OIL CROPS YEARBOOK October 24, 2001 October 2001, ERS-OCS-2001 Approved by the World Agricultural Outlook Board --------------------------------------------------------------------------- OIL CROPS YEARBOOK is published annually by the Economic Research Service, U.S. Department of Agriculture, Washington, DC 20036-5831. This release contains only the text of the OIL CROPS YEARBOOK--tables and graphics are not included. Printed copies of this yearbook will be available from the USDA order desk. Call, toll-free, 1-800-999-6779 and ask for stock # ERS-OCS-2001, $21. ERS-NASS accepts MasterCard and Visa. -------------------------------------------------------------------------- Summary U.S. Soybean Situation, 2001/02 World Oilseed and Protein Meal Situation Situation for Other U.S. Oil Crops Cottonseed Peanuts Sunflowerseed Other Minor Oilseeds Other Fats and Oils Highlights Special Articles Soybean Production Costs and Export Competitiveness In the United States, Brazil, and Argentina Estimating Farm-Level Effects of Adopting Herbicide-Tolerant Soybeans Report Coordinator Mark Ash (202) 694-5289 E-mail: MASH@ers.usda.gov Principal Contributors Mark Ash (202) 694-5289 (Soybeans, Vegetable Oils) Erik Dohlman (202) 694-5308 (Peanuts) Mae Dean Johnson (202) 694-5299 (Statistics) Editor Martha R. Evans Approved by the World Agricultural Outlook Board. Summary released October 24, 2001. Summaries and full text of Situation and Outlook reports may be accessed electronically via the ERS website at www.ers.usda.gov/. To order call 1-800-999-6779 in the United States or Canada. Other areas please call (703) 605-6220. Or write ERS-NASS, 5285 Port Royal Road, Springfield, VA 22161. Summary Based largely on a record harvested area of 72.4 million acres, U.S. soybean production in 2000 increased 104 million bushels to a record 2,758 million bushels. With stronger demand for soybean meal in 2000/01, domestic soybean crushing climbed to a record 1,641 million bushels from 1,578 million the previous year. The lowest prices in 3 decades and the rapid pace of China's soybean imports helped lift foreign demand for U.S. exports in 2000/01 to 998 million bushels, surpassing the 1999/2000 record of 973 million. Consequently, record use of soybeans reduced U.S. ending stocks to 248 million bushels from 290 million in 1999/2000. The 2000/01 U.S. average farm price was $4.55 per bushel compared with $4.63 the previous season. Brisk soybean meal sales, particularly to the European Union (EU), Indonesia, Egypt, and South Korea, helped expand U.S. meal exports in 2000/01 to 7.6 million short tons. Gradually improving livestock-price-to-feed-cost relationships expanded 2000/01 domestic disappearance of soybean meal by 5 percent to 31.85 million tons. Despite a bumper South American soybean harvest, relatively weak growth in their output and exports of soybean meal helped support the U.S. season-average price at $173.60 per ton. A higher soybean crush pace in 2000/01 produced a record U.S. soybean oil output of 18,480 million pounds. Low soybean oil prices helped support steady growth in domestic disappearance of soybean oil to 16,350 million pounds from 16,056 million in 1999/2000. However, estimated U.S. exports only edged up to 1,400 million pounds from 1,376 million in 1999/2000. The large carryover and record output raised stocks throughout the year, which had swelled to a record 2,800 million pounds by September 30. The supply glut depressed soybean oil prices in 2000/01 (averaging 14.15 cents per pound) to their lowest level in 30 years. Virtually all of the increase in world oilseed production in 2000/01 was due to an 8-percent rise in global soybean production to 173.2 million metric tons, which derived from a record area in the United States and an almost ideal growing season in South America. The growth in China's domestic supplies was far exceeded by its import needs, which swelled from 10.5 million to 13.2 million tons. China's domestic consumption of soybean meal surged 19 percent to nearly 15 million tons. And after the European Union banned the use of meat and bone meal in all livestock feeds, EU soybean imports expanded from 15.7 million to 17.2 million tons in 2000/01 and soybean meal imports rose from 19.8 million to 20.2 million tons. Despite strong consumption growth, global ending stocks of soybeans rose to a record 28.5 million tons. Global output of soybean oil totaled 26.5 million tons, up 1.7 million tons from 1999/2000. However, world trade expanded just 0.4 million tons to 7.7 million tons in spite of a large increase in Indian soybean oil imports. Indian soybean oil imports increased from 0.8 million to 1.4 million tons in 2000/01 because of poor domestic oilseed harvests, strengthening consumption, and higher import tariffs on competing vegetable oils. A moderating influence on global soybean oil trade was the huge increase of China's domestic production, which sharply curtailed its 2000/01 soybean oil imports to just 80,000 tons from 1999/2000 imports of 556,000 tons. A severe drought caused poor U.S. cotton yields and substantial area abandonment in 2000, resulting in only a slight increase in cottonseed output to 6.4 million short tons. Weak prices for cottonseed oil diminished crushing margins for cottonseed, reducing crush to 2.7 million tons in 2000/01 from 3.1 million in 1999/2000. But, consumption of cottonseed in livestock feed and planting seed increased 7 percent to 3.75 million tons. Firmer feed prices generally supported the 2000/01 average cottonseed price at $106 per ton. U.S. peanut production fell sharply in 2000, down 563 million pounds to 3,266 million. Although national plantings were up slightly, extremely dry conditions in many areas of the Southwest (New Mexico, Oklahoma, and Texas) during the growing season raised abandonment and cut national harvested acreage to 1.32 million acres, down 8 percent from 1999. As in 1999, poor weather also cut the 2000 average peanut yield, which fell to 2,444 pounds per harvested acre, down 223 pounds per acre (8 percent) from 1999, and the lowest yield since 1995. Domestic food use of peanuts weakened in 2000/01, falling to 2,170 million pounds from 2,233 million in 1999/2000. Ample foreign harvests and a tighter available U.S. supply reduced U.S. peanut exports to 520 million pounds from 727 million in 1999/2000, and the smallest since 1980/81. U.S. sunflower harvested acreage in 2000 fell to 2.6 million acres, down nearly one-fourth from 1999. Domestic sunflowerseed production in 2000 was 3,544 million pounds, down from the previous year's 4,342 million. Sluggish European demand curtailed 2000/01 oil-type sunflowerseed exports to just 45 million pounds. In contrast, U.S. confectionery seed exports increased (to the EU in particular) to a record 400 million pounds, in spite of a shorter domestic supply. But estimated sunflowerseed oil exports, which make up the majority of crushers' sales, fell again in 2000/01 to 575 million pounds. Very low soybean oil prices pressured U.S. season-average prices for sunflowerseed and sunflowerseed oil to $6.80 per hundredweight and 16.2 cents per pound, near historic lows. U.S. Soybean Situation, 2000/01 Despite Strong U.S. Soybean Demand, Huge Foreign Crops Squelch 2000/01 Prices U.S. farmers planted a record 74.3 million acres of soybeans in 2000, up 1 percent from 1999. Stronger corn prices and optimal planting conditions in the spring limited the expansion of soybean area, and plantings were 605,000 acres less than farmers' March intentions. Nearly all of the expansion was in the Northern Plains and Lake States, where crop rotations are still adjusting to incorporate more soybeans. Farmers in Minnesota, Michigan, Wisconsin, Nebraska, Kansas, South Dakota, and North Dakota each set State records for soybean planting. In the spring of 2000, warm and firm soils advanced U.S. soybean planting well ahead of average. Throughout the eastern Corn Belt, relatively mild temperatures and regular precipitation early in the summer ensued, generating soybean yields that neared or equaled records. However, on the western and southern edges of the soybean belt, soybeans struggled through one of the hottest and driest summers of the last century. The worst yield damage occurred in Kansas, Nebraska, and the lower Mississippi River Valley. Although the national soybean yield improved from 36.6 to 38.1 bushels per acre, it was still suppressed well below the trend. Based largely on a record harvested area of 72.4 million acres, U.S. soybean production in 2000 increased 104 million bushels to a record 2,758 million bushels. However, lower beginning stocks raised total supplies just 45 million bushels from 1999/2000. After economic difficulties closed several U.S. crushing mills in 2000 (causing a 6-percent reduction in operating capacity), soybean processing margins generally improved. And, with stronger demand for soybean meal in 2000/01, domestic soybean crushing climbed to a record 1,641 million bushels from 1,578 million the previous year. The rapid pace of China's soybean imports helped lift foreign demand for U.S. exports in 2000/01 to 998 million bushels, surpassing the 1999/2000 record of 973 million. Consequently, U.S. ending stocks of soybeans fell to 248 million bushels from 290 million in 1999/2000. Yet, this expansion of the demand base was possible only with the lowest soybean prices in three decades. The 2000/01 U.S. average farm price was $4.55 per bushel compared with $4.63 the previous season. To supplement their cash receipts from crop year 2000 soybeans, U.S. producers received $2.54 billion of marketing loan benefits (averaging 93 cents per bushel) on 2,730 million bushels. Producers also received additional income in crop year 2000 from Oilseed Program payments. In June 2000, the government authorized $500 million in Oilseed Program payments for producers of 2000-crop soybeans and minor oilseeds (Public Law 106-224). Payments from this authorization were made to producers in February 2001. In August 2001, as part of an emergency package to provide market loss assistance for the 2001 crop, the government authorized an additional $423.5 million for supplemental Oilseed Program payments (Public Law 107-25). Soybean producers received $475 million in February payments and $404 million in August supplemental payments. Strong Meal Demand Spurs Crushing A record 39.4 million short tons of soybean meal was produced in 2000/01 because of a robust demand situation. Domestic consumption of soybean meal grew solidly, as livestock-price-to- feed-cost relationships gradually improved. U.S. hog slaughter rates slowed and import restrictions on EU pork exports favored U.S. exports. Thus, hog prices rallied by about one-third between fall 2000 and spring 2001. In addition, soybean meal use increased because of a very cold winter and a substantial drop in wheat supplies that limited summer wheat feeding. As hog slaughter rates slowed, higher feeding rates contributed to a 5- percent expansion of U.S. soybean meal disappearance for 2000/01 to 31.8 million tons. Brisk soybean meal sales, particularly to the European Union (EU), Indonesia, Egypt, and South Korea, helped expand U.S. exports in 2000/01 to 7.6 million short tons. A shortage of protein feeds following the 6-month EU ban on meat and bone meal boosted European soybean meal demand. EU agriculture ministers subsequently agreed to extend the ban beyond its original June 30 expiration. The EU also prohibited meat and bone meal exports, of which Indonesia had formerly been a major importer. U.S. soybean meal exports also benefited from an Indonesian ban on imports from South America, India, and China, which was attributed to fears of transmitting the foot and mouth disease that was present in those countries. Central Illinois prices for high-protein soybean meal surged from an October 2000 average of $171.50 per short ton to a December peak of $196. But by the first quarter of 2001, large expected South American crops and an uncertain outlook for feed demand in Europe (due to incidences of foot and mouth disease in several countries) weighed on prices. Prices slid to $156 per short ton in March, the lowest level since December 1999. Yet steady demand growth re-emerged later in the spring and the seasonal crush decline prompted a recovery in soybean meal prices. For the entire 2000/01 season, the average U.S. price was $173.60 per ton. Despite bumper South American soybean harvests, relatively weak growth in their soybean meal output also supported prices. Poor Soybean Oil Exports Swell Stocks, Prices Plunge In contrast to the strength of the soybean meal market, weak soybean oil demand limited the expansion of soybean crushing. The share of soybean oil to total processing value fell to an all- time low of 27 percent in 2000/01. Despite a higher crush pace in 2000/01, a lower oil extraction rate limited growth in soybean oil output to 18,434 million pounds compared with 17,825 million in 1999/2000. With very low oil prices, soybean processors sacrificed maximum oil yields to accelerate delivery of the more profitable meal to its buyers. Estimated domestic disappearance of soybean oil rose steadily to 16,240 million pounds from 16,056 million in 1999/2000. In spite of the low soybean oil prices, export gains were hard to make. U.S. exports of soybean oil increased to Canada, India, Pakistan, and North Africa but were offset by sharp declines to China, Mexico, Turkey, and South Korea. Total exports did not revive until near the end of the season, only edging up to 1,400 million pounds from 1,376 million in 1999/2000. The large carryover and record output raised stocks throughout the year, which had swelled to a record 2,865 million pounds by September 30. The supply glut depressed soybean oil prices to their lowest level in 30 years, averaging 14.15 cents per pound in 2000/01. Soybean oil prices bottomed out at 12.4 cents in February 2001, which was 2.7 cents below February 2000. By the summer, a seasonal weakening of the crush pace, lower than expected soybean stocks, declining prospects for 2001 soybean production, and a strengthening of world palm oil prices buoyed the value of soybean oil again. World Oilseed and Protein Meal Situation Global Soybean Output and Trade Surges World oilseed production in 2000/01 increased 2 percent to 310.7 million metric tons. Virtually all of the increase was due to an 8-percent rise in global soybean production to 173.2 million tons. The increase in soybean production derived from a record area in the United States and an almost ideal growing season in South America. Despite growth in consumption, global ending stocks of soybeans rose to a record 28.5 million tons. Although U.S. soybean stocks declined, they were offset by an accumulation in Argentina and China, which kept the pressure on world prices. China was the world's leading soybean importer in 2000/01, which alone accounted for 46 percent of the 6.7-million-ton growth in world soybean