I. Terms of Reference 278 II. Economic Data 279 III. Domestic Support 291 IV. Export Credit Guarantees 293
Download 3.13 Mb.
|
- Navigate this page:
- Upland cotton loan activity (pounds
- Planted and Harvested Upland Cotton Acres
- 210. Are worldwide planted acreage figures available BRA, USA
- (a) to what extent does the use of 1997 survey technological coefficients with annually updated values affect the results
- (b) to what extent do producers base planting decisions on their ability to cover operating costs but not whole farm costs USA
- Cumulative net returns ($ per acre)
- Harvest Futures Prices at Planting Time Compared to USDA Baseline Expected Prices (cents per pound)
207. Please indicate whether any of the measures challenged in this dispute obliges cotton farmers to harvest their crop in order to receive the benefit of the programme (subsidy). 31. One programme at issue in this dispute requires that a cotton farmer harvest upland cotton in order to receive payment: the marketing loan programme. In addition, the user marketing certificate programme (Step 2) requires that upland cotton have been harvested and marketed although payment is made to upland cotton users and not to farmers directly. 208. Please provide data for the marketing years 1992 and 1999-2002 of the "quantity of production to receive the applied administered price" (Agreement on Agriculture, Annex 3, paragraph 8) for purposes of a price-gap calculation of support through the marketing loan programme. 32. The marketing loan programme is a direct payment to support producer income that is "dependent on a price gap," namely, the difference between the loan rate and the Adjusted World Price. Thus, the price‑gap calculation of support would be calculated under paragraphs 10 and 11 of Annex 3 of the Agreement on Agriculture on "non‑exempt direct payments," not paragraph 8 on "market price support." That said, the calculations to quantify the support under these provisions are similar. Under paragraph 10, "non‑exempt direct payments which are dependent on a price gap shall be calculated either using the gap between the fixed reference price and the applied administered price multiplied by the quantity of production eligible to receive the administered price, or using budgetary outlays." 33. Under the 1990 Act, only production produced on cotton base acres or other programme crop base acres was eligible for marketing loans. Based on the 1992 Compliance Report (Exhibit US‑39), 11,164,726 acres programme acres were planted to upland cotton. Based on an average crop yield of 694 pounds per acre275, 7,748,319,844 pounds or 16,142,333 480‑lb bales, would have been eligible for marketing loans in 1992. 34. Under provisions of the 1996 farm bill, upland cotton planted on farms with any programme crop base were eligible for marketing loans. While production data are not collected by the Farm Service Agency, the following quantities of upland cotton were put under loan or collected a loan deficiency payment in 1999, 2000, 2001 and 2002. Upland cotton loan activity (pounds)
Source: USDA, Farm Service Agency, Loan Deficiency Payment and Price Support Cumulative Activity as of 10 December 2003. Available at: HREF="http://www.fsa.usda.gov/dafp/psd/reports.htm#Unofficial" 35. Thus, compared to MY1992, the quantity of cotton placed under marketing loans or receiving loan deficiency payments in MYs 1999, 2000, and 2002 were similar. In MY 2001, as a result of exceptional weather conditions and yields, a much larger quantity of cotton was eligible. 36. As noted above, the price gap calculation involves comparing a fixed reference price to an applied administered price. Under paragraph 11 of Annex 3, the fixed reference price "shall be based on the years 1986 to 1988 and shall generally be the actual price used for determining payment rates." The applied administered price for marketing loan payments is the marketing loan rate. The fixed reference price is the average over 1986 to 1988 of the Adjusted World Price (the "actual price used for determining payment rate"). 37. The average Adjusted World Price for 1986‑88 was 53.65 cents per pound – that is, higher than the loan rate in marketing years 1992 (52.35 cents per pound), 1999‑2001 (51.92 cents per pound), and 2002 (52 cents per pound). The result is that the gap between the fixed reference price and the applied administered price is always negative as would be the AMS calculation. When these price gap calculations for marketing loan payments are utilized, negative numbers result, reflecting the decrease in support from the 1986‑88 level. Similarly, if the applied administered price for marketing years 1999‑2002 were compared to the 1992 applied administered price, the resulting negative numbers would again show the decrease in the level of support from MY 1992. Thus, the large budgetary expenditures for marketing loan payments in recent years obscures the fact that the level of support decided by the United States had declined; the price gap calculation, on the other hand, reflects this reduction in support. As the United States demonstrated in its rebuttal submission, by calculating both deficiency