) sugar production has been growing in less developed countries, especially in Sub Saharan Africa, in response to incentives like the Some countries that have experienced high sugar production growth are Sudan, Ethiopia, Tanzania, Mozambique, Zambia, Mala wi, and Uganda, with the last four being part of the sugar protocol. Some of the factors that have shifted production in their favor are high sugar yields, long milling campaigns (8 to 19 months), low labor and transportation costs, and low processing cost s (Nyberg, 2005). Orden (2008), in an IFPRI discussion paper, ranks the lowest cost sugar producers in the 2004 05 period, with Brazil as the least costly producer, followed by Malawi, Zimbabwe, Australia, Swaziland, Zambia, Guatemala, South Africa, China, Thailand, Mozambi que, Tanzania, India, and Mexico, respectively. This model indicates that if only the European Union liberalized its sugar markets, global consumption would go up 5.8 %, and EU sugar consumption would increase from 17 million tons to about 2 8 million tons p er year, or a 61.2 % increase. Global sugar consumption would increase 8.8 % in the free market scenario due to consumption increases of 60 % in the European Union, 16 % in Thailand, and 14.8% in China, respectively. To meet this demand, sugar production would have to increase in some parts of the world, and it is reasonable to argue that the lowest cost producers would respond positively. For that reason, it is plausible that the supply price in the African region would increase as more demand is anticipated, thus pushing production upwards. According to Conforti et al. (2007), losers
97 from EU trade reforms would tend to be (a) high cost producers such as the Caribbean countries, and (b) countries such as Swaziland and Mauritius who were not eligible for expande d access under the EBA at the time. This is consistent with what this trade model is predicting.
98 Table 5 1. The baseline model PRICE PRODUCTION CONSUMPTION TRADE WELFARE Supply Price US$/Ton Demand Price US$/Ton Production ('000) MT Consumption ('000) MT Imports ('000) MT Exports ('000) MT Producer Surplus $Million Consumer Surplus $Million Africa n G roup 318.14 318.14 3,246.2 5,211.8 2,752.0 789.1 942.44 2,701.14 Australia 243.23 243.23 5,355.0 1,414.1 3,939.4 1,181.03 24,995.10 Brazil 248.60 288.37 27,812.0 13,538.0 14,276.0 5,908.55 12,405.20 Canada 318.12 318.12 105.0 1,849.0 1,746.5 28.33 5,122.24 Caribbean G roup 493.23 493.23 503.1 342.0 267.0 428.0 235.99 818.64 China 315.71 710.35 14,114.0 14,114.0 3,862.50 14,144.90 Colombia 286.29 343.55 2,269.7 2,244.4 25.3 629.29 2,138.21 Cuba 267.42 374.38 1,294.1 790.2 503.5 334.16 936.68 EU 789.29 1,558.97 15,933.0 17,440.0 1,507.0 11,399.20 16,640.00 Guatemala 258.19 309.82 2,256.3 825.7 1,430.5 579.78 2,148.18 India 280.72 449.15 28,001.0 25,412.0 2,588.4 4,004.64 88,794.40 Indonesia 309.99 387.48 2,229.3 5,030.8 2,800.0 557.72 25,308.80 Japan 324.26 324.26 699.6 4,365.9 3,661.9 209.34 7,570.44 Malaysia 314.75 330.48 78.3 1,642.1 1,560.0 16.38 972.08 M iddle East 327.32 327.32 1,270.1 6,934.3 5,670.0 408.14 47,364.00 Mexico 314.45 314.45 5,349.1 6,562.5 1,817.0 604.0 1,543.76 16,799.50 Pacific G roup 258.90 258.90 285.2 63.9 221.4 62.48 658.83 Pakistan 322.14 370.46 3,630.6 4,653.3 1,020.0 1,104.83 7,000.85 Russia 320.96 320.96 3,024.5 7,125.5 4,100.0 878.16 14,594.20 S outh Africa 267.47 267.47 2,386.0 2,169.6 216.2 492.41 2,396.38 S outh Korea 320.01 329.61 0.0 1,536.1 1,540.0 0.00 1,439.30 Thailand 266.24 529.81 7,030.6 2,542.5 4,484.7 1,677.04 1,334.05 USA 534.08 534.08 7,729.2 8,794.1 1,065.2 3,456.13 32,373.50 334.76 421.85 134,601.9 134,601.8 29,506.7 29,506.7
99 Table 5 2. Accuracy of prediction of the model Actual Predicted Consumption Production Consumption Production % Consumption Difference % Production Difference African Group 3,296.9 3,558.8 5,211.8 3,246.2 58.1 8.8 Australia 1,384.1 5,065.5 1,414.1 5,355.0 2.2 5.7 Brazil 12,574.7 30,437.5 13,538.0 27,812.0 7.7 8.6 Canada 1,559.1 104.3 1,849.0 105.0 18.6 0.6 Caribbean Group 327.3 496.0 342.0 503.1 4.5 1.4 China 15,195.4 12,879.0 14,114.0 14,114.0 7.1 9.6 Colombia 1,792.3 2,328.8 2,244.4 2,269.7 25.2 2.5 Cuba 781.7 1,300.0 790.2 1,294.1 1.1 0.5 European Union 20,004.9 15,795.0 17,440.0 15,933.0 12.8 0.9 Guatemala 783.9 2,255.0 825.7 2,256.3 5.3 0.1 India 25,442.0 24,125.0 25,412.0 28,001.0 0.1 16.1 Indonesia 4,820.1 2,013.3 5,030.8 2,229.3 4.4 10.7 Japan 2,637.6 923.0 4,365.9 699.6 65.5 24.2 Malaysia 1,448.6 48.0 1,642.1 78.3 13.4 63.0 M iddle East 6,339.7 1,309.5 6,934.3 1,270.1 9.4 3.0 Mexico 5,709.5 5,587.3 6,562.5 5,349.1 14.9 4.3 Pacific Group 62.2 282.0 63.9 285.2 2.7 1.1 Pakistan 4,544.5 3,471.8 4,653.3 3,630.6 2.4 4.6 Russia 6,457.0 3,082.8 7,125.5 3,024.5 10.4 1.9 South Africa 1,761.3 2,404.5 2,169.6 2,386.0 23.2 0.8 South Korea 1,300.4 0.0 1,536.1 0.0 Thailand 2,284.3 6,643.8 2,542.5 7,030.6 11.3 5.8 United States 8,273.7 7,151.0 8,794.1 7,729.2 6.3 8.1 Average Difference 12.1 3.3