imports. In contrast, world soybean meal imports were up only 2 percent in 2000/01, as stronger European meal imports were partly offset by lower Asian imports. In Brazil, a shortage of corn supplies strengthened prices relative to soybeans, encouraging substantial interest in domestic corn production. Cotton area also increased. The December rally in soybean prices came too late to permit a larger expansion of Brazilian area, so they only edged up 3 percent to 14.0 million hectares. Soybean planting began considerably earlier in 2000 than a year earlier when dry weather delayed progress. Generous rainfall was very favorable for soybeans in the major producing areas of the south and center-west, although drought cut yields in the northern states. Based on slight increases in area and yield, Brazil's 2001 soybean harvest grew to a record 38.4 million metric tons. Between January and September 2001, Brazil's exchange rate depreciated about one-third, which made its crushers' exports more competitive against Argentina's, whose peso was pegged to a strong U.S. dollar. A recovery in yields in the three southernmost states also aided the crushing industry, as nearly 60 percent of processing capacity is concentrated there. Yet, because of limited foreign demand for soybean meal, Brazil's 2000/01 soybean meal exports only edged up to 10.25 million tons. In addition, on June 1 the government began a program to ration Brazilian electrical power use by up to 20 percent. Hydroelectric utilities produce 90 percent of Brazil's electricity, and drought critically depleted the main reservoir levels. Operations by domestic soybean processors in the affected areas were disrupted and the 2000/01 crush rose only modestly to 22.0 million tons. A weak economy slowed domestic soybean meal consumption to 7.3 million tons from 7.2 million in 1999/2000. With a soybean carryover slightly larger than the previous season, greater supply availability helped boost Brazilian exports to a record 15.0 million tons in 2000/01. In spite of higher demand, Brazil's soybean stocks ending September 2001 remained at a relatively high 7.6 million tons, which exceeded the stocks held in the United States. Some Brazilian farmers held off marketing their new crop in anticipation of further weakening of the exchange rate and weather problems in the United States. A bumper Argentine soybean crop was possible in 2000/01 because of a reduction in sunflower area. Also, the rise in soybean prices during planting encouraged more double cropping after winter wheat, swelling total harvested area by 20 percent to 10.3 million hectares. Abundant March-April rains also helped produce excellent yields. Argentine soybean farmers produced a record harvest of 26.5 million tons in 2000/01, compared with 21.2 million a year earlier. By itself, Argentina accounted for 40 percent of the world's expansion of soybean production in 2000/01. However, Argentina's 2000/01 soybean crushing still only edged up to 17.7 million tons from 17.1 million the previous year. Processing declined sharply between October 2000 and April 2001, but accelerated somewhat in May-September 2001 following the newly harvested crop. Despite strong European demand, Argentine processors were hurt by their exclusion from Indonesian purchases. Indonesia temporarily implemented a ban on purchases from countries where foot and mouth disease was present. The disease was confirmed in Argentina in March 2001. Argentine soybean meal production in 2000/01 was limited by a slower crush pace, and meal exports increased only slightly to 13.9 million tons. Conversely, with much of current world demand being propelled by China's soybean imports, Argentine soybean exports surged to 7.2 million tons. Paraguay's soybean-producing region experienced excellent growing conditions in 2001, similar to the neighboring areas of southern Brazil. Paraguayan rainfall was above normal in January-February, and relatively dry weather during March generally favored soybean harvesting. Paraguay's 2001 soybean production was a record 3.4 million tons from a harvested area of 1.25 million hectares. Consequently, Paraguay expanded its 2000/01 soybean exports to 2.5 million tons. India's soybean harvest dropped again in 2000 to 5.25 million tons because of a weak start and early retreat of the summer monsoon. The shortage reduced Indian soybean meal exports 300,000 tons in 2000/01 to 2.05 million. Soybean meal exports were also slowed after late January 2001 by a severe earthquake in western India that damaged the port of Kandla, which ships about two- thirds of India's exports. Meal stocks were stranded at the port and it took several months to fully restore operating capacity. While some deliveries of soybean meal were rerouted through other ports, the disruption forced some crushers to temporarily shut down. The lack of a good conduit for exports reduced soybean meal prices and encouraged an increase in domestic meal consumption to 1.5 million tons. In China, comparatively stronger prices in 2000 shifted another 1.3 million hectares of farmland into soybean production, mostly in lieu of corn. Drought cut yields in northern China and moderated the gain in domestic output from 14.3 million tons to 15.4 million. Yet, the growth in China's domestic supplies was far exceeded by its import needs, which swelled from 10.5 million tons to 13.2 million. Domestic consumption of soybean meal surged 19 percent to nearly 15 million tons. By early 2002, China's entry into the World Trade Organization (WTO) will begin to reduce import barriers for oilseed products. But during 2001, China's accession to the WTO stalled pending resolution of several outstanding issues. China subsequently agreed to limit its agricultural subsidies to 8.5 percent of the value of agricultural production, which settled one of the last hurdles to WTO accession. However, the delay in implementing lower oilseed product tariffs helped preserve crush margins for China's soybean processors. As world soybean meal prices and domestic consumption strengthened, China's 2000/01 soybean crush soared 23 percent to 18.6 million tons. As a consequence, soybean meal imports (most of which came from India) were very slow, totaling just 100,000 tons compared with 633,000 in 1999/2000. Despite the rapid consumption rate, many of the higher soybean imports added to China's ending stocks, which increased to an estimated 5.1 million tons. Chinese buyers may have desired larger stocks because a smaller 2001 soybean area and a spring drought in northern China threatened the potential of the domestic 2001 crop. Soybean purchases may also have been advanced by uncertainty about how China's Government would ultimately administer new regulations on imports of biotech crops. International prices and interest rates were low, and Chinese storage capacity was abundant. Late last year, the European feed market was thrown into turmoil by new cases of bovine spongiform encephalopathy, popularly known as mad cow disease. The spread of this disease has been linked to the inclusion of infected tissues in meat meal used in cattle feed. Scientists believe it is possible that consuming infected beef can transmit a similarly fatal, brain-damaging disease to humans. The crisis caused EU beef sales to collapse. Among the measures taken by the EU Farm Commission to limit the disease's spread and restore consumer confidence in the safety of consuming beef, was adopting a ban on the use of meat meal in all livestock feeds. The resulting protein meal deficit in Western Europe had to be managed through imports. Climate limits the feasible area that soybeans can be grown in Europe, and EU oilseeds production for 2000 had already dropped by 14 percent. The major beneficiary of the meat meal ban was soybean meal consumption, which increased 4 percent in 2000/01 compared with a 7-percent decline the previous year. The relative protein content of soybean meal compared to meat meal helped drive this growth in consumption. Meat meal has a high protein content (50-55 percent) compared with soybean meal (44-48 percent). Another alternative was fish meal, which was excluded from the overall ban, but is still restricted in ruminant feeds. During fall and winter, stronger soybean meal prices in Rotterdam favored soybean imports from the United States. The oil supplies generated by crushing soybean imports also better compensated for the loss of fat supplies available for feed and food applications. Both France and Germany restricted use of beef tallow in livestock feeds. Price slippage in the spring accelerated soybean meal imports from South America. EU soybean imports expanded from 15.7 million tons to 17.2 million in 2000/01, while soybean meal imports rose from 19.8 million tons to 20.2 million. The EU meat meal ban was countered throughout 2001 by the relative weakness of the euro. EU feed compounders have been encouraged to substitute more dollar-denominated imports of soybeans and soybean meal with domestically produced grains, as reforms of the EU's Common Agricultural Policy allowed internal grain prices to decline further. The EU also prohibited exports of EU meat meal to prevent their inclusion in feeds in Eastern Europe and elsewhere. With trade exceeding 500,000 tons per year, the EU was formerly the world's largest exporter of meat and bone meal. But many countries, even after the EU export ban, prohibited imports of meat and bone meal from any country, regardless of origin. One major market to do this was Poland, which in calendar 2000 was importing nearly 300,000 tons of meat and bone meal. Consequently, 2000/01 soybean meal imports by all of Eastern Europe increased by 230,000 tons to 2.8 million. Demand for soybeans and soybean meal by several Middle East and North African nations has expanded strongly in recent years. For 2000/01, soybean imports by Iran were up 130 percent from 2 years earlier and Egyptian imports nearly doubled. Greater domestic crushing moderated imports of soybean oil and meal in both countries. In Tunisia and Algeria, where there are no domestic crushing facilities, there also were strong gains in consumption and imports of soybean meal. An exception to improved soybean demand in the Middle East was Turkey. After February 22, when the Turkish Government allowed the country's currency to float, its lira lost about one-half of its value relative to the U.S. dollar. These events suddenly and sharply increased the cost of all Turkey's imported goods, including agricultural products. Over the last decade, Turkey became a significant importer of soybean meal and soybeans. But the country's weaker economy and more expensive feeds stalled poultry demand in 2000/01. Turkish soybean meal imports fell to 450,000 tons from 509,000 in 1999/2000. Weak Oil Prices Stifle Production and Trade of Rapeseed, Sunflowerseed Smaller European, Canadian, and Australian crops in 2000/01 reduced world rapeseed production by 4.9 million tons to 37.5 million. Given abundant alternate sources of vegetable oil, EU rapeseed prices at planting time had sunk about 30 percent from a year earlier. In addition, direct payments to EU oilseed producers were cut over the next 3 years by Agenda 2000 reforms to equal the area payment for grains. The low rapeseed prices and declining subsidies decreased EU rapeseed area 13 percent in 2000 to 3.1 million hectares. EU rapeseed yields suffered in 2000 because drier than normal spring weather was followed by excessive harvest-time rain in northern France and western Germany. These factors combined to slash the 2000 EU rapeseed crop to 9.1 million tons from 11.4 million the previous year. Eastern European producers responded to the same price incentives by reducing rapeseed plantings by 15 percent, lowering production to 2.2 million tons from 2.6 million in 1999/2000. World trade in rapeseed (which has a heavy European orientation) fell 11 percent in 2000/01 to 9.6 million tons. Similarly, relatively weak vegetable oil prices and a large stock carryover encouraged Canadian farmers in 2000 to shift more area from canola into barley and durum. Canola area harvested fell 13 percent in Canada to 4.8 million hectares. Cool temperatures slowed development and a late spring frost in Alberta forced some replanting in June. Canola production dropped back from a record 8.8 million tons in 1999 to 7.1 million. Canadian 2000/01 exports increased to 4.8 million tons, cutting ending stocks of canola about in half from the previous year. Domestic crushing increased negligibly. The rapid expansion in Australian rapeseed area of recent years was interrupted in 2000/01. Superior returns for barley slashed harvested rapeseed area to 1.3 million hectares, down one-third from the 1999/2000 record. The area reduction cut Australian rapeseed production to 1.7 million tons, compared with 2.4 million in 1999/2000. Exports and domestic crushing fell accordingly. In India, a comparatively better profit outlook for wheat and persistence of last year's dryness in the north reduced rapeseed area from 5.6 million to 5.0 million hectares. A continuation of the drought depressed yields, cutting India's 2000/01 rapeseed crop to 3.7 million tons from 5.1 million. Among major rapeseed producers, China was an exception to the crop reductions in 2000, as farmers there harvested a record 7.5 million hectares. The 12-percent expansion in area stemmed from strong internal crushing demand and a withdrawal of government support for winter wheat and early rice. A mild winter, adequate moisture, and improved varieties raised China's 2000 rapeseed production 12 percent to 11.4 million tons. A larger domestic harvest, smaller foreign crops, and larger soybean oil supplies reduced China's rapeseed imports to 2.4 million tons from 3.7 million in 1999/2000. Global sunflowerseed production in 2000/01 declined to 22.9 million tons from 27.2 million in 1999/2000. Area reductions in the EU and Argentina and drought in Eastern Europe were largely responsible. Weak demand in Western Europe, brisk crushing of domestic harvests in the former Soviet Union, and a small Argentine crop stalled global sunflowerseed exports near 3.5 million tons. Scaled down planting and drought slashed 2000-crop sunflowerseed production in Romania, Turkey, Hungary, Bulgaria, and Yugoslavia. Lower supplies in each of these countries rationed crushing and cut their sunflowerseed exports dramatically. The Russian harvest declined only moderately to 3.9 million tons, compared with the bumper 1999 crop of 4.2 million. Despite a 20-percent reduction in Russian sunflowerseed area, yields were quite good. In contrast, larger sunflowerseed crops in Spain and Ukraine helped offset reductions elsewhere. Spanish output recovered from severe drought in 1999, although fewer plantings in France and Italy moderated the increase in EU sunflowerseed output. Similarly, good yields pushed Ukraine sunflowerseed output up to 3.5 million tons from the previous harvest of 2.7 million. Continuation