 |
|
|
APPENDICES
|
||||||||||||||||||||||||||||||||||||||||||
|
|
Appendix 1 - Economics of separating GMOs on the farm
"Increasingly, GMOs are, in our opinion, becoming a liability for farmers. A two-tier market for grain may develop, where the GMO 'improved' grains will trade at a discount to non-GMO product.... If GMOs trade to a discount, the price premium for the seed may collapse, and they would trade at a discount to traditional varieties and hybrids. That would be an earnings nightmare for Pioneer Hi-Bred and, we would guess, for Monsanto as well" - Deutsche Bank report 'GMOs are dead' [143]
Average farm-gate price: $3.20 / bu = $117 / t Ave export price: $4.80 / bu = $175 / t (1999 prices) Ave farm-gate price: $1.60 / bu = $62 / t Ave export price: $2.30 / bu = $90 / t GM seeds are more expensive than conventional seeds (although non-GM seeds will become more expensive, the lower the allowable tolerances of GM content). Seed companies hope to recoup the large investments they have made in research and development of the new varieties. In turn, it is expected that farmers will justify the extra seed expenditure through higher incomes (from greater yields), and lower other input costs (such as herbicide and pesticide). Some recent studies indicate that this may not be the case, and show either increased herbicide / pesticide requirements or lower yields associated with the GM varieties, or alternatively show that the increased income is smaller than the net increase in input costs [144]. It is clear that the overall effect of GM on production varies between different growing regions, in particular depending on an area's weed and pest profile. But the high rate of uptake by farmers of the new crops indicates that in many cases the GM crops do indeed increase the farmers' margins. Separation costs On top of the differences in the intrinsic economics of growing GM or non-GM crops, farmers will incur costs in keeping the two separate. These costs have been estimated by researchers from the University of Illinois. [145] The study reckons that cleaning a planter takes one hour, with a labour cost of $15. If he just cleaned once (ie did all GM plantings, then cleaned, then did all non-GM plantings), for a crop of 500 acres of soyabeans yielding 40 bushels per acre, the extra cost would be just 0.08 cents per bushel ie negligible. In harvesting, using a 'flush' method (running non-GM through combine harvester to clean out leftover GM grains, then sell this run-through as GM), the combined cost of labour and lost premia on the flush add up to 0.09 cents a bushel. The types of trucks used to take grain from farm to elevator are designed to easily dump and easily sweep clean; the situation in which this stage might create extra costs is if there are longer queues at the elevator due to GMO testing, in which case he might have to stop the harvest while waiting for the truck to return. Thus cleaning of equipment introduces no significant extra costs although it may be inconvenient by requiring extra work at the two times when the farmer is at his busiest, planting and harvest. If harvests are stored on-farm, capital expenditure may be necessary for more bins. US Secretary of Agriculture Dan Glickman has called for federal loans to support farmers investing in new bins for GMO segregation. Some farmers have now decided not to grow GM at all, so they avoid the need to segregate altogether. Perhaps the biggest problem is the risk of cross-pollination of maize, over which the farmer has limited control - it depends on his neighbours. He can either grow large barrier crops or lose flexibility in where he can plant (his non-GM crops will have to be away from neighbours' GM varieties). Opportunities for contamination of non-GM crop Below are listed the stages at which contamination of a non-GM crop may occur, together with estimates (analysed by Dr Charles Hurburgh of Iowa State University) of how significantly it might be contaminated if care is not taken (most of these %ages will be reduced by extra effort or cost) [146]: 1) Seed: Soyabean seed 0.1 - 0.2% contamination. Maize seed 1% contamination (seed company confident estimates) [147]. Soya (being self-pollinating) is far easier to keep separate than (wind-pollinated) maize. 2) Planting: suppose 500 seeds remaining in a planter box (not properly cleaned) à 0.6% contamination. 3) Cross-pollination: maize pollen can easily travel a few hundred metres. Allow 0.3 - 0.5% contamination if separation distances are greater than that (and much higher if distances are lower). 4) Harvest: combine harvester may contain 3-5 bushels (bu) if not fully cleaned. 