100 Table 5 3 The baseline model compared with years 2006 and 2010 Effect of Supply Prices Effect on Production 2009 Supply Price US$/Ton 2006 Supply Price US$/Ton 2010 Supply Price US$/Ton 2006 to 2009 % Change 2009 to 2010 % Change 2009 Production ('000) MT 2006 Production ('000) MT 2010 Production ('000) MT 2006 to 2009 % Change 2009 to 2010 % Change Africa n G roup 318.14 318.97 317.08 0.261 0.003 3,246.2 3,247.6 3,244.3 0.043 0.059 Australia 243.23 244.06 242.17 0.341 0.004 5,355.0 5,358.4 5,350.6 0.063 0.082 Brazil 248.60 249.43 247.54 0.334 0.004 27,812.0 27,839.0 27,777.0 0.097 0.126 Canada 318.12 318.95 317.06 0.261 0.003 105.0 105.0 104.9 0.076 0.095 Caribbean G roup 493.23 494.06 492.17 0.168 0.002 503.1 503.2 503.0 0.018 0.020 China 315.71 315.71 315.71 0.000 0.000 14,114.0 14,114.0 14,114.0 0.000 0.000 Colombia 286.29 286.29 286.29 0.000 0.000 2,269.7 2,269.7 2,269.7 0.000 0.000 Cuba 267.42 268.25 266.36 0.310 0.004 1,294.1 1,294.4 1,293.8 0.023 0.023 EU 789.29 790.12 601.98 0.105 0.237 15,933.0 15,773.0 16,136.0 1.004 1.274 Guatemala 258.19 259.02 257.13 0.321 0.004 2,256.3 2,256.4 2,256.3 0.004 0.000 India 280.72 281.55 279.66 0.296 0.004 28,001.0 28,082.0 27,897.0 0.289 0.371 Indonesia 309.99 310.82 308.93 0.268 0.003 2,229.3 2,231.6 2,226.4 0.103 0.130 Japan 324.26 325.09 323.20 0.256 0.003 699.6 699.9 699.3 0.040 0.050 Malaysia 314.75 315.58 313.69 0.264 0.003 78.3 78.4 78.1 0.177 0.225 M iddle East 327.32 328.15 326.26 0.254 0.003 1,270.1 1,270.2 1,269.9 0.008 0.016 Mexico 314.45 315.28 313.39 0.264 0.003 5,349.1 5,351.5 5,346.2 0.045 0.054 Pacific G roup 258.90 259.73 257.84 0.321 0.004 285.2 285.5 284.9 0.102 0.126 Pakistan 322.14 322.97 321.08 0.258 0.003 3,630.6 3,631.7 3,629.3 0.030 0.036 Russia 320.96 321.79 319.90 0.259 0.003 3,024.5 3,025.9 3,022.5 0.046 0.066 S outh Africa 267.47 268.30 266.41 0.310 0.004 2,386.0 2,389.4 2,381.7 0.142 0.180 S outh Korea 320.01 320.84 318.95 0.259 0.003 0.0 0.0 0.0 0.022 0.022 Thailand 266.24 267.07 265.18 0.312 0.004 7,030.6 7,035.2 7,024.8 0.065 0.082 USA 534.08 534.08 534.08 0.000 0.000 7,729.2 7,729.2 7,729.2 0.000 0.000
101 Table 5 4. Impact of EU liberalization Supply Price US$/Ton % Price Change Production ('000) MT % Production Change Consumption ('000) MT % Consumption Change Imports ('000) MT % Imports Change Exports ('000) MT % Exports Change Producer Surplus $Million % PS Change Consumer Surplus $Million % CS Change African Group 360.18 13.2 3,321.2 3.6 5,000.4 6.3 1,765.0 41.4 88.1 88.8 1,080.49 22.85 2,486.47 12.18 Australia 312.18 28.3 5,638.1 4.9 1,411.3 0.2 4,230.4 6.7 1,560.01 29.96 24,897.70 0.36 Brazil 326.18 31.2 30,336.0 8.6 12,873.0 4.7 17,483.0 21.3 8,164.11 36.07 11,216.80 9.08 Canada 368.18 15.7 110.0 7.0 1,832.3 1.3 1,727.4 2.1 33.71 28.17 5,030.10 2.63 Caribbean Group 342.18 30.6 488.0 3.1 352.8 3.2 100.0 135.2 68.6 161.13 32.58 871.12 6.59 China 350.18 10.9 14,524.0 2.5 13,568.0 3.3 956.0 4,356.02 11.01 13,071.60 6.57 Colombia 342.18 19.5 2,297.6 1.1 2,165.4 3.2 132.3 389.9 756.92 18.68 1,990.34 6.38 Cuba 345.18 29.1 1,320.1 1.9 753.9 4.3 566.0 11.8 435.80 28.72 852.63 8.49 EU 368.18 53.4 14,342.0 9.4 28,323.0 61.2 14,000.0 809.1 5,024.57 52.97 43,887.10 159.78 Guatemala 335.18 29.8 2,262.8 0.3 811.1 1.7 1,453.5 1.4 753.74 28.63 2,072.57 3.36 India 336.18 19.8 33,428.0 17.7 25,089.0 1.2 8,340.4 202.7 5,708.06 38.51 86,553.80 2.30 Indonesia 351.18 13.3 2,343.6 8.0 5,005.1 0.8 2,660.0 7.9 651.90 26.81 25,050.50 1.60 Japan 365.18 12.6 713.3 3.1 4,314.4 1.9 3,598.0 2.8 238.25 21.88 7,392.85 3.68 Malaysia 356.18 13.2 85.2 13.7 1,581.7 5.7 1,500.0 6.4 19.76 32.93 901.96 11.11 Middle East 355.18 8.5 1,274.0 0.6 6,920.2 0.4 5,650.0 0.7 443.58 15.79 47,171.00 0.74 Mexico 364.18 15.8 5,488.2 3.8 6,498.8 1.4 1,616.0 16.5 604.0 0.0 1,813.24 25.95 16,474.70 2.85 Pacific Group 327.18 26.4 308.4 7.7 63.6 0.3 244.5 10.0 82.74 30.57 654.48 0.62 Pakistan 350.18 8.7 3,665.6 1.7 4,603.4 1.9 938.0 14.7 1,207.12 16.77 6,851.60 3.83 Russia 371.18 15.6 3,114.7 4.4 7,038.1 1.8 3,924.0 6.4 1,032.32 25.86 14,238.50 3.56 South Africa 337.18 26.1 2,670.1 11.1 2,101.1 2.9 569.2 152.6 668.65 33.34 2,247.52 5.81 South Korea 361.18 12.9 0.0 1.5 1,501.4 3.5 1,501.0 3.8 0.00 21.10 1,374.90 6.96 Thailand 335.18 25.9 7,409.5 5.0 2,210.1 12.2 5,199.7 14.9 2,174.79 27.71 1,008.05 22.96 USA 524.30 1.8 7,683.1 0.6 8,805.8 0.1 1,123.0 5.4 3,380.70 2.18 32,459.70 0.27 355.8374 6.3 142,823.4 5.8 142,824.0 5.8 40,002.4 32.5 40,002.4 32.5