of an export tax on Ukrainian sunflowerseed exports kept domestic crushing high. Larger sunflowerseed oil output by Ukrainian and Russian processors also eroded traditional Argentine markets. Argentine sunflower area declined sharply in 2000/01 to 1.9 million hectares from 3.5 million because of much lower vegetable oil prices. In addition, a dry spell during a pivotal stage in late February in western Buenos Aires and La Pampa (the provinces where the majority of Argentina's sunflowers are raised) hurt yields. Excess rainfall in March and April delayed sunflower harvesting and extended the crop damage. Consequently, Argentine sunflowerseed output plunged to 3.2 million metric tons, the smallest since 1993/94 and down nearly one-half from a year earlier. Soft European demand also undermined Argentine crushing and exports of sunflowerseed, which slumped to 3.5 million and 120,000 tons, respectively. Global cottonseed production edged up 0.5 million tons in 2000/01 to 33.4 million. The gain was mainly due to much larger harvests in China and Brazil that offset smaller losses in India, Pakistan, and other countries. However, greater feed usage of cottonseed trimmed global crush to 24.5 million tons, which lowered 2000/01 supplies and consumption of cottonseed oil and meal. A larger peanut crop in China was the major reason that global peanut production increased 7 percent in 2000/01 to 31.2 million tons. China's peanut area expanded 14 percent to 4.9 million hectares, raising its output to a record 14.4 million tons. The bumper crop allowed a substantial increase in peanut crushing in China. However, like 1999, India's peanut harvest was again damaged by drought, allowing only a small increase from 5.5 million to 5.7 million tons based on a slightly higher area. Expanded crushing in both countries (which normally account for about 80 percent of the world total) raised global production of peanut meal output by 4 percent to 5.5 million tons. World Vegetable Oil Situation World vegetable oil production expanded to 88.5 million metric tons in 2000/01 from 85.9 million. Global output of soybean oil totaled 26.5 million tons, up 1.7 million tons from 1999/2000. However, in spite of a large increase in Indian soybean oil imports, world trade expanded just 0.4 million tons to 7.7 million tons. Exports increased minimally for Argentina, Brazil, and the United States. Part of the reason was that a huge increase in China's domestic soybean oil production sharply curtailed its 2000/01 soybean oil imports to just 80,000 tons from 1999/2000 imports of 556,000. China has not yet gained accession to the World Trade Organization, which also limited issuance of the country's oil import quotas. Palm Oil Exporters Seek Markets For Bumper Output World palm oil output for 2000/01 expanded to 23.6 million tons, up from 21.8 million in 1999/2000. Malaysian palm oil production in 2000/01 surged from 10.5 million tons to 11.9 million. Similarly, Indonesian output expanded to 7.6 million tons from 7.2 million based on a larger area of mature trees that were planted about 5 years ago. Early in 2000/01, low producer prices encouraged brisk demand for palm oil by many of the major importers, including India, the European Union, China, and the Middle East. World trade surged to 16.5 million tons, compared with 14.3 million in 1999/2000. As palm oil output slowed in the latter stage of the season, global ending stocks dipped to 2.8 million tons from the 1999/2000 record of 3.0 million. When palm oil surpluses accumulate, the area devoted to oil palm plantations will not decline significantly, unlike land planted to annual oilseed crops. Once land is cleared and trees are planted, palm oil producers have quite low marginal production costs. Low prices hurt producers' short-term profits, reduce the fertilizer and labor expenses incurred, and defer the expansion of new plantations. But existing palm plantations resist taking productive trees out of production. With no immediate supply response, palm oil prices must fall sharply to clear the market, so they can expand market share at the expense of rival vegetable oil producers until prices improve. In February 2001, Malaysian crude palm oil prices bottomed out at $193 per metric ton, down 37 percent from a year earlier. Malaysian palm oil exports expanded by 1.7 million tons in 2000/01 to 10.55 million. But exports had slowed following an April 1 tariff hike by India, which stalled the decline in stocks. India is Malaysia's largest foreign market for palm oil. Also in early 2001, Indonesia's unstable political environment and ethnic violence caused its foreign exchange rate to fall to its lowest value since the 1998 financial crisis. Growing supplies and currency depreciation intensified the pressure on palm oil producers to export and increased the competition with Malaysia. Indonesian palm oil exports climbed to 4.45 million tons from 4.0 million in 1999/2000. In response to the continuing high level of stocks, Malaysia's Government developed schemes to subsidize replanting of 200,000 hectares of older palm trees and burn 600,000 tons of palm oil in electric power plants. These measures were intended to support short-term prices by reducing annual supplies more than 1 million tons. However, Malaysia has few diesel-fueled electric generators and burned just 100,000 tons of palm oil. Indonesia's Government pledged cooperation with Malaysia's goals of lowering foreign import barriers and supporting prices. On March 1, Indonesia implemented a higher export tax on crude palm oil and lowered its refined palm oil tax to help develop its domestic refining industry. Later in the year, a peaceful change of government helped strengthen the exchange rate. Even Thailand, a minor palm oil producer, promoted domestic use with tax waivers on biodiesel. Yet, a cyclical slowing of output provided the most effective boost to palm oil prices, which rose to $337 per ton by August 2001 and cooled interest in the stock-reduction plans. Indian Import Tariffs Reorient World Vegetable Oil Trade India's oilseed production for 2000/01 was estimated at 20.7 million tons, down 7 percent from the previous year and 17 percent from 2 years earlier. The shortfall of domestic oilseed crushing has contributed to a rising tide of Indian vegetable oil imports, which alone comprised 18 percent of the world imports in 2000/01. The Government of India tried to aid domestic oilseed farmers by again raising import tariffs on vegetable oil. India's last tariff hikes on April 1, 2001, raised import duties on crude palm oil and crude sunflowerseed oil to 75 percent and refined palm oil to 85 percent. India maintains higher tariffs for refined oils to promote its domestic refining industry. However, poor domestic oilseed crops, steeply discounted prices, and robust consumption growth nullified the impact of higher palm oil tariffs. Indian palm oil imports still expanded to a record 4.2 million tons. In contrast, with the 1994 Uruguay Round Agreement, India's market access commitments eliminated all quantitative restrictions on imports and bound the maximum tariff on soybean oil imports at 45 percent. Soybean oil imports quickly became the preferred source for covering India's acute vegetable oil deficit. Border prices for palm oil were discounted $75-$80 per ton to soybean oil, but the Indian tariff differential nearly equalized domestic prices of palm oil and soybean oil. By August 2001, the price discount for Malaysian palm oil against U.S. soybean oil had narrowed to about $40 per ton. Indian consumption is very price sensitive, so a small change in that price wedge can cause a substantial substitution between vegetable oils. Indian soybean oil imports increased from 0.8 million to 1.4 million tons in 2000/01. India typically imports soybean oil between May and September, so Argentina and Brazil usually supply most of them. U.S. soybean oil exporters had a negligible share of this trade, but they may soon benefit competitively in other foreign markets as India absorbs more supplies from South America. Conversely, lower available sunflowerseed oil supplies from Argentina and higher import duties made them less price competitive for India. India's 2000/01 sunflowerseed oil imports dropped to 475,000 tons from 570,000 tons in 1999/2000. When India's partial retreat from the market created favorably low international prices, other countries were encouraged to expand their 2000/01 imports. European Union imports of palm oil increased 9 percent to 2.9 million tons. China does not produce palm oil but is a leading consuming country. Palm oil imports by China swelled to 1.8 million tons from 1.2 million in 1999/2000. Import quotas were more available for palm oil than for soybean and rapeseed oil, which help protect China's oilseed processors. Both the gains in world trade for soybean oil and palm oil were aided by lower quantities of sunflowerseed oil and rapeseed oil. World sunflowerseed oil output in 2000/01 declined 1.1 million tons to 8.5 million and exports (mainly by Argentina) plummeted 0.8 million tons to 2.9 million. Likewise, global production of rapeseed oil fell 0.5 million tons in 2000/01 to 13.1 million, and exports dropped 0.3 million tons to 2.6 million. Situation for Other U.S. Oil Crops Cottonseed Feed Use of Cottonseed Eclipses Crush Demand U.S. farmers planted 15.6 million acres of cotton in 2000, the second largest area ever. However, a repeat of late-summer heat and drought resulted in substantial area abandonment. The adverse weather also produced below-average dryland cotton yields in Texas and the Delta, although these yields were somewhat better than in 1999. Factoring in an improved seed-to-lint ratio, 2000 cottonseed output increased only slightly to 6.4 million short tons. However, there was no serious deficit in cottonseed supplies because of a record volume of imports from Australia. In fact, ending stocks of cottonseed rose to 424,000 tons because of a sharp reduction in crushing. Oil processors used 2.7 million tons of seed in 2000/01, down from 3.1 million in 1999/2000. Weak prices for cottonseed oil diminished crushing margins for cottonseed. Ample supplies of competing oils cut 2000/01 domestic disappearance of cottonseed oil 21 percent to 657 million pounds, the lowest usage since 1986/87. Similarly, demand for cottonseed oil exports remained slow, sliding to 125 million pounds from 141 million in 1999/2000. But, consumption of cottonseed in livestock feed and planting seed increased 7 percent to 3.75 million tons. U.S. cottonseed exports (mostly to Mexico) also edged up from 198,000 tons to 235,000. Firmer feed prices generally supported the 2000/01 average cottonseed price at $106 per ton. The Government also approved $84.7 million of supplemental payments to U.S. producers and first handlers of the 2000 cottonseed crop. Payments were made to cotton ginners who distributed them to producers. The average payment rate was approximately $13 per ton of cottonseed. Peanuts On December 13, 1999, the U.S. Department of Agriculture (USDA) announced a national peanut poundage quota for the 2000 marketing year of 1.18 million short tons (2.360 billion pounds), the same as for the 1999 marketing year. The quota equaled the estimated quantity of peanuts needed for domestic edible and related uses, excluding seed, in the 2000 marketing year and allowed for potential underdeliveries of up to 18,500 short tons. The national average support price for 2000/01 quota peanuts was announced as $610 per short ton. The support price for additional peanuts was $132 per short ton. Both were unchanged from 1999. Following losses on the 1999 crop, producers were spared from a higher marketing assessment in 2000 when Congress authorized payment of the 1999 losses with assessments from previous years and future assessments to be collected through 2002. 2000 Peanut Plantings Up but Production Declines U.S. peanut area planted in 2000, at 1.54 million acres, was up 1 percent from plantings in 1999. Although national plantings were up slightly, extremely dry conditions in many areas of the Southwest (New Mexico, Oklahoma, and Texas) during the growing season raised abandonment and cut national harvested acreage to 1.32 million acres, down 8 percent from 1999. For the 2000 crop year, 205,000 planted acres were not harvested compared with 84,000 acres in 1999. As in 1999, poor weather also cut the 2000 average peanut yield, which fell to 2,444 pounds per harvested acre, down 223 pounds per acre (8 percent) from 1999, and the lowest yield since 1995. As a result, U.S. peanut production fell sharply in 2000, down 563 million pounds to 3,266 million. Output fell by 304 million pounds in the Southwest and 292 million pounds in the Southeast (Alabama, Florida, Georgia, and South Carolina), but production from the Virginia-North Carolina region was up 6 percent from 1999. Despite the harvest problems, the peanut buyback provision met quota needs by converting 224 million pounds of additionals for domestic use. Although less than 5 percent of total supply, peanut imports in 2000 rose to 211 million pounds (farmer stock equivalent). Imports increased 33 million pounds from the year before and were the largest since the record 1980/81 imports of 401 million. The smaller U.S. harvest mostly reduced the ending stocks from a quite large 1999/2000 carryover of 1,233 million to 1,116 million pounds. But lower demand offset part of the decline in supply, and the 2000/01 national average farm price for peanuts strengthened only slightly to 25.7 cents per pound from 25.4 cents the previous season. The smaller crop reduced the farm value of the 2000 peanut crop to $839 million, off $134 million from 1999, and the lowest farm value since 1983. U.S. Food Use of Peanuts Weakens Domestic food use of peanuts waned in 2000/01, falling to 2,170 million pounds from 2,233 million in 1999/2000. However, overall food use in 2000 was still relatively strong compared with the 1990-99 average of 2,095 million pounds. Rising peanut butter imports partly contributed to the drop in peanuts processed for domestic use. Individual categories of primary product use in 2000/01 saw consumption of peanuts for peanut butter (the major use in the United States) fall more than 2 percent to 753 million pounds. Consumption for snack peanuts dropped more than 8 percent to 362 million pounds, and 'other' use fell by 1 percent to 20 million pounds. Peanut use for candy was virtually unchanged at 355 million pounds, while in-shell peanut use was up by nearly 5 million pounds to 150 million. Lower Exports and Domestic Crush Contribute To Decline in Peanut Disappearance Ample foreign harvests and a tighter supply of export additionals reduced U.S. peanut exports to 520 million pounds from 727 million in 1999, and the smallest since 1980/81. On an in-shell basis, peanuts crushed in 2000/01 totaled 548 million pounds, 23 percent below the previous season and 29 percent below the 1990- 99 