Table of Contents Page I. INTRODUCTION 19 II. CLAIMS OF THE UNITED STATES 19
II.1 Causal relationship and other factors which affected and still affect the world
cotton economy 19
II.2 Exclusion of measures 23
II.3. Interpretation of Article 6.3(c) 24
II.3.1 Effect of the subsidy 24
II.3.2 "In the same market" 24
II.3.3 Time period to be considered 25
II.4 Interpretation of article 6.3(d) 25
II.4.1 World market share 25
II.4.2 Time period to be considered 26
II.5 Threat of serious injury 26 III. EXPORT CREDIT GUARANTEES 26 IV. CONCLUSION 27
1. Argentina would like thank the Panel for this renewed opportunity to submit its views as a third party in these proceedings, and as stated in its submission of 3 October101, it will be commenting on some of the claims made by the United States in its written submission of 30 September.102
II. CLAIMS OF THE UNITED STATES II.1 Causal relationship and other factors which affected and still affect the world cotton economy 2. In paragraph 5 of its submission, the United States asserts with respect to marketing loan payments and step 2 payments that Brazil seeks to ascribe extraordinarily low cotton market prices in recent years to US subsidy payments without presenting or explaining to the Panel the factors that led to this low market price level that in turn resulted in larger US subsidies.
3. Similarly, in paragraph 80 of its submission, the United States claims that Brazil's argument rests largely on the assertion that large US outlays under the challenged measures necessarily demonstrate that US measures caused those price declines.103 4. The United States claims that Brazil has not been able to demonstrate the causal connection between the US measures and their effects, nor has it considered other factors which affected and still affect the world cotton economy.
5. Argentina believes that Brazil, in its First Written Submission, provided a precise and comprehensive description of the world cotton market situation, backing the facts with considerable evidence and documentation.104 Similarly, in its Further Submission, Brazil took account of other factors which also contributed to demonstrating the suppressing or depressing effect on prices of the US subsidies.105
6. Argentina would further like to point out that despite the existence of factors other than the US subsidies that could also have had a depressing effect on international prices (such as the development of synthetic fibres, Chinese trade polices and other factors raised by the United States in its further submission106, some of which will be considered by Argentina further on), Article 6.3 of the SCM Agreement107 clearly states that "serious prejudice … may arise in any case where … the effect of the subsidy...".
7. In other words, Argentina considers that Brazil has demonstrated the causal relationship between the subsidies that the United States has granted and continues to grant to its cotton sector and the fall in international cotton prices.
8. That is, Argentina considers that under the SCM Agreement it is not necessary for a subsidy to be the only factor in the decline in international prices in order to be able to establish a causal relationship between that subsidy and the serious prejudice.108 Rather, the subsidy must be a determining factor or, according to the text of Article 6.3(c), its effect must be a "significant" price suppression or price depression, and this was demonstrated by Brazil.
9. Argentina recalls that the United States, given the size and global impact of its cotton market – with a 41.6 per cent share of the world market – is the international market "price-setter" par excellence.
10. Thus, without the US subsidies which generate a world market surplus, international cotton prices would have been higher or would not have fallen as much. Similarly, if the US share in the world market had not increased as a result of the subsidies, the international price of cotton would have been higher or would have not fallen as much, and as a result, third-country producers, including Argentina, would not have suffered as much prejudice as a result of artificially depressed prices.109
11. Argentina does not agree with the US statement that Brazil has not established a prima facie case because it has failed to provide sufficient evidence of the causal relationship between the enormous budgetary outlays and the low international cotton prices.110
12. Brazil has not based its claims on a mere assertion, but rather, as we have already stated111, the number and quality of the empirical and econometric analyses presented by Brazil in its Further Submission112, which were carried out both by international organizations and by various prestigious US institutions, not to mention the USDA itself, provide irrefutable evidence of the collective and individual effects of each subsidy programme on the price of cotton.