payments and marketing loan payments using a price gap methodology, the upland cotton AMS reveals that in no year have the challenged US measures granted support in excess of that decided during the 1992 marketing year.276 209. It is understood that the data in the graph in paragraph 5 of the US oral statement are as at harvest time, while the data in the graph in paragraph 39 of Brazil's oral statement are as at planting time. Please explain why the trend of US acreage increase/decrease differs between these two graphs. BRA, USA 38. The planted and harvested area differ because of abandonment. Over the period 1965 to 2003, the rate of abandonment (abandoned acres divided by total acres) for US upland cotton averaged 8.3 per cent, but the rate will vary from year to year primarily because of weather, primarily in the Southwest. In 1997, for example, weather in the Southwest was generally good and the abandonment rate for that year was only 3.6 per cent. By contrast, dry weather in Texas, Oklahoma and parts of the Southeast in 1998 led many farmers to abandon their cotton crop because of poor yields, resulting in an abandonment rate of 20 per cent. Planted and Harvested Upland Cotton Acres (1,000 acres)
Source: USDA, National Agricultural Statistics Service, Acreage, various issues. 
39. Comparing the per cent change from the prior year of planted and harvested US upland cotton acreage shows that movements in acreage figures are fairly similar, as one would expect. 210. Are worldwide planted acreage figures available? BRA, USA 40. To our knowledge, planted acreage figures are not available on a consistent basis across countries. No other sources (including ICAC) carry worldwide‑planted area. Harvested area is the standard, but in reality many countries do not have a sophisticated system for data collection. To provide a comparative analysis of US acreage changes to the rest of the world, the United States has therefore used harvested acreage, the most reliable acreage measure available. 211. Brazil presents a graph in paragraph 59 of its further rebuttal submission indicating the increasing cumulative loss incurred by cotton producers. Please comment on the argument that US cotton producers could not continue operating without subsidies. In particular: (a) to what extent does the use of 1997 survey technological coefficients with annually updated values affect the results? 41. Brazil’s cost of production argument is based on erroneous facts and arguments. Brazil points to dated average total cost of production data over a randomly selected period, compares this to market revenue, and proclaims that any "gap" must be covered by subsidies. The United States has identified several fatal conceptual and factual flaws in Brazil’s analysis.277 For example, Brazil ignores the evidence on record that a significant number (approximately 47 per cent) of traditional (and likely high‑cost) upland cotton producers no longer plant upland cotton. This structural shift in the industry is not reflected in cost of production data. In addition, Brazil treats the only sources of farm income as cotton market prices and government payments, ignoring crop diversification and off‑farm sources of income. By ignoring alternative revenue sources, Brazil invalidates its claim that only government payments could serve to cover any alleged cost‑revenue gap. 42. But most importantly, Brazil has no legal basis for its argument. Brazil argues that the Appellate Body in Canada ‑ Dairy (21.5) has stated that total costs are the relevant measure, but that reasoning is inapt here. The only question in that dispute was whether a practice involved an export subsidy within the meaning of Article 9.1(c) of the Agriculture Agreement. Solely because the question was to determine whether certain milk provided to processors constituted a payment for purposes of Article 9.1(c) did the Appellate Body opt to use the average cost of production.278 However, the Appellate Body explicitly recognized that "a producer may well decide to sell goods or services if the sales price covers its marginal costs." The Appellate Body also noted that cost of production can be measured "in at least two ways": (1) per unit average total cost of production and (2) marginal cost of production.279 Here, the issue for which Brazil seeks to use total costs is not whether a subsidy exists but to evaluate the effect of the subsidy, an altogether different analysis. Thus, Canada ‑ Dairy (21.5) provides no support Brazil’s average total cost argument. (a) to what extent does the use of 1997 survey technological coefficients with annually updated values affect the results? 43. As described in detail in previous US submissions280, the combination of, among other things, the boll weevil eradication programme and the extraordinary adoption rates of biotech cotton have combined to lower producers’ costs and enhance net revenues. Despite the difficulty in providing precise figures on the extent of cost savings and net revenue increases for the cotton sector that have occurred since the 1997 USDA ARMS cost and returns survey, the rapid adoption of biotech cotton (over 90 per cent of area in key producing States) suggests farmers are reaping significant benefits in terms of net returns. These cost savings have been analyzed and documented in a wide range of studies. 