average of 774 million pounds. With the reduced crush, U.S. peanut oil production fell to 179 million pounds, down from 229 million pounds the previous year. Nevertheless, a large jump in peanut oil imports--to a record 79 million pounds--helped raise domestic consumption of peanut oil to the highest level since 1975. Most peanut oil imports came from Argentina. Peanut oil prices declined to 32.2 cents per pound, down from 33.6 cents per pound in 1999 and their lowest level since 1992. Peanut meal production also declined, from 146,000 short tons in 1999 to 115,000 short tons in 2000. Ending stocks remained unchanged at a low level of 2,000 short tons and exports declined from 6,500 to 5,700 short tons. Peanut meal prices rebounded slightly from the previous year, up 4 percent to a season-average $153.60 per short ton. Sunflowerseed Despite Smaller Crop, Poor Crush Demand Slashes Sunflowerseed Prices Farmers in the Northern Plains switched from sunflowers to soybeans, canola, and flax in 2000 because of fears that the sunflower disease that afflicted the 1999 crop could return. Thus, U.S. sunflower harvested acreage fell to 2.6 million acres, down nearly one-fourth from 1999. Oil-type sunflowers accounted for most of the decline in total harvested area, falling 0.6 million acres as confection varieties declined another 0.2 million acres. While North Dakota comprised 60 percent of the reduction in sunflower acreage, planting declined in other States, as well. The recovery from a low 1999 sunflowerseed yield moderated the decline in production to 3,544 million pounds from the previous year's 4,342 million. Sluggish European demand curtailed oil-type sunflowerseed exports to just 45 million pounds. In contrast, U.S. confectionery seed exports increased (to the EU in particular) to a record 400 million pounds, in spite of a shorter domestic supply. Domestic use of confectionery sunflowerseed fell sharply to 363 million pounds, keeping ending stocks at a relatively high 159 million. With sharply lower seed supplies and falling oil prices, domestic sunflowerseed crushers were in no stronger position than foreign processors. Sunflowerseed oil exports, which make up the majority of crushers' sales, fell again in 2000/01 to 575 million pounds. Greater supplies of mid-oleic varieties (NuSun) supported domestic sunflowerseed oil consumption equal to the previous season's level of 385 million pounds. But very low soybean oil prices pressured U.S. prices for sunflowerseed and sunflowerseed oil near historic lows of $6.80 per hundredweight and 16.2 cents per pound, respectively. Producers of oil-type and confection sunflowerseed received $132.2 million of marketing loan benefits for the 2000 crop. In addition, producers received Oilseed Program payments of $12.7 million in February 2001 and supplemental payments of $10.8 million in August 2001. Other Minor Oilseeds Canola plantings swelled 46 percent in 2000 to a record 1.6 million acres. Combined with relatively normal yields, the additional acreage pushed U.S. canola production to a record 2,017 million pounds. A larger harvest trimmed Canadian imports to 479 million pounds from 534 million the previous season. In fact, U.S. canola seed exports jumped from 299 million to 486 million pounds. Foreign sales are mostly to Canadian crushing plants, although shipments to Mexico grew rapidly in 2000/01. Larger supplies and less processing of sunflowerseed also helped domestic canola seed crushing expand to 1.7 billion pounds. Consequently, U.S. ending stocks of canola fell from 109 million pounds to 84 million. The slowdown in Canadian crushing also stalled the growth in U.S. imports of canola meal and canola oil. Canola producers received $70.8 million of marketing loan benefits for the 2000 crop. In addition, producers received Oilseed Program payments of $5.7 million in February 2001 and supplemental payments of $4.8 million in August 2001. Likewise, U.S. flax acreage surged 39 percent from 1999 to 536,000 acres. With a relatively normal yield of 20.8 bushels per acre, 2000 flaxseed output rose to 10.7 million bushels from 7.9 million in 1999. The comparatively large domestic crop sharply reduced imports from Canada from 6.6 million bushels to 2.8 million. Canada itself (the world's largest flaxseed producing country) slashed acreage in 2000 given its own huge carryover from the previous season. Consequently, U.S. flaxseed exports also expanded sharply in 2000/01 to more than 1 million bushels. Ending stocks dropped from 1.8 million bushels to 1.3 million. But the still-high Canadian flaxseed surpluses depressed U.S. farm prices again in 2000/01, which dropped from $3.79 per bushel to $3.30 and well below the $5.21 marketing loan rate. As a result, flaxseed producers received $21.8 million in marketing loan benefits for the 2000 crop. Producers also received Oilseed Program payments of $1.3 million in February 2001 and supplemental payments of $1.1 million in August 2001. U.S. farmers decreased safflower plantings in 2000 by 22 percent to 215,000 acres. Safflowerseed yields (at 1,434 pounds per acre) fell off from a very high 1999 level, so production declined 30 percent to 283 million pounds. The smaller harvest helped reduce an unusually large stock carryover from 1999/2000 (82 million pounds) to less burdensome 57 million. Exports of safflowerseed and safflowerseed oil both fell to their lowest levels since the mid-1980s. Although safflower producers received no direct payments under the marketing loan program because of the high price of safflowerseed relative to the 9.3-cents-per-pound loan rate, producers received Oilseed Program payments of $1.7 million in February 2001 and supplemental payments of $1.4 million in August 2001. Other Fats and Oils Highlights Corn Oil U.S. corn oil production fell in 2000/01 to 2,403 million pounds from 2,501 million a year earlier. Production incentives in 2000/01 were diminished by a sharp fall in the price of corn oil, which slumped from 17.8 cents to 13.5 cents per pound. Domestic corn oil output also declined as a consequence of a reduced grind by dry corn millers, whose sales suffered from concerns over the possible presence in their products of an unapproved, genetically modified corn variety. This same process separates the corn germ, from which corn oil is extracted, from the other corn components. Large world supplies of competing vegetable oils curtailed demand for U.S. corn oil exports from 970 million to 890 million pounds. Yet, a low price (an unusual discount against soybean oil for the first half of 2000/01) encouraged a high rate of domestic consumption for this byproduct. U.S. corn oil disappearance swelled to 1,635 million pounds from 1,470 million in 1999/2000. Imported Oils World coconut oil output, particularly in the Philippines, continued a comeback in 2000/01. Production rose from 3.3 million tons in 1999/2000 to 3.4 million. In fact, the supply situation has quickly turned from a deficit into a surplus. Coconut oil prices (Rotterdam) plunged to an average of $319 per ton in 2000/01, compared with $539 and $748 in the 2 previous years. The easing of prices swelled U.S. coconut oil imports to 1,145 million pounds from 926 million in 1999/2000. This allowed domestic disappearance of coconut oil to recover to 1,044 million pounds and a rebuilding of ending stocks to 230 million pounds. World production of olive oil rebounded 8.5 percent in 2000/01 to 2.56 million tons. The EU accounts for about three-fourths of world output, with Tunisia, Syria, and Turkey accounting for much of the remainder. More than 70 percent of world consumption of olive oil also occurs in the European Union. Yet, the United States is a major import market. Health conscious consumers strongly raised U.S. olive oil imports to 470 million pounds from 417 million in 1999/2000. The strength of the dollar versus the euro in 2001 also enhanced the purchasing power of U.S. consumers of olive oil. Animal Fats A decline in U.S. cattle slaughter slowed edible tallow production in 2000/01 to 1,820 million pounds from 1,810 million in 1999/2000. By the summer, tallow prices gradually firmed, averaging 13.4 cents per pound in 2000/01 compared with 13.2 cents the previous season. However, very low vegetable oil prices still pressured the value of tallow. Domestic tallow disappearance fell 5 percent to 1,520 million pounds as a decline in edible use offset gains in industrial use. The major factor buoying the edible tallow price was stronger foreign demand. The EU Commission exempted animal fats from its ban on animal protein in feeds, requiring only that they be completely filtered. But, EU beef production fell and because no meat meal could be sold, rendering of tallow in Europe also declined sharply. And despite a lack of evidence that beef tallow can transmit BSE, France and Germany adopted their own feeding restrictions on tallow. So, lower tallow consumption favored imports of cheaper substitutes such as palm stearin. As European exports faded, U.S. edible tallow exports helped make up the difference by swelling to 330 million pounds in 2000/01 compared with 224 million the previous year. Output of lard by U.S. renderers fell to 1,055 million pounds in 2000/01, from 1,069 million a year earlier. After June 2001, lard prices began recovering from the very low levels of the previous year, and the 2000/01 average price firmed to 14.6 cents per pound from 13.6 cents in 1999/2000. The relatively low prices encouraged a near-record domestic disappearance of lard, which rose to 964 million pounds from 886 million in 1999/2000. In contrast, U.S. lard exports fell by half in 2000/01 to 95 million pounds. Weak sales to Mexico (which imported more tallow, instead) and Hong Kong were largely responsible for the decline. End Uses of Fats and Oils Low prices in calendar year 2000 encouraged a surprisingly strong response for U.S. oils and fats consumption. U.S. output of salad and cooking oils rose a robust 19-percent to 9,155 million pounds. Domestic consumption surged 19 percent to 9,522 million pounds. On a per capita basis, U.S. consumption for these uses increased from 29.4 pounds in 1999 to an all-time high of 30.5 pounds. Exports of salad oils also rebounded to 734 million pounds. Production of baking and frying fats also increased substantially in 2000, rising 11 percent to 6,593 million pounds. U.S. per capita consumption of shortenings jumped to 23.7 pounds from 21.6 pounds in 1999. A 6-year slide in domestic margarine production ended in 2000, as output rose from 2,274 million to 2,398 million pounds. Total margarine consumption also recovered by 5 percent to 2,353 million pounds and per capita consumption edged up to 8.5 pounds. Conversely, consumption of oils for inedible uses dropped 5 percent in 2000 to 6,416 million pounds. Lower consumption of oils for soap and fatty acids outweighed negligible increases in animal feeds and paints. After a strong increase in 1999, the inedible consumption of soybean oil was unchanged in 2000 at 588 million pounds. Consumption of linseed oil declined 2 percent to 81 million pounds. Soybean Production Costs and Export Competitiveness in the United States, Brazil, and Argentina Erik Dohlman, Randall Schnepf, and Chris Bolling 1/----- ----- 1/ Agricultural economists, Market and Trade Economics Division, ERS. ----- Abstract: Argentina and Brazil have become increasingly strong competitors to the United States in international soybean and soybean product markets, as evidenced by steady market share gains in recent decades. A comparison of combined marketing, transportation, and farm-level production costs in the late 1990s reveals that Brazil and Argentina maintained a competitive advantage over the United States in production costs, mainly due to higher imputed land values in the United States. The U.S. production cost disadvantage was partially offset by lower internal transportation and marketing costs, but Brazil and Argentina have reduced these costs considerably in recent years. Key words: Brazil, Argentina, agriculture, soybeans, production costs, competitiveness. Introduction The competitiveness of U.S. agricultural products in export markets is an ongoing concern for domestic producers and U.S. policymakers. The United States has long been the world's leading exporter of soybeans, corn, and wheat, but it has faced increased competition from other exporters for global market share of these commodities. This situation is exemplified by the declining share of U.S. soybean and product exports in global markets since 1980, despite increased domestic production and aggregate exports. For example, the U.S. share of global soybean and soymeal exports (in soybean equivalents) has declined from about 55 percent in 1980 to slightly over one-third in 2000, whereas Brazil and Argentina's combined share of global soy complex exports has grown from about 31 percent to nearly 50 percent (fig. A-1). ----- 2/ The U.S. share of world corn exports fell from an average of 67 percent during 1980-89 to 61 percent during 1998-2000. The U.S. share of world wheat exports fell from an average of 34.3 percent during 1980-89 to 22.8 percent during 1998-2000. Source: USDA, PS&D database. ----- Competitiveness in commodity markets of course reflects the influence of many different factors. These include relative resource endowments and agro-climatic conditions, but also the impact of macroeconomic policies (affecting exchange rates, work incentives, investment, energy costs and availability, etc.), sector-specific policies (e.g., credit subsidies, import or export taxes on inputs or final products), infrastructure (for storage and transportation), and supporting institutions (e.g., credit, regulatory, news and information, etc.) that help markets to work effectively. Export shares and growth trends also depend on domestic demand, relative returns to other crops, and other conditions. However, in its simplest terms, international market competitiveness is the ability to deliver a product at the lowest cost--i.e., with the lowest combined farm-level production, transportation, and marketing costs. On this basis, analysis of 1998/99 cost structures underlying soybean production, transportation, and marketing from principal growing regions to a common export destination, Rotterdam, suggests that the United States lags slightly behind Argentina and Brazil in soybean export cost competitiveness. At the farm level, soybean producers in the U.S. 'Heartland' had the highest overall average costs of production at $5.11 per bushel, ranging from 18 to 25 percent above those of Argentine or Brazilian competitors. 