102 Table 5 5. Impact of simultaneous EU and USA liberalization Supply Price US$/Ton % Price Change Production ('000) MT % Production Change Consumption ('000) MT % Consumption Change Imports ('000) MT % Imports Change Exports ('000) MT % Exports Change Produce r Surplus $Million % PS Change Consumer Surplus $Million % CS Change African Group 386.46 21.5 3,368.1 3.8 4,868.3 6.6 1,500.0 45.5 100.0 1,168.39 23.97 2,356.79 12.75 Australia 3 10.86 27.8 5,632.7 5.2 1,411.4 0.2 4,218.0 7.1 1,552.59 31.46 24,899.50 0.38 Brazil 325.22 30.8 30,305.0 9.0 12,881.0 4.9 17,381.0 21.7 8,135.05 37.68 11,231.10 9.46 Canada 394.74 24.1 112.6 7.3 1,823.4 1.4 1,710.0 2.1 36.66 29.42 4,981.54 2.75 Caribbean Group 342.91 30.5 488.1 3.0 352.8 3.1 100.0 135.0 68.5 161.49 31.57 870.86 6.38 China 348.62 10.4 14,506.0 2.8 13,593.0 3.7 913.0 4,333.40 12.19 13,119.20 7.25 Colombia 344.10 20.2 2,298.6 1.3 2,162.7 3.6 136.0 438.1 761.34 20.98 1,985.35 7.15 Cuba 347.35 29.9 1,320.8 2.1 752.9 4.7 568.0 12.8 438.66 31.28 850.34 9.22 EU 394.70 50.0 14,443.0 9.4 28,081.0 61.0 13,600.0 802.5 5,406.28 52.57 43,139.10 159.25 Guatemala 336.53 30.3 2,262.9 0.3 810.8 1.8 1,452.0 1.5 756.80 30.53 2,071.25 3.58 India 334.47 19.1 33,261.0 18.8 25,099.0 1.2 8,162.0 215.3 5,651.10 41.11 86,622.40 2.45 Indonesia 377.62 21.8 2,417.0 8.4 4,988.6 0.8 2,570.0 8.2 714.84 28.17 24,885.30 1.67 Japan 391.89 20.9 722.2 3.2 4,280.8 1.9 3,563.0 2.7 257.42 22.97 7,278.05 3.86 Malaysia 382.38 21.5 89.5 14.4 1,543.6 6.0 1,450.0 7.1 22.05 34.65 858.97 11.64 Middle East 381.07 16.4 1,277.7 0.6 6,907.0 0.4 5,630.0 0.7 476.61 16.78 46,992.00 0.79 Mexico 392.79 24.9 5,568.2 4.1 6,462.1 1.5 894.0 50.8 100.0 1,971.41 27.70 16,289.30 3.04 Pacific Group 326.53 26.1 308.2 8.0 63.6 0.3 245.0 10.7 82.54 32.12 654.52 0.65 Pakistan 375.89 16.7 3,697.7 1.8 4,557.7 2.1 860.0 15.7 1,301.78 17.83 6,716.16 4.07 Russia 397.58 23.9 3,162.1 4.5 6,992.2 1.9 3,824.0 6.7 1,115.18 26.99 14,053.30 3.71 South Africa 335.79 25.5 2,664.5 11.7 2,102.5 3.1 562.0 159.9 664.94 35.04 2,250.44 6.09 South Korea 387.64 21.1 0.0 1.6 1,479.0 3.7 1,479.0 4.0 0.00 22.14 1,334.28 7.30 Thailand 333.87 25.4 7,402.3 5.3 2,216.4 12.8 5,186.0 15.6 2,165.10 29.10 1,013.81 24.01 USA 396.44 25.8 7,080.7 8.4 8,958.5 1.9 1,878.0 76.3 2,436.87 29.49 33,595.30 3.77 362.85 8.4 142,388.9 5.8 142,388.3 5.8 32.0 32.0
103 Table 5 6. Impact of US only liberalization Supply Price US$/Ton % Price Change Production ('000) MT % Production Change Consumption ('000) MT % Consumption Change Imports ('000) MT % Imports Change Exports ('000) MT % Exports Change Producer Surp lus $Million % PS Change Consumer Surplus $Million % CS Change Africa n G roup 324.48 2.0 3,257.5 0.3 5,179.9 0.6 2,619.0 4.8 701.0 11.2 963.08 2.19 2,668.16 1.22 Australia 249.57 2.6 5,381.0 0.5 1,413.8 0.0 3,967.0 0.7 1,215.09 2.88 24,986.10 0.04 Brazil 254.94 2.6 28,018.0 0.7 13,484.0 0.4 14,539.0 1.8 6,085.70 3.00 12,305.70 0.80 Canada 324.46 2.0 105.6 0.6 1,846.9 0.1 1,740.0 0.4 28.99 2.36 5,110.51 0.23 Caribbean G roup 499.57 1.3 503.8 0.1 341.6 0.1 266.0 0.4 428.0 0.0 239.18 1.35 816.47 0.26 China 315.71 0.0 14,114.0 0.0 14,114.0 0.0 3,862.50 0.00 14,144.90 0.00 Colombia 273.82 4.4 2,263.5 0.3 2,262.0 0.8 1.4 94.4 601.04 4.49 2,171.92 1.58 Cuba 277.07 3.6 1,297.4 0.3 785.7 0.6 512.0 1.7 346.67 3.74 926.03 1.14 EU 795.63 0.8 15,957.0 0.2 17,440.0 0.0 1,483.0 1.6 11,500.40 0.89 16,640.00 0.00 Guatemala 266.25 3.1 2,257.0 0.0 824.2 0.2 1,432.0 0.1 597.98 3.14 2,140.19 0.37 India 287.06 2.3 28,622.0 2.2 25,375.0 0.1 3,240.0 25.2 4,184.30 4.49 88,536.60 0.29 Indonesia 316.33 2.0 2,246.9 0.8 5,026.9 0.1 2,780.0 0.7 571.92 2.55 25,269.00 0.16 Japan 330.60 2.0 701.8 0.3 4,357.9 0.2 3,659.6 0.1 213.79 2.12 7,542.77 0.37 Malaysia 321.09 2.0 79.3 1.4 1,632.8 0.6 1,550.0 0.6 16.88 3.05 961.17 1.12 M iddle East 333.66 1.9 1,271.0 0.1 6,931.1 0.0 5,660.0 0.2 416.20 1.98 47,320.00 0.09 Mexico 322.51 2.6 5,371.7 0.4 6,552.2 0.2 1,180.0 35.1 100.0 1,587.00 2.80 16,746.60 0.31 Pacific G roup 265.24 2.4 287.4 0.8 63.8 0.0 223.6 1.0 64.29 2.91 658.43 0.06 Pakistan 328.48 2.0 3,638.5 0.2 4,642.0 0.2 1,000.0 2.0 1,127.89 2.09 6,966.93 0.48 Russia 327.30 2.0 3,035.9 0.4 7,114.5 0.2 4,080.0 0.5 897.39 2.19 14,549.00 0.31 S outh Africa 273.81 2.4 2,411.9 1.1 2,163.4 0.3 249.0 15.2 507.64 3.09 2,382.63 0.57 S outh Korea 326.35 2.0 0.0 0.2 1,530.8 0.3 1,530.0 0.6 0.00 2.06 1,429.28 0.70 Thailand 272.58 2.4 7,065.5 0.5 2,511.9 1.2 4,550.0 1.5 1,721.76 2.67 1,302.14 2.39 USA 326.16 38.9 6,749.7 12.7 9,042.5 2.8 2,295.4 115.5 1,950.90 43.55 34,227.90 5.73 330.99 0.02 134,636.3 0.0 134,636.9 0.0 1.1 1.1