44. In June 2002, the National Center for Food and Agricultural Policy (NCFAP) compiled 40 case studies of 27 crops to document the benefits of biotechnology.281 These case studies were done by various universities. Among other findings, one study found that adoption of insect resistant biotech cotton in states in the Southeast and Southwest experiencing high infestations of budworm resulted in a $20 per acre increase in net income. Another study that examined the use of herbicide‑resistant cotton in several Mid‑South states estimated producers saved $133 million annually in weed control costs. 45. The post‑1997 updates of the cost of production data assume the same technological coefficients as the 1997 survey – for example, pounds of seed per acre, the number of pesticide applications per acre, etc. Brazil correctly notes that the ERS/USDA updated COP data from 1997 show increased seed costs, which reflects the use of higher‑cost biotech seed.282 To the extent those inputs become more costly (for example, as biotech seed replaces conventional), cost increases are captured by the updating process through input price indexes. What is not captured is the cost savings from technological changes that alter the mix of production activities and inputs. New survey data will incorporate new technological coefficients as well as changes in such practices as direct pesticide costs, changes in tillage, application and cultivation trips, and handweeding. Many of the cost‑saving aspects of biotechnology or other new practices (no‑till farming) cannot be accurately captured by simply updating old cost data by price indices. Thus, relying on such updated cost data that reflects an outdated technological mix is in error. (b) to what extent do producers base planting decisions on their ability to cover operating costs but not whole farm costs? USA 46. As explained in some detail in the US Further Rebuttal Submission of 18 November, the agricultural economics profession is clear that short‑run production decisions are made based on the ability of a producer to cover his variable or operating costs.283 All economic models that attempt to capture supply response (producer planting behaviour) use variable costs in the equations, not total costs. Examples include the FAPRI baseline projections model (a variation of which was used by Dr. Sumner), the ERS baseline projections model, and the Economic Research Service’s FAPSIM model, the results of which are cited by Brazil.284 No economic model of which we are aware looks to total costs as the relevant costs for producer planting decisions. 47. One can do the same exercise as done by Brazil in paragraph 59 of its further rebuttal submission, but using the economically correct variable costs instead of total costs.285 Even using the technologically‑ and structurally‑dated cost‑of‑production data based off the 1997 ARMS survey, in all years except the extraordinary year of 2001, average market returns more than covered variable costs, allowing producers to earn a sufficient margin to pay off other fixed costs, a conventional agricultural business practice, as noted by Christopher Ward.286 Instead of a cumulative loss of $332.79 per acre over the 6‑year period as claimed by Brazil, producers had a cumulative net margin of $592.65 per acre. Clearly, if all years were like 2001, US cotton farmers would go out of business.287 But because most US cotton farmers regularly cover their variable costs – and then some – they can survive a year like 2001. Cumulative net returns ($ per acre)
Source: USDA, Economic Research Service, HREF="http://www.ers.usda.gov" 212. Brazil states in paragraph 37 of its oral statement that studies of Westcott and Price found that the effect of the programme on cotton is to add an additional 1 to 1.5 million acres during marketing years 1999-2001 and to suppress US prices by 5 cents per pound. Does the US reject these findings? Why or why not? 48. While an interesting "academic" analysis of the impacts of the US marketing loan programme, the Westcott and Price study Analysis of the US Commodity Loan Programme With Loan Provisions (Exhibit BRA‑222) is not relevant for the Panel’s assessment of the matter before it. In this study acreage decisions are based on an expected net returns which includes as the expected price term as the higher of the lagged market price or the loan rate plus additional marketing loan facilitated revenue. Since the period of analysis for the study is 1998 through 2005, rather than actual data, the authors used USDA’s 2000 baseline. This baseline incorporates actual data for years prior to marketing year 1998 and partial marketing year 1999 to make projections about prices and other factors for marketing year 1999 forward. Thus the study is based on projections except for MY1998.288 49. A panel, however, cannot base its findings on hypothetical market conditions instead of actual conditions. That is, Brazil must show that US domestic support has actually caused serious prejudice in a given year based on actual market conditions and not that under some assumed conditions US domestic support programmes have impacted prices. Since the study is based on projected prices, the analysis is not useful for the Panel in determining whether US support programmes have caused serious prejudice to Brazil. 