13. It is therefore difficult for Argentina to understand how the United States can claim that other factors, and not its subsidies, were the cause of the dramatic fall in cotton prices over the past few years. Nor does Argentina understand how the United States can disregard the evidence provided by Brazil113 to the effect that during the marketing years 1999 to 2002, the total value of US cotton subsidies amounted to almost US$13 billion while the average cotton subsidization rate was 95 per cent.114 14. In the paragraphs that follow, Argentina will refute some of the arguments put forward by the United States concerning other factors that may have influenced the fall in international cotton prices:
15. FIRST: the United States claims that the explosion in the production of synthetic fibres played a considerable part in causing cotton prices to fall. Argentina submits that the contrary appears to be true.
16. Indeed, the "Fibre Prices" Table in paragraph 23 of the US Further Submission shows that polyester prices have always been lower than cotton prices (see: "US mill" as compared to "US spot" and "Asia poly") and, moreover, they appear to follow cotton prices (see 1995, when cotton prices reached their record level for the series and polyester happened to follow the same trend).
17. The fact that polyester had to adapt to cotton prices, and not the reverse as the United States claims, is confirmed by the very close correlation between cotton and polyester prices. On the other hand, the last few years do not show any correlation between the price per barrel of oil and the price of polyester fibres.
18. SECOND: Argentina does not understand how the United States can claim that without the increase in US retail consumption, international cotton prices would currently be lower115 (which is not being questioned), while disregarding the role that its enormous subsidies have played and still play in the fall of international cotton prices.
19. As we have already pointed out116, price movements of the US Cotlook "A" Index and third country (e.g. Brazilian and Argentine) market prices are directly interconnected, and this is an unquestioned and irrefutable fact. This being the case, the amount of the US subsidies granted to the cotton sector added to the scale of US production and exports are decisive when it comes to determining the extent to which the subsidies affect the fixing of both international and third market prices.117
20. THIRD: The United States correctly points out that since the world's cotton trade is managed in US dollars, an appreciation of the dollar will lead to a fall in the price of cotton both in the United States and in third markets.118 What the United States does not explain in connection with this fact is why this appreciation of the dollar by some 37 per cent between 1995 and 2002119 did not also result in a fall in production, and consequently, in US cotton exports. The National Cotton Council (NCC) gives us the answer, pointing out that without its subsidies, the United States' share in the world cotton market would have declined.
21. FOURTH: Regarding the United States argument that China is the giant of the world cotton industry, and hence the impact of its trade policies and stocks, we note that neither Brazil nor third parties, such as Argentina, overlooked this fact. We repeat, Argentina has already pointed out120 that while there were a great many cotton producing countries, four of them (China, the United States, India and Pakistan, in descending order) alone account for two thirds of world cotton production.
22. However, Argentina also pointed out that most of the cotton was used in the producing country itself, and that the great exception to that rule was the United States, which exported over a half of the cotton it produced and was the world's leading exporter. This was why the level of subsidization in the United States was the main factor in determining the world cotton market price. In other words, while China may be the giant in the world cotton industry, the United States is the giant in world cotton trade.
23. FIFTH: The United States claims that the decisions of farmers are based on expected cotton prices for the upcoming crop year and not prices from the previous crop year as cited by Brazil, and that US cotton producers are not insulated from international price movements.121 We can only repeat some of the questions that we addressed to the United States during the consultations122, namely:
24. If this is so, how does the United States explain the fact that in 2001 – the fifth year of falling prices – US cotton producers achieved a record harvest of 20.3 million tons, an increase of 42 per cent compared to 1998, and that the cotton planted area increased by 6 per cent during the same period? Why does the USDA estimate a 10 per cent drop in the world production for 2002 – reflecting the impact of world prices on investment – and at the same time estimate for this year another record harvest in the US – the fourth biggest ever recorded? How does the US explain the reasons for the increase in the volume of its cotton exports from 946,000 tons in 1998 to 1.8 million tons in 2001, while there was a drop in international prices?