5) Wagons: 1 bu remaining in wagon à 0.2% contamination of 500 bu load. 6) Storage bins: 30 bu remaining à 0.3% contamination of 10,000 bin. 7) Elevator: 10-100 bu contamination depending on size of equipment à 0.1% in 10,000 - 100,000 bu bin. 8) Accidental mixing: one 800-bu truck à 1% of an 80,000 bin. 9) Export elevator: 0.1% There should be minimal contamination in railcars, barges and ships, as these are cleaned out before each use, whatever the cargo. If this level of contamination with GM crops occurs at each of these levels, the final crop will arrive at the processor containing 3.8% GMO. Premia Since production of non-GM does involve extra effort and cost for the farmer, and potentially less favourable growing economics, he will need an incentive. Thus some elevators pay farmers a premium price for non-GM harvests, or equivalently, discount the GM crop price. The farmer of course wants to maximise his profit, so will weigh the size of premia against production costs. Estimates of the premia paid at the farm gate (by Cargill, ADM and others) in 1999 vary from 3-10% - depending on the growing region, target market etc. This is for post-harvest separation (ie with no growing contract), to meet demand from Japan for crops that were at least 95% non-GM. [148] For tighter tolerances, according to agricultural consultancy PG Economics, estimated costs cited range from 15-25% of the farm-gate price for tolerances of 1-2% GMO, to 150% where the tolerance set is no detectable residue (in reality equal to about 0.01 per cent tolerance which is about the limit of current commercial testing). [149] Contract growing Often non-GM crops are grown under contract between farmer and either elevator, exporter or customer (food processor / packager), where the contract sets a pre-determined price to be paid for the harvest. Up till now, the majority of the contracts have been with the customers (especially in Japan). This has necessitated an identity-preservation form of supply, rather than segregation, as a specific batch of produce must travel from the farm to the buyer who has ordered it. However, we should expect to see an increase in contracts between elevator and grower, and between exporter / processor and elevator, and these will favour a segregation-based approach, with a larger, continuous flow of non-GM grain. There are also moves towards contract-free supply of non-GM on an open market - as in for example the Tokyo Grain Exchange selling non-GM soyabean futures. This too leads to segregation rather than IP. Cargill, which favours identity preservation over segregation, has all of its IP non-GM crops grown under contract. [150] In 2000, 1.5 million acres non-GM soyabeans were grown under contract through DuPont Speciality Grains, representing 2% of US plantings. [150] In these cases, Consolidated Grain and Barge, ADM and Protein Technologies International (a DuPont subsidiary) signed contracts with both growers and elevators, specifying the production practices required (keeping seed receipts, cleaning equipment, training staff, applying clear segregation or IP systems etc) and the premia (10-30 cents a bushel at farm gate, plus about another 5 cents for the elevator). Risk management Since a farmer must decide what to plant at least six months before he tries to sell it, he carries a lot of risk, and the uncertainty in the GM debate has a major impact on farmers. It is in this way that the farmer has been a victim of the controversy. On the one hand, there is the risk that rejection of GM crops will increase (eg by expanding into animal feeds), so if the farmer plants GM he may find there is no market for his harvest. On the other hand, he might opt for non-GM, incur the extra costs, and then discover that there are no non-GM premia available; and with margins in farming already extremely tight, this is an outcome the farmer simply cannot afford. Growing contracts for non-GM crops provide one way out of this risk. Farmers will also look to ease their uncertainty from guarantees by the large grain-handling firms. For example, Cargill promised farmers that for crops planted in 1999 and 2000, all of its grain elevators would accept GM maize and soyabean varieties [150], and this did much to calm farmers' nerves over GMOs. The ADM announcement calling for separated non-GM supplies in September 1999 had the opposite effect. Summary In summary, the feasibility of non-GM production depends principally on three regional / environmental factors: 1) the degree of weed and pest problems (and hence the degree of economic advantage to GM crops); 2) the planting decisions of neighbouring farms (determining the potential for cross-pollination); 3) the availability of premia for non-GM from local elevators. Thus GM and non-GM supplies may perhaps be achieved most efficiently - especially for maize where the potential for cross-pollination is high - by the dividing up producing countries into regions where GM is allowed and regions where it is not. |
||||||||||||||||||||||||||||||||||||||||||