50. Putting aside the issue that the study is based on projections and not actual market conditions for the 1999‑2001 period, the United States believes the study results overstate the impacts of the US marketing loan programme for two reasons: (1) the expectation of prices used and (2) the overstated additional marketing loan facilitated revenue. 51. As previously discussed, the authors used USDA’s 2000 baseline as the input into the USDA FAPSIM model. To represent farmers’ price expectations, the simulation uses lagged prices from the projections in the USDA’s 2000 baseline. The model uses the higher of lagged market prices or the loan rate plus additional marketing loan facilitated revenue (fixed at 14 cents per pound for cotton). Using the price projections in the USDA 2000 baseline, farmers would expect the marketing loan programme to kick in for the period 1999‑2001. The problem here is that the price expectations used by the model are not consistent with the price expectations the farmers actually held at the time of planting for prices at harvest. As the United States provided in its opening statement at the Second Meeting of the Panel with the Parties, the futures prices at the time of planting indicated that prices would be above the marketing loan rate during this period.
As the futures prices demonstrate, market expectations at the time of planting for marketing years 1999‑2001 was that prices would be above the marketing loan rate and hence no marketing loan benefits. Therefore, the marketing loan programme would not have had the impact this study found. 52. In addition to having the wrong expectations about price levels and the marketing loan programme being tripped, the study has overstated the potential additional market loan facilitated revenue that can be achieved. The authors used a fixed rate of 14 cents to represent this additional revenue above the loan rate when the marketing loan programme kicks in. Their justification for this figure is that this was their calculation for 1998. The authors do not provide any discussion as to why conditions in 1998 were indicative of conditions to continue for the near future that would keep this margin at 14 cents. This margin is based on the fact that farmers can pick the date to make the claim for the marketing loan gain/deficiency payment and then sell the cotton at a later date. The premise is that a farmer is able to sell when prices have increased relative to the date they made their claim. However, in reality, there is no such guarantee. It is just as possible that prices will fall below the price when the claim was made. In fact, the additional revenue has not been as large as in MY1998.289 In MY1999, the annual average was 6.1 cents, MY2000 was 5.8 cents, and MY2001 was 1.3 cents. As Exhibit US‑126 demonstrates, the margin fluctuates from month to month, with the value in several months even negative, implying that a farmer that did not sell his crop at the time he received the marketing loan payment earned less than the marketing loan rate. Using a much lower value for this additional revenue above the loan rate when the marketing loan rate kicks in would have reduced the impact of the study’s result. 53. By request of the Payment Limitations Commission, Westcott and Price updated this analysis using actual prices for MY2001. As the United States has discussed at the panel meetings and in its further rebuttal submission, this will overstate the impact of the marketing loan programme because it assumes that farmers had perfect foresight. That is, the model is calibrated to the actual values that occurred in that year while, in fact, producers could not have anticipated such events when planting decisions occurred. This overstates the effects of the programme because the model assumes outcomes that were unanticipated by producers when they made their planting decisions. 54. This overstatement is similar to the other third party studies that used actual outlays for marketing loans when calculating the price wedge. As the United States argued when critiquing those models, a more appropriate method to determine the impact is to look at expectations based on futures prices relative to the marketing loan rate. If the futures prices are above the loan rate, the programme will have a negligible impact on planting decisions since farmers are not expecting benefits from the programme. 55. For these reasons, the United States finds the results from these studies not relevant for the Panel in making its assessment of the effect of the marketing loan programme. Directory: english -> tratop e -> dispu e dispu e -> Annex g list of exhibits submitted by the parties dispu e -> World Trade Organization dispu e -> World Trade Organization Download 3.13 Mb. Share with your friends:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||