25. Besides, the United States is not a low-cost producer123 (despite its claim that pest eradication programmes and the adoption of genetically modified varieties of cotton have lowered its production costs124), and its productivity levels are lower than those of other exporting countries. Nevertheless, while international prices have fallen by some 54 per cent since the middle of the 1990s, the United States has expanded the area under cotton and increased its production. How does the United States explain the lack of correlation between the world cotton price and US cotton production?
26. Brazil has given answers to all of these questions. Argentina has also pointed out that if the United States did not grant subsidies to its cotton sector, the US cultivated acreage and production would diminish. US exports would also diminish and, the US being the world's leading supplier of cotton, international prices would be higher or would not have decreased as much.125 27. Argentina considers that the evidence submitted with respect to the increase in US production and exports which took place entirely independently or in isolation from the fall in the international price clearly demonstrates that US cotton producers are immune to changes in the market prices for cotton.126 28. SIXTH: The United States claims that the eradication of pests and the adoption of genetically modified varieties of cotton have lowered production costs.127 We stress that even so, there continues to be a widening gap between those costs and market prices.
29. We repeat what has already been stated by Brazil and Argentina128, namely that while US cotton production costs are among the highest in the world, US producers' market prices have fallen from US$0.60 to US$0.30 per pound.129 The only possible explanation of how the United States has been and continues to be able to bridge the widening gap between production costs and market prices is subsidies, since without them many US producers would have been or would be compelled to cease cotton production.
30. SEVENTH: Argentina does not understand how the United States can claim that US cotton producers are highly sensitive to price changes when in spite of a 54 per cent fall in international prices since the middle of 1990s, the area under cotton and the production of cotton in the United States expanded considerably. In other words, contrary to what the United States has claimed, the area under cotton has responded to the fall in international prices by increasing steadily. The only way to achieve such a result is to grant enormous subsidies, since without them the cultivated area, and hence production, would have decreased.
31. EIGHTH: Argentina does not understand how the United States can state that its cotton producers show greater sensitivity to price changes than is demonstrated by third markets when, for example in Argentina – which is basically a "price-taker" in the international cotton market – the cultivated area shrank by 76 per cent during the marketing year 2001/2002, while production fell by 63 per cent compared to 1998.130 II.2 Exclusion of measures 32. Regarding the US argument that some of its domestic support measures should not be included in the analysis131, Argentina considers that the Panel should examine the collective effects of all of the support measures that are not green box measures. Argentina does not agree with the US statement that direct payments and counter-cyclical payments should be excluded from the analysis simply because their individual effects may not be that significant.
33. It is the collective impact of all of the US subsidies that has effects on the cultivated area, production, exports and prices.
II.3. Interpretation of Article 6.3(c)
II.3.1 Effect of the subsidy
34. Argentina considers that Brazil has made a proper prima facie case with respect to its claim of inconsistency with Article 6.3(c) of the SCM Agreement, conclusively demonstrating the significant price suppression or depression effect.
35. Firstly, Argentina does not understand how the United States can simply brush aside the Panel's findings in Indonesia-Automobiles132, since this was the only dispute under the GATT-WTO which dealt with the interpretation of the term "significant".
36. Secondly, Argentina fails to understand how the United States can claim that Brazil argued that it is the effect on producers that must be significant, and not on prices, when Brazil has submitted copious evidence, based on numerous empirical and econometric analyses, of the effects of the subsidies on prices.
37. It is remarkable that the United States should completely disregard these analyses, especially considering they were conducted by international organizations and by different prestigious US institutions, not to mention the USDA itself133, in an attempt to distort Brazil's evidence.
38. Argentina repeats that over and above any endorsement that may be given to the conclusions of any one of these studies (and each study's estimate of the price effect of the subsidies), an increase in the world price of cotton would be significant, even if international price suppression or depression were to amount to only one per cent per pound, since such an increase would enable countries such as Brazil and Argentina to recover their competitive positions in the world cotton market.134
39. Finally, at no time does the United States seem to suggest that a suppression or depression effect of 12.6 per cent on international cotton prices is not "significant" within the meaning of Article 6.3(c). The Panel should therefore find that the subsidies in question have caused and still cause significant depression of cotton prices in the world market resulting in serious prejudice to Brazil's interests.