3/ ----- ----- 3/ The Heartland is defined as western Ohio, Indiana, Illinois, Iowa, northern Missouri, western Kentucky, and parts of Nebraska, Minnesota, and South Dakota. ----- Total production costs were lowest in Argentina's central soybean growing region (southern Santa Fe and northern Buenos Aires Provinces) and in Brazil's interior expansion zone (the State of Mato Grosso), at about $3.90 per bushel in both regions (fig. A- 2). Production costs in Brazil's coastal State of Parana (in its traditional agricultural heartland) were estimated at $4.16 per bushel. High imputed land costs in the United States account for much of the difference in overall production costs. The U.S. production cost disadvantage is partially mitigated by internal transportation and marketing cost savings. In Brazil and Argentina, these costs are two to three times higher, on average, than in the United States, despite important efficiency gains in recent years. Freight charges to Rotterdam are also higher from South America. As a result, the delivered cost of Argentine and Brazilian soybeans at Rotterdam ranged from 2 to 12 percent less than U.S. costs in 1998/99. Methodology Behind the Cost Comparisons The export cost competitiveness of U.S., Brazilian, and Argentine soybean producers is examined by comparing the components and distribution of farm-level production costs, the costs of internal marketing and transportation, and shipping costs to a common export destination. Cost data for each country were from local 1998/99 marketing years, the most recent year for which detailed comparisons were possible. First, production costs were separated into their variable- and fixed-cost components. Variable costs include the use of inputs such as seed, fertilizer, chemicals, fuel, machine repair, interest on operating capital, and other direct costs incurred during crop production. Land costs--e.g., rental, maintenance, etc.--are not included with variable costs of production, but are combined with fixed production costs following Economic Research Service (ERS) methodology that uses land rental rates to value the opportunity cost of all land farmed. Fixed costs include costs that are not directly tied to the production decision, such as land payments on principal, interest and taxes, depreciation of machinery and equipment, and farm overhead. --Box Story: Why compare costs? In addition to providing an overview of current cost conditions in each country, cross-country comparisons of production and marketing costs can be a useful tool for decision-makers considering production, investment, or policy alternatives, and can help guide expectations of future market developments. For example, a country that can produce and transport a commodity to an export destination at lower cost would be expected to increase production and gain market share relative to its competitors, holding other factors equal. In addition, information on the contribution of particular cost components to total production and marketing costs can be used to interpret the impact of changing input prices on production incentives in different countries. A sustained rise in fuel prices, for instance, could have a greater negative impact on Brazilian soybean supply and export growth than in the U.S. or Argentina since the costs of transporting soybeans from production regions to ports are disproportionately large in Brazil, especially from the country's interior Center-West region. This is due to the greater reliance on road (truck) transportation to ports in Brazil than in the United States (where commodities are generally transported by barge), and greater average distances to port than in Argentina (average distance from farm-gate to the Argentine port of Rosario is about 330 kilometers, compared with about 1,500 kilometers from Brazil's Center-West to Atlantic ports). Similarly, natural gas prices may have a stronger impact on corn- soybean planting tradeoffs in the United States than Argentina since (natural-gas based) nitrogen fertilizers are more heavily used by U.S. corn producers. The contribution of internal transportation costs to final port prices can also inform policy- makers and private investors about the potential impacts of transportation infrastructure projects. Other investment decisions, such as the construction of new processing facilities, can be guided by information on the cost-competitiveness of production in different countries and regions within each country. --End Box Cost data from the U.S. Heartland region, where most U.S. soybean production takes place, were chosen to represent the United States. U.S. data are based on surveys by the National Agricultural Statistics Service (NASS), using the Agricultural Resource Management Study (ARMS). The data are compiled and published by ERS for regional and national aggregates. 4/----- ----- 4/ For soybean cost-of-production data, see http://www.ers.usda.gov/data/costsandreturns/car/soybean2.htm. ----- For Brazil, data from USDA and Brazilian Government sources were compiled for two regions: the State of Parana, a leading soybean producer in the South; and Mato Grosso, the largest soybean producing State in the Center-West. In Argentina, average variable cost-of-production data for northern Buenos Aires/southern Santa Fe (the heart of the corn- soybean region) were obtained from Margenes Agropecuarios (January 1999) based on no-till, Roundup Ready soybean production for high-yielding corn and soybean land. The lower end of the average yield range of 3.4 to 3.8 tons per hectare (50.6 to 56.5 bushels per acre) was used in the per-bushel cost calculations. Argentine land rents are also based on data from Margenes Agropecuarios (July 1999) for rental rates in the northern Buenos Aires production region. Other fixed cost data were adapted from Vieira and Williams (1996). A detailed and comparable breakdown of variable production costs for the Buenos Aires/Santa Fe region was not available, but the distribution of variable production costs based on suggested practices in the northern Province of Chaco was available, and is presented in table B-1 for comparison purposes. 5/----- ----- 5/ Chaco is primarily a cotton growing region, but soybean production has emerged there in the past decade. According to Hinrichsen (2001), 350,000 hectares of soybeans were planted in Chaco in 1999, making it the fifth leading soybean Province in Argentina, by area planted. ----- Internal marketing and transportation costs in the United States and Brazil are estimated by calculating the average monthly spread between farm-level soybean prices and the f.o.b. (free on board) port prices during calendar years 1998 and 1999. These spreads should reflect differences in transportation, storage, drying, loading and unloading, taxes, and other costs associated with bringing soybeans from farm to cargo vessel. Port prices are from the U.S. Gulf ports and the port of Rio Grande in Brazil. 6/------ ----- 6/ Although other major ports in Brazil (e.g., Santos and Paranagua) lie closer to the production regions in Parana and Mato Grosso, a consistent series of f.o.b. prices was available only for the port of Rio Grande. Nevertheless, f.o.b. prices for Rio Grande should be reflective of f.o.b. prices at other ports in Brazil's South since they all lie in relatively close proximity to oceangoing cargo vessels. ----- For Argentina, monthly farm-level prices were not available, so internal marketing and transportation costs were estimated in two steps. First, port and associated charges (including a 3-percent export tax) were estimated as the difference between f.o.b. port prices and f.a.s. (free alongside ship) Rosario terminal prices-- reflecting port charges (loading, export tax, and quality control). Next, costs of bringing soybeans from farm to port were estimated using information from other sources on internal transportation charges at the average distance to port in 1998, plus estimates of other marketing costs (loading/unloading, and brokers' commission). -----7/ ----- 7/ Estimates of freight and other charges from farm to port are based on data from the Brazilian oilseed crushing association (ABIOVE), cited in Verheijden and Reca (1998), and data provided by the Argentine beokerage firm Cortina-Beruatto (Frogone, 2001). ----- The third factor affecting the competitiveness of U.S. and South American soybeans in export markets is the cost of bringing the soybeans from the point of embarkation to their export destination. These costs are estimated by examining the average monthly spread between f.o.b. port prices and the c.i.f. (cost, insurance, and freight) price at a destination port, in this case Rotterdam during 1995-99. The European Union is the world's largest importer of soybeans and soymeal--accounting for about 35 percent of global soybean imports and about 40 percent of soymeal imports during the 1998 and 1999 marketing years--and Rotterdam is the leading port of entry for these products. Table B-1 summarizes the production cost data on a per-acre and per-bushel basis, and table B-2 presents estimates of the overall 'export cost' from the different production regions using a 'landed' soybean price in Rotterdam--calculated by adding the estimated shipping charges and internal marketing and transportation costs to the farm-level costs of production for each country. The comparisons made here are only rough indicators of competitiveness. Comparisons of farm-level costs of production, in particular, are difficult and potentially imprecise for a number of reasons. For example, the methods used to calculate costs vary considerably from country to country, with certain components of cost included by one country and omitted by others. In addition, cost estimates may be based on different production practices (such as single- or double-cropping, till or no-till production) or slightly different time periods (based on local growing seasons). Estimates are further complicated by exchange rate conversion issues, differences in financial versus economic accounting, the impact of policy distortions, and the fact that data reflect production and marketing costs for regions that bear different relationships to national averages in their respective countries. Data presented here may not correspond exactly with source data due to certain assumptions and the omission or reformulation of some data to make them as comparable as possible. Soybean Production Cost Structure Favors Argentina and Brazil With their favorable natural resource endowments and climates, Argentina and Brazil are naturally low-cost producers of soybeans, giving them a strong competitive edge in international markets. Based on 1998 farm-level soybean production cost and yield data, total per-bushel costs in Brazil's Mato Grosso ($3.89 per bushel) and Argentina ($3.92 per bushel) were 23-24 percent lower than the U.S. Heartland's $5.11 total cost per bushel. Production costs in Parana ($4.16 per bushel) were 19 percent lower. Similarly, total per-acre soybean production costs were highest in the U.S. Heartland, averaging about $235, some $60-$70 more than in Brazil and about $35 an acre higher than in Argentina during 1998/99 (table B-1).8/----- ----- 8/ Total per-acre soybean production costs in the Heartland are slightly above the U.S. national average, largely reflecting higher land costs, but higher yields led to somewhat lower (about $0.25/bushel) per-bushel costs of production than the national average. We exclude the opportunity cost of unpaid labor from the U.S. data. It is likely also excluded from Argentine and Brazilian data. ----- The relatively high overall costs in the United States are attributable largely to high fixed costs of production, particularly the large imputed land costs faced by U.S. producers. This is especially true in comparison with Brazil, where estimated rental rates are just $6 (in Mato Grosso) to $14 (Parana) per acre, compared with $88 in the U.S. Heartland and $63 for prime land in northern Buenos Aires Province. The particularly low rental rates in Brazil's Center-West reflect the abundance of cerrado soils still available for conversion into agricultural production. Recent reports indicate that high yielding land in Mato Grosso can still be purchased for as little as $200 an acre, compared with over $2,000 per acre in the U.S. Corn Belt. Differences in land costs clearly play a crucial role in assessments of competitiveness based on overall production costs. For example, if land costs are excluded from overall production costs, the United States would rank ahead of Brazil, but still behind Argentina, in production-cost competitiveness. 9/----- ----- 9/ Previous studies (Ortmann et al., 1989; Vieira and Williams, 1996) show similar results. ----- Based on variable costs alone, soybean growers in the U.S. Heartland are the low-cost producers. In Parana, greater fertilizer and labor costs (due to small-scale and labor- intensive production practices) inflate variable costs. In Mato Grosso, higher fertilizer and chemical costs (due most likely to higher prices rather than greater intensity of application) keep variable costs high. Low expenditures on lime or fertilizers keep Argentine variable costs closer to U.S. costs. A previous ERS study (Trapido and Krajewski, 1989) also showed that the main Argentine producing Provinces (Buenos Aires and Santa Fe) had slightly higher variable costs per ton of production than the U.S. Corn Belt/Lake States, but another study (Ortmann et al., 1989) calculated per- ton variable costs to be slightly lower in Argentina. Also favoring soybean farms in Argentina and Brazil's Mato Grosso are their much larger size (averaging over 1,000 hectares) relative to soybean farms in the U.S. Heartland (120-150 hectares), or Brazil's Parana (about 30 hectares)--where land is scarcer and a large class of landless or near-landless labor exists. Large farm size spreads overhead costs over more acres, resulting in much lower per-unit costs. As a result, average depreciation of machinery and equipment costs were significantly lower in Mato Grosso and Argentina than in the United States. The United States had higher soybean production costs than Parana throughout the 1990s. U.S. average soybean costs rose steadily from $185 per acre in 1989 to $235 per acre in 1998, slightly below the general pace of consumer inflation. 