104 Table 5 7 Free world trade model Supply Price US$/Ton % Price Change Production ('000) MT % Production Change Consumption ('000) MT % Consumption Change Imports ('000) MT % Imports Change Exports ('000) MT % Exports Change Producer Surplus $Million % PS Change Consumer Surplus $Million % CS Change African Group 405.29 27.4 3,401.7 4.8 4,773.6 8.4 1,372.0 50.1 100.0 1,232.14 30.74 2,265.99 16.11 Australia 338.68 39.2 5,746.9 7.3 1,410.3 0.3 10.2 1,710.91 44.87 24,860.30 0.54 Brazil 344.05 38.4 30,918.0 11.2 13,127.0 3.0 4,340.0 24.7 8,711.61 47.44 11,662.80 5.98 Canada 413.57 30.0 114.5 9.1 1,817.1 1.7 1,700.0 2.7 17,798.0 38.80 36.98 4,947.26 3.42 Caribbean Group 361.74 26.7 490.0 2.6 351.4 2.7 100.0 67.5 170.70 27.67 864.23 5.57 China 413.80 31.1 15,282.0 8.3 16,202.0 14.8 920.0 139.0 5,304.26 37.33 18,639.90 31.78 Colombia 362.93 26.8 2,308.0 1.7 2,221.6 1.0 242.0 804.72 27.88 2,094.92 2.02 Cuba 366.18 36.9 1,327.1 2.6 793.0 0.3 86.4 6.0 463.60 38.74 943.17 0.69 EU 413.53 47.6 14,514.0 8.9 27,908.0 60.0 13,400.0 789.2 534.0 5,678.98 50.18 42,611.90 156.08 Guatemala 355.36 37.6 2,264.4 0.4 818.5 0.9 1.1 799.44 37.89 2,110.74 1.74 India 353.30 25.9 35,104.0 25.4 25,760.0 1.4 1,446.0 261.2 6,294.93 57.19 91,246.60 2.76 Indonesia 405.44 30.8 2,494.3 11.9 5,021.8 0.2 2,530.0 9.6 9,350.0 783.17 40.42 25,218.60 0.36 Japan 419.71 29.4 731.5 4.5 4,245.7 2.8 3,515.0 4.0 277.64 32.63 7,159.42 5.43 Malaysia 410.20 30.3 94.2 20.3 1,531.5 6.7 1,440.0 7.7 24.61 50.25 845.58 13.01 Middle East 399.90 22.2 1,280.4 0.8 6,897.5 0.5 5,620.0 0.9 500.70 22.68 46,862.00 1.06 Mexico 411.62 30.9 5,620.9 5.1 6,438.0 1.9 817.0 55.0 100.0 2,076.78 34.53 16,167.80 3.76 Pacific Group 354.35 36.9 317.6 11.4 63.6 0.5 14.7 91.25 46.06 652.75 0.92 Pakistan 394.72 22.5 3,721.2 2.5 4,615.7 0.8 895.0 12.3 254.0 1,371.65 24.15 6,888.39 1.61 Russia 416.41 29.7 3,196.0 5.7 6,959.5 2.3 3,770.0 8.0 1,175.05 33.81 13,921.90 4.61 South Africa 354.62 32.6 2,741.2 14.9 2,084.0 3.9 203.9 715.85 45.38 2,211.01 7.74 South Korea 415.46 29.8 0.0 2.2 1,465.7 4.6 1,470.0 4.5 657.0 0.00 31.35 1,310.44 8.95 Thailand 361.69 35.9 7,555.2 7.5 2,949.8 16.0 2.8 2,373.20 41.51 1,795.73 34.61 USA 415.27 22.2 7,169.5 7.2 8,936.0 1.6 1,767.4 65.9 4612 2,571.07 25.61 33,426.80 3.25 386.43 15.4 146,392.5 8.8 146,391.2 8.8 32.9 32.9
105 Table 5 8. Producer surplus comparison under different models Baseline Model EU Only Liberalization US Only liberalization EU & USA Liberalize Free World Trade African Region 942.44 1,080.49 963.08 1,168.39 1,232.14 Caribbean Region 235.99 161.13 239.18 161.49 170.70 Pacific Region 62.48 82.74 64.29 82.54 91.25 Total 1,240.90 1,324.36 1,266.55 1,412.42 1,494.09
106 Table 5 9 ACP sugar protocol quotas pre reform and exports in 2006 EU Quota ('000MT) Percent of Quota Total Supply ('000MT) Exports ('000MT) Exports / Total Supply Share Exported to EU Congo (Brazzaville) 10.20 1.50 106 41 0.39 0.10 Cote d'Ivoire 10.20 1.50 254 7 0.03 0.04 Kenya 0.00 777 11 0.01 0.00 Madagascar 10.80 1.50 276 10 0.04 0.04 Malawi 20.80 3.00 341 70 0.21 0.06 Mauritius 491.00 70.00 689 548 0.80 0.71 Mozambique 530 220 0.42 0.00 Swaziland 117.80 16.80 729 320 0.44 0.16 Tanzania 10.20 1.50 553 20 0.04 0.02 Zambia 0.00 300 140 0.47 0.00 Zimbabwe 30.20 4.30 469 149 0.32 0.06 Sub Total 701.20 100.00 5,024 1,536 0.31 0.00 Barbados 50.30 11.80 58 32 0.55 0.87 Belize 40.40 9.40 225 114 0.51 0.18 Guyana 159.40 37.20 243 161 0.66 0.66 Jamaica 118.70 27.70 274 140 0.51 0.43 St. Kitts and Nevis 15.60 3.60 6 0 2.60 Trinidad and Tobago 43.80 10.20 105 33 0.31 0.42 Sub Total 428.20 100.00 911 480 0.53 0.00 Fiji 165.35 100 445 270 0.61 0.37 Overall Total 1,294.75 6,380 2,286