II.3.2 "In the same market"
40. Contrary to what the United States has claimed135, Argentina considers that Brazil has presented sufficient evidence with respect to the significant effect of subsidies on prices for each relevant geographical market, including the United States, Brazil, the African Countries, other producer countries and Brazilian export markets.
41. The United States provides no legitimate reason why the Panel should not consider the world market. As stated earlier136, price movements of the United States, of the Cotlook "A" Index and third country (e.g. Brazilian and Argentine) market prices are directly interconnected, and this is an unquestioned and irrefutable fact.
42. Moreover, US cotton forms part of Cotlook's "A" Index basket, so that the Panel cannot ignore the fact that the US subsidies have a decisive impact on the price of cotton in the world market.
43. Indeed, this being the case, the amount of US subsidies granted to the cotton sector added to the scale of US production and exports is decisive when it comes to determining the extent to which the subsidies affect the fixing of both third market and world market prices.
II.3.3 Time period to be considered
44. Regarding the US argument that subsidies that have ceased to exist can have no "effect"137, Argentina would like to recall the Panel's remarks in the Indonesia-Automobiles concerning the irrelevance of serious prejudice having been caused by programmes that are no longer in force. Upon examining whether the subsidies caused serious prejudice to the interests of the complainants, the Panel in the said case rejected the argument that the effects of an expired subsidy programme could not be considered.138
45. Moreover, Argentina considers that it is necessary to consider a sufficiently extensive period to reflect market trends, and a period of one year as suggested by the United States is not sufficient.139 II.4 Interpretation of article 6.3(d) II.4.1 World market share
46. The United States errs in its interpretation of the expression "world market share" by trying to identify it with "share in world consumption".140 If, as the United States contends, the expression "world market share" in Article 6.3(d) refers to the increase in consumption of the country granting the subsidy, it would be contrary to the object and purpose of Articles 5 and 6 of the SCM Agreement, namely to avoid adverse effects of subsidies to the interests of other Members.
47. Indeed, if there were an increase in the share in world consumption of the product subsidized by the Member granting the subsidy, this would very probably lead to an increase in the international price of the product in question, and hence, there would be no adverse effects for other Members. In other words, to identify "world market share" with "share of world consumption" would completely subvert the underlying rationale of Article 6.3(d).141 48. Moreover, if we take account of the immediate context of Article 6.3(d), i.e. footnote 17, which states "unless other … rules apply to trade … ", we have a clear indication that the expression "world market share" can only refer to the share in world exports.
II.4.2 Time period to be considered
49. Contrary to the US claim that the trend in the period considered is not consistent because it includes years in which the United States world market share decreased rather than increasing142, the fact is that there will always be peaks and troughs in agricultural production and export for climatic and other reasons. This does not mean that a trend over the years cannot be "consistent".
50. In other words, the word "consistent" cannot be interpreted in such a way as to allow a decrease in world market share during a given year to invalidate a trend over several years. On the contrary, the word "consistent" should be interpreted in the context of the investigation period, disregarding the market variations.
II.5 Threat of serious injury 51. Argentina considers that since Brazil has established the existence of serious prejudice caused by the US subsidies, the threat of serious prejudice is clearly foreseeable and imminent as a result of the even higher subsidies planned under mandatory US legislation for the marketing years 2003-2007. Consequently, Argentina maintains that Brazil has established a prima facie case that these subsidies threaten to cause serious prejudice to Brazil.
52. Argentina contends that this guaranteed flow of subsidies will unquestionably lead to a higher level of US cotton production and exports. This will inevitably result in price suppression and depression as well as an increasing and inequitable US market share for cotton, thus creating a source of permanent uncertainty that confirms the threat of serious prejudice generated by the subsidies.