| Appendix 2 - Supply chain economics of segregation and identity preservation (IP)
It is widely agreed that the greatest costs of segregation or identity preservation occur post-farm gate. How many elevators segregate? One of the biggest bottlenecks in the supply chain is the elevator sector, and a number commentators have suggested that most elevators will not be able economically to separate GM from non-GM harvests. However, the number of elevators which keep non-GM separate has been steadily rising. • A survey during the 1999 harvest by commodity research company Sparks Companies Inc found that 8% of Midwest grain elevators separated soyabeans and 11% maize. 3% were paying a premium for non-GM soya, and 1% for maize. [153] • A survey of the 2000 harvest, commissioned by seed company Pioneer Hi-Bred International, found that 25% of elevators said they would separate GM from non-GM maize, and just over 20% soyabeans. 10% would pay a premium for non-GM maize, and 15% for soya. What's more, 12% of elevators will not accept Roundup Ready soya at all, and nearly 20% won't take Bt or LibertyLink maize. [154] • Later in 2000 a survey by the American Corn Growers Foundation found that 42% of elevators require or suggest on-farm segregation; 31% require or suggest separation at the elevator gate; and 22% are paying non-GMO premia. [155] Mechanics of separation at the elevator Elevators have been designed for single streams of each crop, and some have had difficulty separating non-GM. Part of the problem is that the elevator business operates to very small margins, so relies on large volume turnover, and segregation (in particular testing) will slow operations down. Also, it obviously gets extremely busy at harvest time, and so adding the new complication will be difficult. While cleaning out pits (into which farmers empty their trucks) and bins (in which grains are stored) before handling non-GM crops is quite possible, cleaning 'legs' (the mechanical conveyor system with built-in buckets / shovels, which lifts grain from pits to bins) is much more difficult, as there are many working parts in which grains can get stuck. Preserving crop identity within a low level tolerance for GM contamination may not be economically feasible by cleaning, as it would probably require dismantling the leg's many parts, and there is not time for this at harvest. Depending on the layout and mechanics of the elevator, it may be possible to flush GM grains out of the leg by running a quantity of non-GM grain through it into the standard, GM / mixed bin, before beginning the non-GM run. One option for handling non-GM crops is to accept only non-GM on certain days of the week, to avoid the need for two separate queues of farmers, and to reduce the risk of commingling. This approach lends itself to IP, where the same facilities are used for both GM and non-GM, and cleaned in-between. Direct costs of segregation and IP at the elevator Segregation of non-GM crops is a new activity, and there is still insufficient information available to accurately describe the economics. Estimates of the costs vary widely, as do the available premia. However, recently a number of studies have been published, [156] and this is certainly a growing area. One attempt to estimate the extra cost of IP, based on semi-empirical data, is shown in the table below. This excludes premia paid to farmers - ie the table shows post-farm gate costs. Researchers at the University of Illinois surveyed 35 elevators in spring 1998 [157] for their estimates of the costs of IP for high-oil corn (maize) and STS soyabeans (both non-GM speciality crops). A study by the Economic Research Service of the US Department of Agriculture [158] then applied theoretical considerations to translate these figures into the costs of IP for non-GM. Table - extra cost (in US $) per bushel at a grain elevator of identity-preserved handling of non-GM soyabeans and maize. Most non-GM crops are still exported, as most demand for non-GM is outside the USA. A large number come from farms near rivers, so they are taken by truck from the farms to river elevators, then by barge to an export elevator, where they are loaded onto a ship. Some others are delivered from the farm to a country elevator, from which they are then carried by train to a river elevator. Obviously, segregation or IP incur more costs with the three-elevator process than with two elevators. Assuming the per bushel costs above are applicable to each of the three different types of elevator [159], we add up the elevator cost (ie not including premia paid to farmers, nor transportation or processing costs) of IP in the table below. The elevators able to handle non-GM grains most efficiently will be those which are able to dedicate some facilities to GM and some to non-GM; or even dedicating whole elevators, by dividing all the elevators in a locality between GM and non-GM - ie employing a segregation approach. These latter elevators, by specialising, will avoid the extra storage, handling and marketing costs of IP, leaving just testing and risk management costs. What's more, the original University of Illinois study showed that by specialising, risk management costs fall to zero for maize and $0.01 for soyabeans. The former segregation approach (dedicating facilities within each elevator) will incur extra costs somewhere between those of IP and those of dedicated whole elevators. Table - total extra costs ($ / bu) at all elevators through supply chain, of segregation or IP of non-GM crops