107 CHAPTER 6 SUMMARY AND CONCLUSION S Summary This study was motivated by the idea that with reform of the E uropean U nion (EU) sugar polic y there was likelihood that the A frican, C aribbean and P acific (ACP) countries could be adversely affected There was a concern that by liberalizing its sugar trade regime E uropean U nion prices c ould decline and sugar farmers in the ACP countries would no longer be paid two to three times world prices for the sugar they exported to the E uropean U nion Specifi cally, preferential imports from the ACP w ere considered to be development aid, and were tied to the survival of the sugar Common Market Organization (CMO). This potentially presented problem s if some countries were encouraged by EU policies to expand suga r production, while diversification to other crops could have been a better long term alternative. A sudden demise of the CMO potentially implied far reaching social costs in poor nations. The study intended to quantify the implication of sugar reforms by s pecifically understand ing what would be the effects of EU sugar pol icy reform on world production. How w ould sugar production in the ACP and rest of the world be affected? Related to th is was understanding what it meant for world sugar prices, i.e. what w ould be the direction of change as liberalization is implemented? The study uses producer and consumer surplus to quantify welfare gains and losses. P ressure to reform EU sugar policies ha s been building over time. In the 1990s the consensus in agricultur al policy circles, globally, was to move away from price support programs. So while other sectors were being reformed, sugar policy in most countries had been left un changed because of its sensitive political nature. This fact,
108 however, served to highlight sugar as a commod ity that needed urgent reforms. Factors such as European Union enlargement, W orld T rade O rganization commitments to the Uruguay Round, concessions to LDCs through Everything but Arms (EBA), and the Balkan Free Trade Agreement added pressu re to change sugar policies in the European Union. Within the EU, the CMO for sugar was controversial, and it faced several legal challenges regarding its fairness and anti competitive nature. The fairness argument centered on sugar beet farmers receiving support that was not available to other crops, and the distortion s introduced such as that the CMO might be encouraging more European farmers into sugar beet production, than otherwise would be the case under market conditions. In short, domestic pressure in the EU came from competition monitors such as national authorities, the OECD, sugar user industries and consumer groups L ack of competition is generally attributed to the fundamental terms of the CMO, which limited the ability of most efficient producers to develop, imposed limits on production of competing products and created barriers to entry of new producers. Some saw th e arrangement the EU had with ACP countries as a way to protect itself from external competition. Research Steps Literature review was one of the first steps done to determine what model to be utilize d in the study, understand how to best implement it, and compare its pros and cons to other trade models. Literature review covered work on the benefits of trade which strongly supports the notion that opening up economies to trade outweighs protectionist p olicies by increasing the variety of goods available in the market and raising productivity by making less expensive intermediate goods more available. Sugar trade models dating back to the 1960s were reviewed. The types of models that have
109 been utilized i n other research varied from econometric simulation, non spatial equilibrium and general equilibrium models like GTAP. Although the base periods for studies reviewed were different, results from this dissertation do not deviate by much from what other rese archers found to be the impact of liberalization on world prices. A spatial equilibrium model was chosen as the tool to analyze the impacts of liberalization, hence demand and supply elasticities were required along with transportation costs to build th e model. To understand the impact of the change in EU policies, the main sugar producing and co nsuming countries in the world we re incorporated into the model. A world sugar trade model wa s developed to determine the direction of flows of sugar from produc tion to consumption regions, and to determine equilibrium prices. The process involve d estimating domestic supply and demand equations for all