53. Argentina further considers that the link between US cotton subsidies and the threat of significant price suppression and depression and of an increase in the US world market share for cotton stems from the fact that the future subsidies will be as necessary as the current ones for US producers to bridge the gap between market prices and their total production costs. This will enable US producers to continue competing with more efficient third country producers, especially considering that the USDA itself forecasts an increase in total production costs.143
III. EXPORT CREDIT GUARANTEES 54. In its Further Submission, the United States reverts to its argument in connection with the negotiating history for the interpretation of Article 10.2 of the Agreement on Agriculture, maintaining that there are no disciplines regulating the use of export credit guarantees.
55. Argentina repeats what it stated in its oral submission of 24 July 2003, namely that Article 10.2 in no way provides an exception to the general disciplines on export subsidies, and in particular, to the applicability to Article 10.1. As Argentina has pointed out, "if that had been the intention, then the negotiators would have expressly said so."144 IV. CONCLUSION 56. For the reasons set forth both in this statement and in previous submissions, Argentina requests the Panel to issue the findings and recommendations requested by Brazil throughout these proceedings.
ORAL STATEMENT OF BENIN
8 October 2003
Mr. Chairman, members of the Panel,
1. My name is Eloi Laourou from the Mission of Benin. I am joined by Mr. Nicholas Minot of the International Food Policy Research Institute in Washington, D.C., the co-author of the study that has been annexed to our Third Party submission, “Effect of Falling Cotton Prices on Rural Poverty in Benin”. I will ask Mr. Minot to speak to you in a moment. The other members of our delegation are our legal advisers, Mr. Brendan McGivern and Mr. Stefan Ramel, both from White & Case.
2. As noted in our Third Party submission, this is the first time that Benin has participated in a WTO dispute. We have not taken the decision to participate lightly. Indeed, it was only the serious threat posed to the economic and social stability of our country by massive, WTO-inconsistent US subsidies on cotton that has led us to take this unprecedented step.
3. The cotton farmers of Benin are efficient producers. The World Bank has estimated that the cost of producing cotton in West Africa is about 50 per cent of the cost of production in the United States. Moreover, the cotton sectors of both countries have undergone considerable structural reforms.
4. Yet the economic efficiencies of our producers, and the painful reforms they have accepted, have in the end proved to be almost completely irrelevant. US subsidies have had a ruinous effect on the world price of cotton, which in turn has had devastating effects on the economies of West Africa.
5. US cotton subsidies do not just dwarf the cotton sectors of West Africa. They dwarf all economic activity in the region. As noted in our submission, the subsidies paid by the United States to its prosperous 25,000 cotton farmers exceed the gross national income of Benin, Chad, Burkina Faso, Mali, Togo and the Central African Republic.
6. Oxfam estimates that for the period from 1999/2000 to 2001/2002, Benin suffered a total cumulative loss of $61 million in export earnings. Mr. Chairman, this is not an abstract, anodyne statistic. This translates into genuine suffering on the ground, as hundreds of thousands of people, deprived of export earnings, are pushed from bare subsistence to absolute poverty. Indeed, Dr. Minot estimates that a 40 per cent reduction in farm-level prices of cotton has the effect of pushing an additional 334,000 thousand people below the poverty line in Benin.
7. This is an important point, one that should be considered carefully by the panel as it assesses the meaning of “serious prejudice” to one of the poorest countries in the world.
8. With your permission, I would now ask Mr. Minot to summarize briefly the results of his study on how depressed world prices for cotton translate into poverty in Benin.
9. Thank you for the opportunity to present the results of a study that I hope will be relevant to the dispute. Before I begin, I would like to provide some background. I am a Research Fellow at the International Food Policy Research Institute, a Washington-based international organization whose mandate is to generate information to address problems of hunger and poverty in developing countries. I received my Ph.D. in agricultural economics from Michigan State University and have worked on issues of agriculture in developing countries for more than 15 years, including four years living in sub-Saharan Africa.