Indirect costs of segregation or IP at the elevator On top of the direct costs outlines above, there may be hidden, opportunity costs to handling IP non-GM crops, arising from the loss of flexibility. These have been analysed by two researchers at the university of Missouri-Columbia [160], and consist of three types: 1) storage margin opportunity cost. The elevator sector in the USA has massive over-capacity, meaning that a large quantity of storage space is just not being used. Maximisation of the proportion of an elevator's available capacity which is actually used is therefore a major competitive issue. Clearly, handling a greater number of distinct types of grain (eg GM and non-GM) will reduce an elevator's ability to fully fill its bins. 2) grind margin loss. Margins in elevator storage are small, so some elevators try to boost their margins by grinding some of their stored crops into feed, which they then sell on to local livestock farms. Feed demand in the USA is mostly serviced by the bulk commodity crop, so grinding of more valuable non-GM crops for feed would write off their extra value. 3) spread opportunity cost. One of the important skills an elevator manager uses to make his profits is exploiting 'spreads' in the market - in other words, holding onto stocks when the expected future price of the commodity is higher than the sum of the current price and storage and lost interest costs. Since most non-GM crops are sold under contracts, which specify delivery times, the elevator manager loses his ability to hold grain for spreads. The study looks at three different elevator configurations, and estimates that spread opportunity costs can range from $ 0.07 to 0.22 / bu, and storage margin opportunity costs can range from $0.06 to 0.15 / bu. However, it should be noted again that for an elevator specialising in non-GM only, there is no storage margin opportunity cost. Probably also, assuming a dedicated elevator has some flexible supply and sale (ie does not only sell through contracts), spreads will still be possible too, although opportunities for spreads will be smaller for smaller markets. Thus in the segregation case the major hidden cost is the loss of grind margin. Transportation Like with elevators, much extra cost in the transportation phase of separating non-GM is due to loss of flexibility, rather than extra physical work or resources being required [161]. Railcars, barges and ships are used for a wide range of different cargoes, so are routinely cleaned out in between, whatever they are carrying. Thus cleaning adds no extra costs to segregation / IP for non-GM grains. Testing costs, as we have noted are low. Barges do not have segregated holds, so cannot simultaneously carry both GM and non-GM grains. Capacity tends to be around 55,000 bushels of soyabeans or maize (elevator storage bins hold about 100,000 - 400,000 bushels). Extra costs will arise from partially filled barges eg a small elevator with 130,000 bushels each of GM and non-GM will require six barges (three for each), whereas 260,000 unsegregated will only require five. Ships, on the other hand, often do have separate holds, each hold generally carrying 200,000 - 300,000 bushels, or 5000 - 8000 tonnes. Again, the export company will aim to fill each hold as full as possible, which should be possible, as processors' orders will tend to be around 50,000 tonnes - as long as this matches the storage capacity in the destination country. Premia We noted differing values of farm-gate premia for non-GM crops. There is even more variation in estimates of the add-on costs of IP through storage and transportation. George Henni, Cargill's protein manager for Europe, reports that non-GM soyabeans, delivered at Rotterdam at a 1% tolerance, are currently demanding premiums of $12-17 per tonne ($0.33 - 0.47 per bushel), or about 5-10 % of the price [162]. GMO testing company Strategic Diagnostics Inc puts the price rather higher, at approximately $16 per tonne ($0.44 / bu) over the Chicago Board of Trade price, PLUS another $10-15 per tonne ($0.28 - 0.41 / bu) in transport costs [163]. John Haas, premium grains specialist at Consolidated Grain and Barge, is the least optimistic, and states that in 