10/----- ----- 10/ U.S. data prior to 1997 are for the North Central region, and for the newly defined Heartland in 1997 and 1998. ----- The increase was due mainly to rising fixed costs, particularly land. Increased chemical costs were responsible for a slight growth in variable costs. However, fluctuations in the Brazilian currency render U.S. dollar-valued representations somewhat misleading. For example, in dollar terms, costs of production in Parana have fluctuated considerably in the last 10 years. After declining sharply from $256 per acre in 1989 to $134 in 1991, total costs of production rose again to $169 per acre in 1992. Production costs ranged between $158 per acre and $205 per acre during 1993-98, before falling to a decade low of $129 per acre in 1999 (according to just recently available data). In local currency terms, however, total production costs in Parana rose nearly 30 percent between 1995 and 1999, so the apparent decline is largely a reflection of the weakening Brazilian currency, particularly after the real was allowed to float freely in international exchange markets. In Mato Grosso, most of the increase in total production costs between 1991 and 1998 (from $99 to $162 per acre) was due to higher chemical costs and interest on operating capital. Limited data from Argentina suggest that soybean producers there have had lower farm costs than U.S. producers throughout the 1990s. Internal Marketing and Transportation Costs are Lowest for United States The Brazilian and Argentine advantage in farm-level production costs is largely offset by much higher internal marketing and transportation costs. However, significant reductions in these costs since 1992 in Argentina and after 1996 in Brazil have boosted their soybean export competitiveness in recent years. During 1998-99, internal marketing and transportation costs for soybeans destined for export averaged two to three times higher in Brazil and Argentina than in the United States, tending to dampen farmgate prices. Based on average farm-to-port distances, these costs averaged $49 per metric ton ($1.33/bushel) from Mato Grosso, $31 per ton from Parana, and $30 per ton for Argentine producers. In the United States, these costs amounted to just $16 per ton. For producers in Mato Grosso, transportation and marketing costs were equivalent to one-quarter of the average f.o.b. port price during 1998. These figures correspond with the combined freight-to-port and port charges estimated by ABIOVE (Brazilian vegetable oil industry association) for each country. According to ABIOVE, at the average distance to port, these charges totaled $18 per ton for the United States and $25 per ton in Argentina (including export taxes but not a broker's commission of $2-$5 per ton) in 1998. For Brazil, these charges were estimated at $41 per ton. Since the mid-1980s, the average U.S. producer-to-f.o.b. port price spread has remained relatively constant at $16-$18 per ton. In Argentina and Brazil, however, privatization and deregulation of railways and ports, and the elimination or reduction of export controls have lowered transportation and marketing costs in recent years. In Argentina, the margin between the terminal cash price at Rosario and the f.o.b. price of soybeans at Argentine ports has narrowed from an average of $68 per metric ton during 1980-91, to just $11 per ton since 1991. Nevertheless, farmgate-to-terminal transportation costs remain high due to a heavy reliance on trucking for bulk transport, high toll rates on private highways, and seasonal transportation bottlenecks. In Brazil, similar internal cost reductions may have resulted in part from transportation infrastructure improvements, but also reflect the elimination (through rebates) of a 13-percent value- added tax on soybean exports in 1996. For Mato Grosso producers, whose soybeans must traverse roughly 1,500 kilometers to reach an east coast seaport, the producer-f.o.b. price spread averaged $76 per ton from 1983 to 1997. Since 1997, they have averaged an estimated $47 per ton. In Parana, where soybeans have a much shorter distance to oceangoing vessels, substantial internal cost reductions have also occurred as the producer-f.o.b. price spread has fallen from an average of $52 per ton during 1983-97 to $29 since 1997 (fig. A-3). Lower transport and marketing costs for the United States reflect, in part, the efficient barge transportation system that can transport grains long distances at low cost. In Argentina and Parana, the fact that most soybean production takes place within 250-300 kilometers of ports has kept their costs significantly below those of Mato Grosso. Shipping Charges to Rotterdam Favor United States The United States has a small advantage ($0.11 per bushel) over Argentina and a somewhat larger one over Brazil ($0.19 per bushel) in shipping charges to Rotterdam. This further narrows the export cost differentials when the combined production, marketing, and transportation costs are compared at the import destination of Rotterdam (table B-2). The difference between the f.o.b. export price and c.i.f. import price spreads for the United States and South American countries is mostly attributable to distance (to Rotterdam), but may also reflect higher insurance rates and demurrage costs for ships originating from South American ports. With even greater relative distances to East Asian ports (e.g., Japan, South Korea, and China), Brazilian and Argentine soybean exports face a larger disadvantage (compared with the United States) in shipping rates to these destinations. The gap between shipping rates from the United States and Brazil to Rotterdam has remained relatively constant over the last 15 years. But for Argentina, the average f.o.b.-to-c.i.f. price spread has narrowed from $26 per ton during 1984-94 to $18 per ton during 1995-99. Producer Revenues With substantially higher total costs of production and similar yields, per-bushel and per-acre net revenues based strictly on a market price (ignoring LDPs, production flexibility contract payments, emergency supplementary income payments, and subsidized crop insurance) for U.S. Heartland soybean producers fall short of those for producers in Brazil and Argentina, assuming similar producer prices. However, higher internal transportation and marketing costs have depressed Brazilian producer prices to levels below those in the United States. In October 1998, producer prices of $4.81/bushel in Parana and $4.58/bushel in Mato Grosso lagged the $5.16/bushel received (excluding LDPs) in the U.S. Heartland. In Argentina, average producer prices were estimated at $4.98/bushel in October 1998. 11/----- ----- 11/ Argentine producer prices were based on the difference between actual October 1998 f.o.b. prices ($213/ton) and the estimated costs of internal marketing and transportation ($30/ton). ----- Nevertheless, in 1998, estimated per-bushel and per-acre net producer returns in Argentina were the highest among the three countries, followed by Brazil and the United States. Argentine producers received an estimated $1.06/bushel in 1998, compared with $0.69/bushel in Mato Grosso, $0.65/bushel in Parana, and just $0.05/bushel in the U.S. Heartland. 12/----- ----- 12/ The net revenue figure of 5 cents per bushel for U.S. Heartland producers is based on market prices only and does not include potential extra revenue from marketing loan benefits. When prices are below the loan rate, U.S. producers can realize gross revenues above the loan rate of $5.26 per bushel by receiving benefits under the marketing loan program early in the market year when prices are typically lowest, and then by selling their crop later in the marketing year when prices have risen. In the 1998 marketing year, for example, the weighted average marketing loan benefit (marketing loan gains and loan deficiency payments) for the soybean crop was $0.44 per bushel. This benefit augmented the season-average price of $4.93 per bushel, raising the average per-unit gross revenue for soybeans to $5.37 per bushel, $0.11 above the national soybean loan rate. ----- Despite relatively low market-based returns in 1998 and consistently higher costs of production in the United States than in Brazil, estimated per-acre net revenues from soybean production in the United States have actually exceeded those of producers in Parana over much of the past decade (fig. A-4). Between 1989 and 1996, per-acre net returns in Parana exceeded those of U.S. North Central/Heartland soybean producers only once, in 1991. From 1997 to 1999, however, net revenues in Parana surpassed those in the United States, and were especially strong in 1998. 13/----- ----- 13/ The trend comparisons made here are based on local harvest- period prices, rather than adjusting prices to the same month (October 1998) as done elsewhere in this analysis. In the U.S., average producer prices are from October; average March-May producer prices were used for Brazil. For the U.S., data prior to 1997 are for the North Central region, and for the newly defined Heartland in 1997, 1998, and recently available 1999 data. ----- Reduced internal transportation and marketing costs, as well as declining production costs (in dollar terms), have seemingly improved the bottom line for Brazilian producers since 1996. From limited data, it appears that net revenues in Mato Grosso have equaled or exceeded those in Parana during the 1990s, which is consistent with the trend toward increased production (and economies of scale) in that region. Conclusion: Argentina Appears Most Competitive Both Argentine and Brazilian soybeans have become more competitive in recent years due to declining internal marketing and transportation costs, including the reduction/elimination of export taxes on soybeans. Brazilian soybeans have also benefited from substantial currency depreciation since 1999. In 1998/99, the underlying cost structures for producing, transporting, and marketing soybeans from Argentina's southern Santa Fe/northern Buenos Aires region and Brazil's two principal growing areas allowed them to bring soybeans to Rotterdam at prices slightly below U.S. soybeans grown in the Corn Belt. These cost advantages help explain the rapid expansion of soybean production and soybean/product exports by Argentina and Brazil during the last decade. In the future, increased soybean plantings by Argentina, holding other factors constant, may be restrained by limitations on the ability to expand total area devoted to agricultural production. In contrast, increased soybean production in Brazil's Center-West (e.g., Mato Grosso) appears especially promising, given abundant, inexpensive land available for cultivation. References Foreign Agricultural Service, USDA. 'Argentina Oilseed and Products Annual Report,' various years. http://www.fas.usda.gov/scriptsw/attacherep/default.asp Foreign Agricultural Service, USDA. 'Brazil Oilseed and Products Annual Report,' various years. http://www.fas.usda.gov/scriptsw/attacherep/default.asp Frogone, Jose, Cortina-Beruatto SA (Argentina). E. Dohlman, personal correspondence, March, 2001. Fundacao Getulio Vargas. Precos Recibidos, Soya, Agroanalysis, selected issues; http://www.fgv.br/. Goldsby, Thomas J. 'A Comparative Analysis of Agricultural Transportation and Logistics Systems in the United States and Argentina,' Argentina Report 3, MATRIC Research Paper 00-MRP 3, Midwest Agribusiness Trade Research and Information Center, Iowa State University, Ames IA, August 2000, www.card.iastate.edu/matric/home.html. Hinrichsen, J.J., S.A. 'J.J. Corredor -- Broker.' Annual yearbook on oilseed markets, Buenos Aires, Argentina, 1998-2001. Larson, Donald, and Norman Rask. 'Industry Note: Changing Competitiveness in World Soybean Markets.' Agribusiness, Vol. 8, No. 1, 1992. pp. 79-91. Lence, Sergio H. 'A Comparative Marketing Analysis of Major Agricultural Products in the United States and Argentina,' Argentina Report 1, MATRIC Research Paper 00-MRP 2, Midwest Agribusiness Trade Research and Information Center, Iowa State University, Ames IA, August 2000, www.card.iastate.edu/matric/home.html. Margenes Agropecuarios, various issues. Oil World Weekly, various issues. Ortmann, Gerald, Norman Rask, and Valter Stulp. 'Comparative Costs in Corn, Wheat, and Soybeans Among Major Exporting Countries.' Research Bulletin 1183, Ohio Agricultural Research and Development Center. Ohio State University, August, 1989. SAGPyA, Argentine Ministry of Agriculture, agricultural database, http://www.sagpya.mecon.ar/agricu/agricultura.htm. Trapido, P.J. and R. Krajewski. 'Soybean Costs of Production in Argentina, Brazil, and the United States: The Implications for Trade Liberalization,' ERS, Staff Report No. AGES-89, August, 1989. Trapido, P.J. and R. Krajewski. 'Soybean Costs of Production: The Cases of Argentina, Brazil, and the United States,' Developing Economics: Agriculture and Trade Report, RS-89-4, ERS, USDA, August, 1989, pp. 37-40. USDA, Economic Research Service. PS&D View, January, 2001; http://www.ers.usda.gov/data/psd/. USDA, Economic Research Service. Commodity Costs and Returns: Soybeans, January 2001 http://www.ers.usda.gov/data/costsandreturns/car/soybean2.htm. Verheijden, Sylvia and Alejandro Reca. The Crushing Industry in Brazil and Argentina, Special Report, Food & Agribusiness Research, Rabobank International, September 1998; Rabobank, International, 1998. Vieira, Rita T., and Gary W. Williams. 'The Brazilian Soybean Industry,' TAMCR International Market Research Report No. IM 4- 96, Texas A&M University, College Stations, TX, August 1996. Estimating Farm-Level Effects of Adopting Herbicide-Tolerant Soybeans William Lin, Gregory K. Price, and Jorge Fernandez-Cornejo 1/---- ----- 1/ William Lin and Gregory K. Price are agricultural economists of the Market and Trade Economics Division, and Jorge Fernandez- Cornejo is an agricultural economist with the Resource Economics Division, Economic Research Service (ERS). ----- Abstract: The farm-level effects of adopting herbicide-tolerant soybeans obtained from various data sources are compared and evaluated. In 1997, adopters' yields were estimated to be only 3 percent higher than for nonadopters. In the Heartland region, where about two-thirds of U.S. soybeans are grown, adopters' weed control costs were estimated to be 11 percent lower than for nonadopters, a savings of $3.50 per acre. Weed control cost savings were estimated to range from 1 percent to 34 percent in other regions. While the impact on herbicide use (measured in pounds of active ingredients) is mixed among production regions-- a decline for the Heartland and Prairie Gateway but an increase for all other regions--overall there is a slight increase. Still, adoption of herbicide-tolerant soybeans may lead to positive environmental and health benefits. According to the elasticity- based estimates, the benefits to U.S. farmers in 1997 were estimated to have been about $60 million, or about 20 percent of the total benefits to all stakeholders from the adoption of the technology. Keywords: Herbicide-tolerant soybeans, farm-level effects, crop yields, weed control costs, herbicide use. Introduction The adoption of herbicide-tolerant soybeans has been particularly rapid in the United States, increasing from less than 10 percent of soybean acreage in 1996, when the technology was introduced, to nearly 70 percent in 2001 (USDA, 2001; fig. 1). The rapid adoption has reflected the benefits of potential increases in crop yields and savings in pest control costs from this technology. But more importantly, herbicide-tolerant soybeans offer producers the simplicity and flexibility of a weed control program that relies on one herbicide to control a broad spectrum of weeds without crop injury or crop rotation restrictions (Carpenter and Gianessi, 1999[b]). Thus, estimates of the benefits from adopting herbicide-tolerant soybeans and their distribution among the stakeholders require accurate information about