10. From 1998 to 2000, I led a German-funded study of the impact of agricultural reforms on farmers in Benin. In collaboration with a local research firm, we carried out four surveys in Benin: surveys of farmers, traders, market managers, and village cooperatives. In 2002, I was contracted by the World Bank to use these survey data to examine the impact of falling world cotton prices on poverty in Benin. My co-author, Lisa Daniels, and I finished the report later that year and a version of it was distributed as an IFPRI working paper in November.
11. Cotton prices are affected by competition with synthetic fibres, weather-related supply shocks, the rate of growth in the global economy, and government policies, among other factors. Cotton prices are pushed below what they otherwise would be by government support to cotton growers. The International Cotton Advisory Committee estimates that worldwide direct assistance to cotton growers was US$ 4.9 billion in 2001/02. Of this amount, the United States accounted for US$2.3 billion, equivalent to 24 cents per pound of cotton produced. Other sources, using a broader definition of assistance, estimate that the government provides US$ 3.9 billion to the cotton sector.
12. Until 2002, US cotton policy consisted of various programs, including two (the marketing loan program and loan deficiency payments) that ensured that farmers receive at least 52 cents/pound. This has the effect of insulating US farmers from falling world prices. In 2001, in spite of low world prices, the US posted record cotton production and near-record export volumes. Furthermore, US subsidies to cotton have increased since these studies were carried out. The 2002 Farm Bill introduced target prices for the major commodities and programs that effectively pay US farmers most of the difference between market prices and the target price. For upland cotton, the target price is 72 cents/pound. In addition, by allowing farmers to update their “base acreage”, the new policy provides incentives for farmers to expand production.
13. Several recent studies have attempted to assess the impact of subsidies on world prices. The Centre for International Economics in Canberra uses a five-region world model of fibre, textile, and garment markets in 2000-01 to simulate the impact of US and European subsidies on cotton production and export. They find that removing US and European subsidies to cotton growers would raise the world cotton price by 6 cents/pound or 11 per cent. Another study, carried out by ICAC, estimates that removing US production subsidies would have increased the world price by 11 cents/pound in 2001/02. And most recently, Sumner estimates that, in the absence of US subsidies, the world cotton price would have been 12.6 per cent higher over 1999-2002.
14. The adverse impact of lower cotton prices on export revenue and GDP in cotton exporting ations is clear, but does this translate into higher incidence of rural poverty? If cotton is grown mainly by larger farmers with relatively high incomes, then the effect of changes in cotton prices on rural poverty may be modest. Even if cotton is grown primarily by small farmers, the magnitude of the effect on rural poverty will be small if few farmers grow cotton or if it accounts for a small share of rural income. Assessing the direct impact of changes in cotton prices on rural poverty requires detailed household-survey data on incomes and expenditures.
15. The paper examined the impact of changes in cotton prices on rural poverty in Benin. In particular, it had two objectives:
to describe the living conditions and level of poverty for cotton growers and other farmers in Benin; and
to estimate the short and long-run impact of lower cotton prices on the income of cotton growers and on the incidence of poverty in rural Benin.
16. The Republic of Benin has a population of about six million, 59 per cent of whom live in rural areas. Its rural economy is based on maize, sorghum, millet, yams, cotton, and livestock production. The per capita gross national product is US$380, placing Benin among the poorest countries in the world. Indeed, the per capita income of Benin is lower than the average for sub-Saharan Africa.
17. In 1989, Benin entered a period of economic and political reform. It made a peaceful transition from a military government to a constitutional multi-party democracy. It also began to move from a quasi-socialist economy to a free-market economy. In the agricultural sector, state farms and cooperatives were disbanded, food crop prices and marketing were liberalized, and many state-owned enterprises were privatized or closed. In January 1994, the CFA franc was devalued by 50 per cent, effectively doubling the price of imports and the returns to exports. Although this imposed hardship on manufacturing firms and urban consumers, it stimulated the local production of cotton, rice, and other tradable goods.