1999 exporters received premia from foreign processors of $8 per tonne ($0.22 / bu) for non-GM soyabeans [164]. Cargill's UK subsidiary reckons that for both soyabeans and maize, Japanese buyers were prepared to pay premia (for crops with 5% GMO tolerance) of up to $18 dollars per tonne (delivered Japan) - a fairly mid-range estimate, equivalent to 20% extra on maize, and 10% on soya [165]. Note that these premia need to cover extra elevator costs, transportation costs, and farm-gate premia. We can see that the premia comfortably cover the costs of segregation, but will only cover IP costs in some cases. Really? There are suspicions that costs are being over-estimated, in order to squeeze the non-GM market, as we noted earlier. We now see for example, Cargill's own poultry subsidiary Sun Valley is now selling chickens reared on non-GM feed (to the European standard) at no extra cost [166]. Since Cargill and ADM control so much of the supply chain, both horizontally and vertically, it is difficult to find out what the real costs are. The recent academic studies examining these economics, on which these appendices have drawn heavily, acknowledge that this is a new area, and much more study is needed, as well as practical experience. The most significant study we have referred to was based on a questionnaire for elevator managers. It may be that their costs will fall below those that they gave here, as they discover how to more efficiently run segregation and IP systems. Much of the cost of IP is incurred in setting up new supply chains - sourcing, training, testing etc. Summary It is difficult to accurately estimate the costs of separating non-GM crops through the distribution system, as there is so little experience available. We can conclude however that segregation is a significantly cheaper approach than identity preservation, and that additional costs tend to be more attached to loss of flexibility, or hidden opportunity costs, than to extra real physical inputs. |
|||||||||||||||||||||||||||||||||||||||||||
| Appendix 3 - Cargill and ADM's share of the market
ADM has a major focus on processing, especially high-value end-products. Thus it has more interests than Cargill in secondary processing, including animal feed, ethanol, and soya-based vegetarian foods. It is also a large player in animal feeds (especially in Europe), but not especially in meat production.
In 1998 Cargill sold most of its seed businesses to Monsanto - including seed research, production and testing facilities in 24 countries, and sales and distribution operations in 51 countries. The deal did not include American, Canadian or British seed businesses [167]. Shortly afterwards, Cargill struck a deal to sell AgrEvo (now Aventis) its North American seed business, Cargill Hybrid Seeds. The deal however was postponed when Pioneer Hi-bred (a DuPont subsidiary) sued Cargill Hybrid Seeds for theft of its patented seed, and AgrEvo eventually pulled out altogether in February 1999. Cargill finally settled with Pioneer for $100m in May 2000, which left it clear to try again to sell the division. It is currently looking for a buyer [168]. Cargill Hybrid Seeds began selling Bt maize and Roundup Ready soyabeans in January 1997 [169]. Through subsidiary Inter Mountain Canola, it also supplies Roundup Ready canola (oilseed rape) [170]. InterMountain Canola, based at a research station in Fort Collins, Colorado, is developing oilseed rape varieties with output traits - including oil that has stability, health and flavor advantages [171]. While the oilseed rape research station at Fort Collins is Cargill's biggest biotechnology lab, the company also researches maize (particularly accelerated breeding) at a seed research station in Aurora, Illinois, and sunflower and canola breeding (to improve disease resistance) at Boissay, Francecl [172]. Cargill greatly expanded its storage and distribution capacity when it acquired the grain businesses of rival Continental Grain in 1999. Globally, ADM has 750 elevators [179]. In South America it has 82 elevators with 1.3m tonnes storage capacity, plus leasing arrangements giving it a further 0.5m tonnes [180]. In addition to this, ADM adopts a strategy of forming joint ventures and partnerships with other storage and distribution companies, especially co-operatives, to extend the volume of grain it has access to. These include Growmark, Countrymark and United Grain Growers. Growmark, for example, is made up of more than 270 member-owner co-ops in grain marketing (plus over 120 in farm supplies retail) [181]. ADM owns 1000 trucks, 15,000 railcars and 2,200 river barges, mainly in North and South Americ a [182]. ADM has the 2nd largest fleet in the USA, after Artco, and followed by ConAgra subsidiary Peavey; these 3 companies control 53% of US covered barges [183]. On top of all this, it uses 5 million