countries in the model as well as estimating the cost of moving a unit of sugar between trading nations The dem and and supply el asticities and transport costs we re then used to set up a baseline model which computes net social welfare, defined as the sum of producer and consumer surplus less transportation costs, using 2009 as the base year. The social welfare func tion is optimized using the Generalized Algebraic Modeling System (GAMS), a mathematical programming language. Main Findings The study considered the impact of five different policy scenario s. First, the 2009 baseline results are compared to the impact o f price floor changes of 2006 and 2010. In the second, the EU is assumed to completely liberalize, in the third, both the EU and the US are assumed to completely liberalize, in the fourth only the US liberalizes, while in the fifth scenario, all countries of the world remove all trade restrictions.
110 Scenario one, reveals that world production, consumption and trade are almost insensitive to changes in the price floor, where the quantified effects are within one percent of the base values for all countries. I n Scenario two, the model predicted that the European Union would cease to be an exporter and revert to a net importer status as it was decades ago. Specifically, African and Pacific farmers would realize increased welfare benefits (producer surplus) from liberalization while Caribbean farmers would experience a producer surplus decline due to the strong reliance of Caribbean agriculture on EU imports. When the European Union and the United States were assumed to completely liberalize, supply prices declin ed sharply in the European Union, in the United States, and in the Caribbean. Allowing these two regions to liberalize their sugar economies makes them net importers, which benefits American and E uropean consumers who get access to cheaper sugar under new policies By assuming the United States alone liberalize s its sugar markets, most of the gains would be realized by American consumers who would pay lower sugar prices, and global exporters who would find a bigger market to sell to in the United States. T his would likely be a cause for concern for American farmers, as opening the US market would mean they would have to compete with countries with lower production costs. It is conceivable therefore that American sugar production would decline as a result s pecifically sugar beets Therefore farmers would have to diversify into other areas of agriculture. The final scenario, where all countries are assumed to follow liberalization policies indicates that supply prices would go up in a majority of the countri es, with only
111 the Caribbean, the USA and the EU facing supply price declines. Global production would increase by almost nine percent, and most of that production would end up in the US and EU, two regions that would import more as a result of liberalizati on. Notably aggregate producer surplus would increase for ACP countries. One of the main shortcomings of this study was the lack of supply and demand elasticities for ACP countr ies individually Since this production and price information was not availabl e by country the study used regional groupings to circumvent the data challenges. It can be inferred from export data that, it is highly likely that Mauritius, a country that had more than half of its exports sent to the EU, is likely to experience a more sign ificant revenue drop than Z imbabwe which exported less than 10% of its sugar to the EU, as a result of liberalization. C ountries that exported more than 50% of their sugar to the EU like most Caribbean countries w ould experience negative producer surp lus changes as liberalization occurs Their exports shares to the European Union are on average higher than most ACP countries, so a 36% drop in support prices would affect them negatively. This model indicates that if only the European Union liberalized its sugar markets, global consumption would increase about 6%, and EU sugar consumption would increase to about 28 million tons per year. Global sugar consumption would change increase by about 9% with full market liberalization. To meet this demand, sugar production would have to increase in some parts of the world, and it is reasonable to argue that the lowest cost producers would respond positively. This is the gap that is likely to be covered by the efficient producers in Africa and the Pacific.