18. In the past two years, Benin has greatly reduced the role of the state cotton marketing board, introducing competition in the distribution of inputs and the marketing of cotton. The fall in world cotton prices has led to political pressure for the government to support the domestic price or even to re-assume control of the sector to protect farmer interests. Cotton represents 90 per cent of agricultural exports and around 70 per cent of its total exports (excluding re-exports).
19. Because the reliability of the results depends heavily on the quality of the survey data, it is worth briefly describing the survey methods. The survey, called the Enquête des Petits Agriculteurs (EPP) or Small Farmer Survey, was carried out in 1998 by the IFPRI and a local research firm, the Laboratoire d’Analyse Régionale et d’Expertise Sociale (LARES). The survey used a 24-page questionnaire covering 16 topics. The households were selected using a two-stage stratified random sample procedure based on the 1997 Pre-Census of Agriculture. In total, one hundred villages were selected. In each village, nine households were randomly selected using household lists prepared for the pre-Census of Agriculture. Due to some variation in the number of households interviewed in each village, the final sample was 899 rural households. The interviews were carried out in local languages by two teams of Benin interviewers, supervised by staff from LARES and IFPRI.
20. In order to study poverty, we need to define it. In this analysis, the poor are defined as those living in households whose per capita expenditure is below the 40th percentile in rural areas. Expenditure is used instead of income because it is more reliably measured and is a better measure of household well-being. It includes cash spending on consumption goods, the value of home-produced food, and the rental equivalent of owner-occupied housing. The resulting poverty line is equivalent to US$123 per person per year. It is worth noting that this is a low poverty line, far below the US$1 per day frequently used by the World Bank.
21. We simulated the impact of various percentage reductions in cotton prices on the incomes of rural households using the concept of producer surplus. The details of the calculation are shown in the paper, but these are standard formulas used in economic analysis. In the short run (before households respond to lower prices), the change in income of each household is simply the percentage change in the value of cotton production multiplied by the quantity produced. In the long run, lower cotton prices will cause farmers to substitute away from cotton, so the impact is smaller. We simulated the impact of these cotton price changes in the short and long run on each of the 899 household in the sample to generate estimates of the impact on rural income and poverty.
22. Before turning to the simulation results, I will describe the role of cotton in the rural economy and the characteristics of cotton growers. According to the IFPRI-LARES survey, cotton is grown by roughly one-third of the farm households. Cotton accounts for about 18 per cent of the area planted by farm households and 22 per cent of the gross value of crop production. In value terms, cotton is the second most important crop, after maize. Among cotton farmers, the average area planted with cotton is 2.3 hectares, producing 2.7 tons of seed cotton. The value of this output is US$ 901 per cotton farm. Cotton accounts for about one-third of the value of crop sales (these figures are shown in Table 2 of our paper).
23. Cotton growers tend to have farms that are larger than other farmers, but they are similar to other farmers in terms of the poverty rate and average per capita expenditure. The larger farms do not translate into a higher standard of living because cotton production is concentrated in the north, which is more arid and has fewer opportunities for non-farm employment. It is not that cotton farmers are poorer than average, but rather that almost all farmers in Benin, including cotton farmers, are quite poor.
24. Turning to the simulations, the short-term impact of a 40 per cent reduction in the farm-gate price of cotton reduces the income of cotton growers 21 per cent. Taking into account the incomes of non-growers, which do not change in this simulation, the average income of rural households falls 7 per cent. Smaller reductions in the cotton price cause roughly proportional changes in income, as shown in Table 3 of our paper.
25. With a 40 per cent fall in the cotton price, the average incidence of poverty, including both cotton growers and other farmers rises 8 percentage points, from 40 per cent to 48 per cent. In absolute terms, this implies that about 334 thousand people would fall below the poverty line. A 40 per cent decrease in the price of cotton results in a 40 per cent increase in the depth of poverty (P1) and a 61 per cent increase in the severity of poverty (P2).