the technology's farm-level impacts on crop yields, pest control costs, and the ease of weed control management. Estimates of the farm-level effects differ significantly, depending on the data source. For example, a recent study of the distribution of benefits from biotech adoption by Falck-Zepeda, Traxler, and Nelson assumed that adopters' yields for 1997 herbicide-tolerant soybeans were 13.0 percent higher than nonadopters in the Corn Belt based on data from the U.S. Department of Agriculture's Agricultural Resource Management Study (ARMS) survey. In contrast, Moschini et al. studied the welfare effects of herbicide-tolerant soybean adoption by assuming no yield difference, adapted from a costs-of-production budget for Iowa (Duffy and Vontalge). Differentials in mean crop yields between adopters and nonadopters from the ARMS survey reflect the combined effect of biotechnology and other confounding factors, such as production practices. The effects on pest control costs differ even more dramatically across data sources than the impacts on crop yields. Variations among differing estimates warrant further evaluation of these farm-level effects and a concerted effort to find ways to reconcile the differences. Accordingly, the purpose of this article is to compare and evaluate the farm-level effects of adopting herbicide-tolerant soybeans obtained from various data sources. The analysis focuses on 1997 herbicide-tolerant soybeans in the United States for which the latest detailed ARMS survey data are available. The key data sources that we compared and evaluated include: (1) the means of the ARMS survey, and (2) the elasticity-based estimates obtained by isolating the effect of biotechnology through econometric analysis of the ARMS survey data (Fernandez-Cornejo et al.). Yield and herbicide use elasticities from their adoption-impact model are further analyzed in this study to show yield and per-acre weed control cost differentials between adopters and nonadopters. In addition, costs-of-production budgets or agronomic research from selected States, where applicable data are available, are compared with the ARMS data analysis and elasticity-based estimates. Results are used to compare alternative estimates of producer benefits from the adoption of herbicide-tolerant soybeans and implications for herbicide use. Data Sources The estimates of potential increases in crop yields and savings in weed control costs induced by the adoption of herbicide- tolerant soybeans are among the most difficult variables in measuring the benefits from adopting this technology. This section discusses two data sources that were used to estimate the farm-level effects of adopting herbicide-tolerant soybeans. The ARMS survey, a Nationwide producer survey conducted by USDA to monitor economic and environmental indicators in the U.S. farm sector, is a data source used by some researchers to estimate the farm-level impacts (Falck-Zepeda, Traxler, and Nelson). Farm financial and chemical use data are reported for all crops in the ARMS survey each year, while detailed enterprise production practices and cost data are collected for several commodities (including soybeans) on a rotating basis every 4 to 7 years (McBride).2/----- ----- 2/ The 1997 ARMS data are used in this study because these are the latest available for soybeans. The 1997 data are based on a relatively small number of adopters. Only about 12 million acres (17 percent) of herbicide-tolerant soybeans were planted that year. Availability of herbicide-tolerant soybean seed varieties suited for northern States (Minnesota, Wisconsin) was quite limited in 1997. ----- There are four characteristics of the ARMS data that make it particularly useful for assessing the farm-level impacts of biotechnology adoption (McBride). First, the ARMS survey has a broad coverage, including all major States producing a particular commodity, and generally covers more than 90 percent of the acreage of targeted commodities. Second, the ARMS survey uses a stratified random sample where each farm represents a known number of similar farms in the population based on its probability of being selected. Each farm is weighted by the number of farms it represents so that the ARMS sample can be expanded to reflect the targeted population. Third, ARMS enterprise costs-of-production data contain sufficient detail about specific inputs to isolate the seed and pest control costs used to produce a given commodity. Finally, enterprise costs of production can be estimated for each observation in the ARMS data so that a distribution of costs can be developed. However, data from farm surveys like the ARMS are more expensive to obtain and are more difficult to use than State cost budgets. More importantly, the data do not easily lend themselves to estimating the farm-level impacts solely attributed to the adoption of biotechnology. Mean crop yield and pest control cost differentials between adopters and nonadopters often reflect not only the effect of biotechnology adoption but also other factors, such as production practices, soil productivity, farm size, and the managerial ability of farm operators. An alternative to estimating mean differentials from the survey data is to estimate the impacts of biotechnology adoption by statistically isolating the effects of the technology through econometric analysis (Fernandez-Cornejo, Klotz-Ingram, Jans; Fernandez-Cornejo and McBride). The econometric model is also estimated from ARMS survey data but takes into account the fact that farmers' adoption of biotechnology and pesticide-use decisions may be simultaneous. In addition, the model corrects for self-selectivity to prevent biasing the results. Self- selection arises because farmers are not randomly assigned to one of two groups--adopters and nonadopters; instead, they make the adoption choices themselves. Therefore, adopters and nonadopters may be systematically different from each other, and these differences may manifest themselves in farm performance, which could confound the effect of adoption. The results of this two-stage impact model are expressed in elasticity form. In terms of the impact on crop yields (an elasticity of +0.03), the adoption of herbicide-tolerant soybeans has a positive and statistically significant effect, but it is small--U.S. herbicide-tolerant soybean yields in 1997 are estimated to have increased by 0.3 percent for a 10-percent increase in adoption across the Nation. This yield effect is generally consistent with other studies (Gianessi and Carpenter; Carpenter and Gianessi, 2000; Carpenter and Gianessi, 1999(b); Moschini, Lapan, and Sobolevsky; Duffy and Vontalge). The impact of adopting glyphosate-tolerant soybeans on herbicide use, based on this econometric model, varies across active ingredients. Glyphosate is the active ingredient in Roundup, the broad-spectrum herbicide that Monsanto developed its Roundup- Ready soybeans to resist. Formerly, only pre-emergence applications of glyphosate on soybeans were possible without crop injury. An increase in the adoption of glyphosate-tolerant soybeans is estimated to have led to statistically significant reductions in the use of herbicides other than acetamides (such as metolachlor and alachlor) and glyphosate (an elasticity of - 0.14) and a significant increase in the use of glyphosate (an elasticity of +0.43). Thus, use of other synthetic herbicides is estimated to have decreased by 1.4 percent for a 10-percent increase in adoption of herbicide-tolerant soybeans. In contrast, use of glyphosate is estimated to have increased by 4.3 percent. The change in acetamides use was not statistically significant. Estimated Impacts on Crop Yields A fundamental question that needs to be addressed before estimating the benefits from adopting herbicide-tolerant soybeans is: 'How much of the difference in crop yields between adopters and nonadopters is attributable to the adoption of the technology?' This section discusses the impacts of adopting herbicide-tolerant soybeans on crop yields mainly across two data sources: (1) the mean values for adopters and nonadopters obtained from the ARMS survey, and (2) the analysis based on the elasticities derived from the adoption-impact model. Other data sources, such as State variety trials or agronomic research, are also discussed. The impact of adopting herbicide-tolerant soybeans on crop yields, based on one year of data (1997), appears to vary significantly across data sources and production regions (fig. 2). There are significant differences in the impacts of adopting herbicide-tolerant soybeans between estimates from the mean ARMS data and the elasticity-based estimates. In the Heartland region, while adopters' 1997 yields are shown to have been 14.2 percent higher than those of nonadopters (averaging 44.4 bushels per acre), based on the mean ARMS data, the elasticity-based estimate indicates only a 3-percent higher yield for adopters (fig. 3). Heartland producers account for 64 percent of U.S. soybean acreage and 56 percent of the herbicide-tolerant soybean area. The small, 3-percent increase in yields based on the elasticity estimate statistically removes factors other than biotechnology, such as production practices, farm operator's managerial ability, soil productivity, and weather, which affect crop yields. Thus, the elasticity-based estimate reflects the impact on crop yields that is attributable to the technology, and is consistent with findings of other studies. That is, the adoption of herbicide- tolerant soybeans has little or no overall impact on U.S. soybean yields (Gianessi and Carpenter; Duffy and Vontalge; Moschini, Lapan, and Sobolevsky). Similar patterns also exist for the Southern Seaboard, Prairie Gateway, and Northern Great Plains regions (table 1). Thus, studies using mean yields from the ARMS survey (e.g., Falck-Zepeda, Traxler, and Nelson) would overestimate the benefits from biotech adoption for U.S. soybean farmers. Despite conflicting evidence, yield trials and costs-of- production budgets from selected States suggest that, overall, there is little difference in yields between herbicide-tolerant soybeans and conventional varieties. In 1998, yields of herbicide-tolerant soybeans were reported to have been 4 percent lower than conventional varieties (about 1 bu./ac.) based on variety trials from more than 3,000 side-by-side comparisons across 40 university performance tests in eight States (Oplinger). However, comparisons in yield trials are made under weed-free conditions and do not necessarily represent farm conditions where imperfect weed control leads to some yield losses, particularly for the case of conventional varieties. Many analysts believe these yield drags will disappear as more backcrosses are made to capture the yield potential in the parent lines. In addition, agronomic research in Minnesota concluded that there was no difference in yields between herbicide-tolerant soybeans and conventional varieties (Breitenbach and Hoverstad). Estimated Impacts on Weed Control Costs Another fundamental question that needs to be addressed before estimating the benefits from adopting herbicide-tolerant soybeans is: 'How much of the difference in weed control costs between adopters and nonadopters is attributable to the adoption of the technology?' This section discusses the impacts of glyphosate- tolerant soybeans on weed control costs (including expenses associated with herbicides, herbicide application, scouting, and cultivation) across the data sources. Weed control costs for glyphosate-tolerant soybean adopters were lower in 1997 than those incurred by nonadopters (table 2). 3/--- ----- 3/ This result is consistent with the finding by Marra et al. that in 1996, total herbicide costs decreased despite an increase in glyphosate use--glyphosate costs rose by $13/ac. while expenditures for other herbicides fell by $24/ac. In 1999, 0.98 pound per acre of herbicides were applied on soybeans, down from 1.01 pounds per acre in 1995 (Carpenter and Gianessi, 2000). ----- However, adopters' savings in weed control costs based on means of the ARMS survey were generally much higher than those based on the elasticity-based estimates. For example, while soybean weed control costs in the Heartland were 31 percent lower for adopters than an average of $33.05 per acre for nonadopters based on the mean ARMS data, the saving is estimated with the elasticity-based approach to be 11 percent. Similar patterns exist between the two data sources for other regions. The elasticity-based estimates of the difference in weed control costs between adopters and nonadopters are based on the elasticities determined by Fernandez-Cornejo, Klotz-Ingram, and Jans (Box 1). Herbicide material costs for adopters are estimated by acounting for any potential savings in herbicide use, ingredient-by-ingredient, assuming application rates of each active ingredient are the same for adopters and nonadopters. Savings in herbicide material costs are based on herbicide use elasticities estimated by Fernandez-Cornejo, Klot-Ingram, and Jans--+0.43 for glyphosate and -0.14 for 'other synthetic herbicides.' The -0.14 elasticity for other synthetic herbicides means that adopters' herbicide use would have been about 14 percent lower than nonadopters' herbicide use. Thus, expenses in 1997 associated with the other herbicides were $24.30 per acre for nonadopters compared with $20.90 per acre for adopters. No significant change occurred in the use of acetamides. In contrast, per-acre glyphosate material costs were about $1.16 per acre higher for adopters in the Heartland, increasing from $2.70 for nonadopters to $3.86 for adopters. However, the decline in expenses associated with the use of other herbicides--$3.40 per acre--more than offset the increase in expenses for glyphosate. As a result, adopters' expenses for all herbicide materials totaled $25.64 per acre, lower than the $27.89 per acre for nonadopters. 