112 Recomm endations f or Future Researc h Given that the main aim of the study was to look at the impacts of policy changes on individual African, Caribbean and Pacific countries, there are still unanswered questions As stated before, there was a lack of data, hence each of the eighteen ACP countries could not be incorporated explicitly into the mathematical programming model. As a way of developing this research further, there is a need to gather data and estimate demand and supply elasticities for each of the count ries involved. The model as presented was insensitive to changes in the EU price floor alone. That is adjustments to the price floor do not result in changes in production, import and exports that are greater than one percent in magnitude. That is a puzzli ng result that requires further examination. It could be that the mathematical specification of the model is incorrect, even though other studies seem to suggest it is the right way to model price floors. So there is still room for more research into under standing how to model price floors. Another area of concern is that t he study does not take into consideration t hat most countries keep sugar stocks from year to year to smooth consumption. The study assumes a long run focus, were the effects of stocks is assumed away. However by using 2009 as the base year, it remains questionable whether this indeed was a long run study. So one more area of improvement could be to allow for stocks to be incorporated into the model and then assess the impacts on the resul ts.
113 LIST OF REFERENCES Abbot, G. C. 1990 Sugar. Routledge, London & New York. Anderson, K., W. Martin and D. Van der Mensbrugghe. 2006. Distortions to World Trade: Impacts on Agricultural Markets and Farm Incomes Review of Agricultural Eco nomics 28(2): 168 194. Anderso n, K., and R. Tyers 1991. Global Effects of Liberalizing Trade in Farm Products Trade Policy Research Center, London. Askari, H., and J.T. Cummings 1977. Estimating Agricultural Supply Response with the Nerlove Model: A Surv ey Inte rnational Economic Review 18(2) : 257 292. Ballinger, R.A. 1971 A History of Sugar Marketing Washington DC: U.S. Department of Agriculture, Economic Research Service. Agricultural Economic Report No. 197. Bates, T.H., and A. Schmitz 1969. A Spatial Equilibrium Analysis of the World Sugar Economy Giannini Foundation Monograph Number 23. California Agricultural Experiment Station. Binkley, J. K., and B. Harrer 1981 "Major Determinants of Ocean Freight Rates for Grains: An Econometric Analys is American Journal of Agricultural Economics 63(1): 47 57. Bo r rell, B., R.C. Duncan ., and S.V. Marks. 1993 A Survey of World Sugar Policies In: S.V. Marks and K.E., Maskus ed. The Economics and Politics of World Sugar Policies The Uni versity of Mic higan Press pp 15 48. Borrell, B., and L. Hubbard 2000 Global Economic Effects of the EU Common Agricultural Policy Economic Affairs 20 ( 2 ) : 18 26. Borrell, B., and D. Pearce. 1999. Sugar: The Taste Test of Trade Liberalization Paper presented at t he Conference on Agriculture and New Trade A genda from a Development P erspective. Geneva Switzerland, 1 2 October. Chenery, H., S. Robinson, and M. Syrquin 1986. Industrialization and Growth Oxford University Press. Clarke, M.A. 1988 Sugarcane Processing: Raw and Refined Sugar Manufacture In M. A. Clarke and M. A. Godshall ed. Chemistry and processing of sugar beet and sugarcane Denver, Colorado. Counci l of The European Union. 2007 Council Decision to Denounce the ACP Sugar Protocol Brusse ls, 28 September 2007. 13345/07 (Presse 213). Cournot, A.A. 1838 Mathematical Principles of the Theory of Wealth Translated by N.T. Bacon (1897), MacMillan, New York.
114 Devadoss, S., J. Kropf, and T. Wahl. 1995. Trade Creation and Diversion Effects of N orth American Free Trade Agreement of U.S. Sugar Imports from Mexico Journal of Agricultural and Resource Economics 20(2): 215 230. Dollar, D. and A. Kraay 20 Trade, Growth and Poverty Economic Journal 114 : 22 49. Dornbusch, R. 1992 The Case for Trade Liberalization in Developing C ountries Journal of Economic Perspectives 6( 1 ) : 69 85. Dye, A., and R. Sicotte 2004. The US Sugar program and the Cuban Revolution. The Journ al of Economic History 64 (3): 673 704. EC Report 2004. A description of th e Common Organization of the Market in Sugar. European Commission Agriculture Directorate General. EN AGRI/63362/2004. Elobeid, A., and J. Beghin 2005. Multilateral Trade and Agricultural Policy Reforms in Sugar Markets. Working Paper 04 WP 356. Cente r for Agricultural and Rural Development, Iowa State University. Elobeid, A., and J. Beghin 2006 Multilateral Trade and Agricultural Policy Reforms in Sugar Markets. Journa l of Agricultural Economics 57( 1 ) : 23 48. Enke, S. 1951 Equilibrium among Spat ially Separated Markets: Solutions by Electric Analogue. Econometrica 19: 40 47. Frandsen, S. E., H. G. Jensen, W. Yu, and A. Walter Jorgensen. 2003 Reform of EU Sugar Policy: Price Cuts Versus Quota R eductions. European Revie w of Agricultural Economics 30( 1 ) : 1 26. GAIN Report Number E35225. 2005 Global Agriculture Information Network Washington DC: USDA Foreign Agriculture Service. Gemmill, G. 1976 The World Sugar Economy: An Econometric Analysis of Production and Policies. on, Michigan State University. Gemmill, G. 1980 Form of F unction, T ast e and the D emand for S ugar in S eventy T hree N ations European Economic Review 13 (2) : 189 205. Gotor, E., and M.E. Tsigas. 2006 Effects of the EU Sugar Trade Reforms on Poor Households in Africa: A General Equilibrium Analysis. present ed at Ninth Annual Confere nce on Global Economic Analysis. Addis A baba Ethiopia, June 15 17 Hafi, A., P. Connell ., and R. Sturgiss 1993 Market Potential for Refined S ugar from Austral ia ABARE Research Report 93.17, Canberra.