26. Does it matter what poverty line we use? By looking at the cumulative distribution of income with and without the price change, we can evaluate the sensitivity of the results to alternative poverty lines. As shown in Figure 2 in our paper, similar results would have been obtained for higher and lower poverty lines. The results are not very sensitive to the elasticity assumption.
27. In summary, our paper analyzed the impact of changes in world cotton prices on farmers in Benin. Both quantitative measures of per capita expenditure from household surveys and qualitative responses to our 1998 survey suggest that rural living conditions improved over the 1990s. Furthermore, farmers tended to attribute this improvement in rural living conditions to economic factors such as crop prices, availability of food, and access to non-farm employment. Although the causal link is difficult to establish with certainty, it appears the economic reforms of the 1990s (including the 1994 devaluation) and the growth of cotton production during this period contributed to a noticeable improvement in rural standards of living.
28. The link between world cotton markets and rural living conditions can, however, work gainst farmers as well. The analysis in this paper is based on the 39 per cent decline in the world price of cotton between January 2001 and May 2002. We combined farm survey data from 1998 with assumptions about the decline in farm-level prices to estimate the short- and long-term direct effects of cotton price reductions on rural income and various measures of poverty.
29. The results indicated that there is a strong link between cotton prices and rural welfare in Benin. A 40 per cent reduction in farm-level prices of cotton is likely to result in a reduction in rural per capita income of 7 per cent in the short-run and 5-6 per cent in the long-run. Furthermore, poverty rises 8 percentage points in the short-run, equivalent to an increase of 334 thousand in the number of people below the poverty line. In the long run, as household adjust to the new prices, the poverty rate settles down somewhat, remaining 6-7 percentage points higher than it was originally.
30. Furthermore, these estimates may well underestimate the actual effect of lower cotton prices on rural poverty in Benin. First, in an economy with unemployed resources and excess capacity, an external shock affecting income (such as a change in cotton prices) has a multiplier effect. Changes in cotton farmer income result in changes in demand for goods and services produced by their non-cotton-growing neighbours, which in turn influences their income and their demand for goods and services. Estimates for four countries in sub-Saharan Africa suggest that the multiplier is in the range of 1.7 to 2.2, meaning that the total effect on income (positive or negative) is 1.7 to 2.2 times greater than the direct impact.
31. Second, we assume that farm prices change by the same proportion as world prices. In competitive markets with a fixed marketing margin, the percentage change in farm prices will be greater than the percentage change in world prices. Until recently, the effect of changes in world prices on farm-level prices in Benin was muted by government regulation of the market which stabilized prices. Under market reforms being carried out in Benin and elsewhere in West Africa, markets are becoming more competitive and changes in farm prices will closely match changes in world prices.
32. Third, our estimates do not take into account other indirect effects associated with declining cotton production. An earlier analysis of the Small Farmer Survey data from Benin indicated that cotton farmers are three times more likely to apply fertilizer to their maize crops compared to non-cotton farmers. This is because growing cotton gives farmers access to fertilizer on credit, some of which they “divert” to their maize fields. The implication is that lower cotton prices will indirectly reduce the yields of food crops.
33. Overall, the results in this paper challenge the stereotype of the rural poor in developing countries as consisting of subsistence farmers that are relatively unconnected to, and thus unaffected, by swings in world commodity markets. At least in the case of Benin, to the extent that fluctuations in world cotton prices are transmitted to farmers, they will have a significant effect on rural incomes and poverty. The broader implication is that policies that subsidize cotton production in the United States and elsewhere, dampening world prices, have an adverse impact on rural poverty in Benin and (by extension) other poor cotton-exporting countries. Thank you, Mr. Chairman and members of the Panel.
34. Thank you for allowing Dr. Minot to present his paper, and for allow Benin to present its views.
35. This concludes our oral statement. We respectfully ask this Panel to find that the United States is in breach of its WTO obligations, including by causing serious prejudice to the interests of Benin and other Members. We would be pleased to answer any questions that you may have.