4/----- ----- 4/ With the expiration of Monsanto's patent on glyphostae in 2000, the cost of the chemical has dropped as generic products are now available. This will likely expand the current cost savings from the 1997 estimates. The competition from Roundup probably forced other herbicide producers to cut their prices to maintain market share. Glyphosate price declined from $56.70/gallon in 1997 to $43.30/gallon in 2000. ----- ---Box text Procedures to calculate the difference in per-acre weed control costs between adopters and nonadopters of herbicide-tolerant soybeans The first step to calculate the impact of adopting herbicide- tolerant soybeans on weed control costs is to estimate expenses associated with herbicide material for nonadopters in a specific production region, such as the Heartland region. Based on NASS' chemical use data, the herbicide application rate per crop year and the percent of area applied with herbicide are tabulated by herbicide ingredient at the regional level, based on State data (table 3). Herbicide-active ingredients are grouped into three categories: acetamides, glyphosate, and other synthetic herbicides. Price data for herbicide-active ingredients are obtained from NASS (1998b) and from a database of 1996 prices developed by Gianessi and Marcelli. In cases where herbicide prices were expressed in terms of dollars per pound of active ingredient, no adjustment of the price data was necessary. However, in cases where price data were shown for final products in dollars per gallon, the final product price was divided by the product-active ingredient conversion ratio (lb/gal) to obtain prices for active ingredients ($/lb). Multiplying the active ingredient price by a weighting factor, which is the product of the application rate per crop year and the percent of area applied, gives the expense associated with a specific active ingredient for nonadopters. Continuing this calculation for all active ingredients and adding up the expenses across active ingredients result in the total per-acre expense associated with herbicide materials --$27.89 for nonadopters of herbicide-tolerant soybeans in the Heartland. Including expenses for herbicide application, scouting, and cultivation (taken from the ARMS data) bring the total weed control cost to $32.78 per acre for nonadopters. Per-acre weed control costs for adopters ($29.28) are estimated by taking into account the elasticities, ingredient-by-ingredient. ---End box text Adopters of herbicide-tolerant soybeans in the Heartland region are estimated to have realized a 10.68-percent weed control cost saving (or $3.50 per acre) if herbicide applications, weed scouting, and cultivation expenses, together with herbicide materials, are all included in the calculation of weed control costs (fig. 4). According to the elasticity-based estimate, weed control cost savings are estimated to range from 1 percent to 34 percent, depending on the production region. The above $3.50-per-acre savings in weed control costs in the Heartland are reaffirmed by other studies. A recent University of Nebraska study shows that weed control costs for herbicide- tolerant soybeans ($10.45/ac, excluding a seed technology fee of $6/ac.) were about $3/ac. lower than a conventional herbicide program using Pursuit Plus as a herbicide in 1998 (Rawlinson). The costs-of-production budget for Iowa assumes that in 2001, herbicide program costs for herbicide-tolerant soybeans ($25.2/ac.) were about $4.8/ac. lower than for conventional varieties (Duffy and Vontalge). The smaller savings in weed control costs for herbicide-tolerant soybean adopters from the elasticity-based approach contradicts the much larger savings assumed in the Moschini, Lapan, and Sobolvsky study, where the savings in herbicide expenses ranged from 48.9 percent to 66 percent, depending on the number (either one or two) of over-the- top Roundup treatments. Implications for Producer Benefits and Herbicide Use The above analysis has important implications for the benefits to U.S. soybean farmers and the use of herbicides for controlling weeds in soybean production. This section estimates the benefits to U.S. farmers resulting from the adoption of herbicide-tolerant soybeans and the implications of their adoption for herbicide use in soybean production. Producer Benefits According to the elasticity-based estimates from this study, the benefits to U.S. farmers in 1997 are estimated from an on-going ERS study to have been about $60 million, or about 20 percent of the total benefits to all stakeholders (including U.S. farmers and consumers, Monsanto, seed companies, and rest-of-the-world producers and consumers) from the adoption of herbicide-tolerant soybeans. This analysis is based on an economic analysis that takes into account the farm-level impacts in terms of the increase in soybean yields and weed control cost savings, price effects resulting from the adoption of the technology, as well as seed premiums and the technology fee (about $7 per acre) for herbicide-tolerant soybean seeds. U.S. farmers who adopted herbicide-tolerant soybeans on 12 million acres of cropland in 1997 gained all these benefits. The $60-million benefit to U.S. farmers in 1997 based on this study are considerably lower than the $220-million benefit accrued to U.S. farmers who adopted Roundup Ready soybeans in 1998 as estimated by Gianessi and Carpenter. However, these two studies differ in their approaches. The former study compares net benefits between adopters and nonadopters of herbicide-tolerant soybeans in 1997, region-by-region. The latter study attributes the bulk of the benefits to savings in herbicide costs between 1995 and 1998 by taking into account the decline in herbicide prices during this period and the seed technology fee ($6 per acre) paid by the adopters. Similarly, the $60-million benefits are much smaller than the $808-million benefits to U.S. farmers (based on the scenario in which supply elasticity of 0.22 is assumed) in 1997 as estimated by Falck-Zepeda et al. The use of the 1997 mean ARMS data in the latter study, which show much larger yield and herbicide cost impacts than the elasticity-based estimates, is a primary factor that contributes to the difference. The benefits obtained from this study also are smaller than the $156-million benefits to U.S. farmers in 1999 estimated by Moschini et al., which were also found to account for about 20 percent of the total benefits to all stakeholders. Herbicide Use The above analysis of the effect on weed control costs also has important implications for herbicide use and weed control. Biotechnology promises to simplify pest management, reduce the use of chemical inputs, and increase flexibility in field operations. If the adoption of glyphosate-tolerant soybeans leads to savings in weed control costs as well as herbicide use, then the technology will not only benefit producers but also have positive environmental and health benefits. On a per-acre basis, the use of herbicides in soybean production has been trending down since the introduction of herbicide- tolerant soybeans in 1996, declining from 1.18 pounds in 1997 to 1.06 pounds in 2000. In contrast, glyphosate use has been steadily increasing from 0.23 pound to 0.59 pound during this period. As a result, U.S. farmers used 75.2 million pounds of herbicides on soybeans in 2000, a decline from 78.2 million pounds in 1997 despite a larger planted acreage for the 2000 crop. The use of glyphosate increased from 14.9 million pounds in 1997 to 41.8 million pounds in 2000 (USDA, 1998 & 2001; fig. 5). The impact of adopting herbicide-tolerant soybeans on herbicide use is not completely clear. The main reason is that while the adoption of glyphosate-tolerant soybeans is estimated to lower the use of 'other herbicides' by 14 percent and have no effect on acetamides for adopters, it is--estimated to raise the use of glyphosate by 43 percent. Since average application rates vary across herbicide active ingredients, the net effect of substituting one for another may be an increase or a decrease in total pounds used. As a result, the impact on herbicide use is being pulled in opposite directions, depending on whether the decrease in 'other herbicide' use outweighs the increase in glyphosate use. The offsets are most likely in soybean-producing areas that applied herbicides at lower rates and glyphosate at higher rates than the national average--mostly in the South. For example, while Heartland producers may have traded off glyphosate for other herbicides, Southern producers were applying much lower rates of herbicides (e.g., 5.72 lb per acre per crop year in Arkansas in 1997 vs. 9.26 lb in Illinois) but higher rates of glyphosate (e.g., 1.01 lb per acre per crop year in Arkansas vs. 0.66 lb in Illinois). So, when Southern producers started adopting glyphosate-tolerant soybeans and using more glyphosate, there was a much larger proportional increase in total herbicide use, given the differing elasticities estimated for these ingredients. The USDA-NASS chemical usage data supports this relationship (USDA, 1998[a]). Results of the analysis show that the impact of adopting herbicide-tolerant soybeans on herbicide use is mixed--a decline for the Heartland and Prairie Gateway, but an increase for all other regions (fig. 6). Overall, the impact is a slight 3-percent increase in herbicide use (measured in pounds of active ingredients) Nationwide as a result of adopting herbicide- tolerant soybeans. Measuring herbicide use in pounds of active ingredients implicitly assumes that a pound of any two ingredients has equal impacts on human health and/or the environment. However, 'other herbicides' being replaced by glyphosate, as a result of the adoption of herbicide-tolerant soybeans, are at least three times as toxic and persist in the environment twice as long as glyphosate (Heimlich et al.). Thus, adoption of herbicide-tolerant soybeans may lead to positive environmental and health benefits. The decline in herbicide use (in terms of active ingredients applied) for adopters ranges from 1.61 percent in the Heartland region to 3.26 percent in the Prairie Gateway. For example, the use of glyphosate increased from 0.19 pound per acre for nonadopters to 0.27 pound per acre for adopters in the Heartland (table 3). In contrast, the use of 'other herbicides' decreased from 0.70 pound per acre for nonadopters to 0.60 pound per acre for adopters (table 3). The end result is a decline in the use of all herbicides from 1.00 pound for nonadopters to 0.98 pound for adopters, or a decline of 1.61 percent, which is lower than the 8-percent decline in herbicide material costs for adopters. The increases in herbicide use for other regions are as follows: 8.35 percent in the Northern Crescent; 11.74 percent in the Mississippi Portal; and 10.89 percent in the Southern Seaboard. Herbicide-tolerant soybeans offer producers the simplicity and flexibility of weed control, which largely explains why U.S. farmers have adopted this technology so rapidly. Adopters of this technology can rely on one to two post-emergence herbicide applications, instead of three or more, to control a broad spectrum of weeds without crop injury. In addition, the adoption of this technology is comparable with the adoption of conservation tillage practices and narrow-row plantings and imposes no restriction on crop rotation (Gianessi and Carpenter). Herbicide-tolerant soybeans also can reduce the foreign material content in soybeans, which is subject to price discounts. Conclusions Estimates of the farm-level effects of adopting herbicide- tolerant soybeans differ significantly, depending on the data source. A key challenge to analysts is to isolate the effects of biotech adoption so that estimated farm-level effects can be attributed solely to the technology itself. Estimates of the farm-level effects derived from the elasticity-based approach appear to be more plausible than the mean ARMS data because the farm-level impacts obtained from the former source are attributed exclusively to biotechnology. Nonetheless, the use of one year of data (1997) in this study has its limitations. As more data become available in the future, further analyses including multiple years would provide a more complete assessment of the effects of biotechnology versus management practices and weather. The farm-level effects of adopting herbicide-tolerant soybeans on crop yields vary across data sources and production regions. Adopters' yields are estimated to be not much different from nonadopters'--only a 3-percent increase for adopters. Adopters' weed control costs in the Heartland are estimated 11 percent lower than for nonadopters (a savings of $3.50/ac.), with the effect ranging from a 1- to 34-percent reduction in other regions. The cost savings have implications for the price premiums necessary to encourage farmers to plant non-biotech soybean varieties instead of the herbicide-tolerant varieties. References Breitenbach, F. and T. Hoverstad. 'Roundup Ready Soybeans,' Crop News, 4(29): pp.162-163. December 18, 1998. Carpenter, J. and L. Gianessi. 'Case Study in Benefits and Risks of Agricultural Biotechnology: Roundup Ready Soybeans' National Center for Food and Agricultural Policy (2000). http://www.ncfap.org/pup/biotech/icabr.pdf 'Herbicide Tolerant Soybeans: Why Growers are Adopting Roundup Ready Varieties.' AgBioForum 2(1999a): pp.65-72. 'Why U.S. Farmers are Adopting Genetically Modified Crops.' Economic Perspectives (1999b). http://www.ncfap.org/pup/biotech/economic%20perspectives.htm Duffy, M. and A. Vontalge. 'Estimated Costs of Crop Production in Iowa 1999.' Iowa State University Extension, Fm-1712, January 1999. Http://www.extension.iastate.edu/Publications/FM1712.pdf. Falck-Zepeda, J. B., G. Traxler, and R. G. Nelson. 'Rent Creation and Distribution from Biotechnology Innovations: The Case of Bt Cotton and Herbicide-Tolerant Soybeans in 1997.' Agribusiness 16(2000): pp.21-32. Fernandez-Cornejo, J. and W. McBride. Genetically Engineered Crops for Pest Management on U.S. Agriculture: Farm-Level Effects. AER-786, ERS-USDA. April 2000. Fernandez-Cornejo, J., C. Klotz-Ingram, and S. Jans. 'Farm-Level Effects of Adopting Genetically Engineered Crops in the U.S.A.' Presented at the International Conference 'Transitions in Agbiotech: Economics of Strategy and Policy' NE-165. Washington, DC. June 24-25, 1999. Gianessi, L.P. and J. Carpenter. Agricultural Biotechnology:Benefits of Transgenic Soybeans. National Center for Food and Agricultural Policy. April 2000. Gianessi, L. P. and M. B. Marcelli 'Prices of Pesticide Active Ingredients (1996)' National Center for Food and Agricultural Policy Oct. 1996. Heimlich, R.E., J. Fernandez-Cornejo, W. McBride, C. Klotz- Ingram, S. Jans. 'Genetically Engineered Crops: Has Adoption Reduced Pesticide Use?' Agricultural Outlook, ERS-USDA, Aug. 2000. Marra, M., G. Carlson, and B. Hubbell. 'Economic Impacts of the First Crop Biotechnologies.' http://www.ag- econ.ncsu.edu:8250/faculty/marra/FirstCrop/sld001.htm McBride, W. D. 'Using Cost and Return Data in Estimating Benefits from Biotechnology Adoption.' Presented at the American Agricultural Economics Association Annual Meeting, Tampa, FL. July 30 - August 2, 2000. McBride, W.D. and N. Brooks. 'Survey Evidence on Producer Use and Costs of Genetically Modified Seed,' Presented at the NE-165 Conference on Transitions in Agbiotech: Economics of Strategy and Policy in Washington, D.C., June 24-25, 1999. Moschini, G., H. Lapan, and A. Sobolevsky. 'Roundup Ready(r) Soybeans and Welfare Effects in the Soybean Complex.' Agribusiness 16(2000): pp.33-55. Oplinger, E. 'Performance of GMO's in Northern U.S.,' University of Wisconsin, Department of Agronomy, 1998. Rawlinson, J. and A. Martin. 'Weed Management Strategies in Soybeans,' University of Nebraska, Nov. 10, 1998. U.S. Department of Agriculture, NASS. Agricultural Chemical Usage: 1997 Field Crops Summary. Ag Ch 1(98) May 1998(a) & May 2001. NASS. Agricultural Prices: 1997 Summary. Pr 1-3(98) July 1998(b). - END_OF_FILE