115 Hammer, J.S., and O. Knudsen. 1990 Agricultural Trade Liberalization: Developing Country Responses. In Ian Goldin and Odin Knudsen ed. Agricultural Trade Liberalization: Implications for developing countries Hannah, A.C. and D. Spence 1996 The International Sugar Trade. John Wiley and Sons, Inc. Herrmann, R., and W. Dietmar 1995 A Welfare Analysis of the EC ACP Sugar Protocol. The Journ al of Development Studies 31 ( 6 ) : 918 941. Hertel, T.W. 1997. Glo bal Trade Analysis: Modeling and Applications Cambridge University Press. Intriligator, M.D. 1978 Econometric models, Techniques and A pplications Prentice Hall, Inc., Englewood Cliffs, NJ. Koo, Won, W. 2002 Revie w of Agricultural Economics 24( 2 ) : 336 352. Laszlo, T. and A.B. Rizzuto. 1990 History of Cane and Beet Sugar. In N L. Pennington and C W. Baker: Van Nostrand Reinhold ed. Sug to Sucrose New York, NY. Marks, S.V., and K.E. Martin, M. V and D. A. Clement 1982. An Analysis of Port Specific International Grain Freight Rates: The Case of the Lower Columbia River Port Area. Transportation Journal 22(1) :18 26. McDonald, S. 1996 Development Policy Review 14(2): 131 149. Milner C., W. Morgan, and E. Zgovu. 2004 The European Journal of De velopment Research 16( 4 ) : 708 808. Ner love, M. 1956 Estimates of the Elasticities of Supply of Selected Agricultural Commodities. Journal of Farm Economics 38(2): 496 509. Nyberg, J. 2005. "Sugar International Market Profile." Markets and Trade Division Food and Agriculture Organization of the United Nations. Polopolus, L.C. 2002 World Sugar Markets and Entangled Gove rnment Programs. In A. Schmitz, T.H. Spreen, W.A. Messina and C.B. Moss ed. S ugar and Related Sweetener Markets. International Perspectives
116 Poonyth, D., P. Westhoff A., Womack, and G. Adams 2000 Impacts of WTO Restrictions on Subsidized EU Sugar E xp orts. A gricultural Economics 22( 3 ) : 233 245. Rao, J.M. 1980 Agricultural Supply Response: A Survey Agricultural Economics 3(1): 1 22. Reddy, M. 2010 Preferential Price and Trade Tied Aid in Fiji: Implications on Price Stability, Certainty and Output Supply. Botswana Journal of Economics 6(10): 71 84. World Bank 2011. Commodity Markets http://www.worldbank.org/prospects/commodities. Roland, H., and W. Dietmar. 1995 A Welfare Analysis of the EC ACP Sugar Protocol. The Journal of Development Stud ies 31( 6 ) 918 941. Samuelson, P.A. 1952 Spatial Price Equilibrium and Linear Programming. The America n Economic Review 42( 3 ): 283 303. Schmitz, A. 2002. The European Union's High Priced Sugar Support Regime. In A. Schmitz, T.H. Spreen, W.A. Messina an d C.B. Moss ed. Sugar and Related Sweetener Markets. International Perspectives Schmitz, A., T.G. Schmitz, and J. L. Seale 2006 CAFTA and U.S. Sugar. University of Florida, IFAS Extension. FE578. Serrano, K. A. 2007 New Sugar Regime Become J ournal of South Pacific Law 11: (2) :169 193 South Center 2007 The Reforms of the EU Sugar Sector: Implications for ACP Countries and EPA Negotiations. Geneva, Switzerland. Takayama, T., and G. G., Judge 1971 Spatial and Temporal Price and Allocation Models North Holland Publishing Company. Thore, S. 1986 Spatial Disequilibrium. Journal of Regional Science 26 ( 4 ) : 661 675 Thursby, J. and M. Thursby. 1984 How Reliable Are Simple, Single Equation Specifications of Import Demand? The Review of Economics and Statistics 66(1): 120 128. Tramel, Thomas E., and A.D. Seale, Jr. 1959. "Reactive Pro gramming of Supply and Demand Relations Applications to Fresh Vegetables." Jou rnal of Farm Economics 41: 1012 22.
117 Tweeten, L., J. Sharples, and L., Evers Smith 1997. Impact of CFTA/NAFTA on U.S. and Canadian Agriculture. International Agricultural Trade Research Consortium. Working Paper # 97 3. Viton, A., Pignalosa, F. 1961 Tr ends and Forces of World Sugar C onsumption. Commodity B ulletin no 32. United Nations Food and Agriculture Organization, Rome. Wooldridge, J.M. 2003 Introductory Econometrics : A Modern Approach, 2e. Thompson Learning.
118 BIOGRAPHICAL SKETCH Sibusiso holds Bachelors of Science degrees in Agricultural Economics from the University of Zimbabwe, and University of Pretoria. In 2004 he graduated from Virginia Tech in Blacksburg VA with an MS in Applied Economics then joined pharmaceutical consultin g firms in the Philadelphia, PA area working as a statistical analyst. In 2009 he enroll ed a t the University of Florida to work on a PhD in Applied Economics specializing in international trade and graduated in December 2012 Besides development economics, his other interests are in mathematics, quantitative analytics and statistical programming politics and half marathons.