trucks and 500,000 railcars belonging to other companies each year [184]. The ADM / Countrymark partnership is one of the largest suppliers of feed grains and soyabeans to the southeast USA market [185]. ADM has taken on joint ventures and acquisitions in South America the last few years; their purchase of subsidiaries of rival Glencore in Brazil and Paraguay, for example, gave a 4% increase in their share of the world soyabean trade [186]. Cargill and its subsidiaries operate more than 40 oilseed processing plants worldwide. European operations include 11 crushing plants and 10 refineries, plus two bottling facilities, a hardening and a protein plant [188]. Cargill is the largest oilseed processor in the UK [189], and in Europe (25-35% market share) [190]. ADM is also the US' largest wet maize miller, and 3rd largest dry maize miller [197]. It runs three maize wet mills and two dry mills in the US that process 1.6 million bushels a day [198]. Globally, ADM has more than 355 processing operations [199]. Cargill controls 37% percent of the delivery points for the Chicago Board of Trade (i.e. the depots and facilities where all the goods traded on the exchange have to be actually delivered) [203]. The CBOT is by far the largest futures exchange in the world for grains and oilseeds; its prices are used as reference prices elsewhere in the world. Cargill owns 50% of storage capacity on the Illinois River (the delivery area for CBOT soya contracts), and 31% of capacity in the northern part of the river (delivery area for CBOT maize) [204]. Cargill also accounts for 25% of Argentina's maize, soya and wheat exports [205]. According to the European Commission, Cargill sells less than 10% of soyabeans, and 10-20% of rapeseed, although these figures only include the beans and seed traded as such, and not those which are processed 'in-house' [206]. Cargill subsidiary Hohenberg Bros is one of the world's largest cotton merchants, and sells cotton to ginners, buyers and textile mills in over 50 countries on all six continents [207]. ADM owns 75% of trading company AC Toepfer International, which conducts about 40m tons of trade from 43 offices worldwide [211]. Between them, ADM, Cargill and Zen Noh account for 81% of US maize exports, and 65% of soyabean exports212. Together, Cargill, ADM and Bunge have 59% of US port facilities [213]. ADM controls 33% percent of the delivery points for the Chicago Board of Trade (ie the depots and facilities where all the goods traded on the exchange have to be actually delivered) [214]. The CBOT is by far the largest futures exchange in the world for grains and oilseeds; its prices are used as reference prices elsewhere in the world. It owns 25% of storage capacity on the Illinois River (the delivery area for CBOT soya contracts), and 25% of capacity in the northern part of the river (delivery area for CBOT maize) [215]. ADM Investor Services is a full service futures commission merchant, fully represented on the main US markets [216]. ADM has 2 export facilities in South America, where it is the largest exporter of soyabeans [217]. Cargill is the largest European producer of crude oilseed meal, with 25-35% of total European production (although much of this is sunflower meal, which is non-GM) [220]. ADM and AC Toepfer International (see above) have a 10-20% share of the European animal feed market (which includes grain, oilseed meals, maize gluten, animal meal, fishmeal, citrus pulp etc), and a 20-30% share of the oilseed meals market. This includes 25-35% of the rapeseed meal market [222]. ADM and AC Toepfer subsidiary Arkady Feed (UK) are the 3rd and 5th largest feed companies in the UK. While it is difficult to calculate their market share due to both companies also being involved in other businesses, together with ABN (the 4th largest) the 3 companies have between 20% and 24% of the market [223]. It produces some maize products such as dextrose, starch, syrups and sweeteners, and soya products such as soya flour and textured vegetable protein [229]. However it has much less involvement in these processing markets than rival ADM. ADM is the USA's largest producer of ethanol, with 46% of the market [230[, which is used both in industrial applications and in drinks. ADM is the USA's largest producer of the grain alcohol used to make gin, vodka, and liqueurs [231]. Products manufactured for use by the food and beverage industry include syrup (eg maltodextrin), starch, glucose, dextrose, crystalline dextrose, high fructose sweetener, crystalline fructose and grits [232]. ADM's sweeteners form the 2nd largest ingredient (after water) in Coca-Cola [233]. In 1996 it paid $258 million for 22% of Mexico's Gruma S.A., the world's biggest maize flour maker [234]. ADM also produces the amino acids lysine, tryptophan, and threonine (used in animal feeds); the organic acids citric, lactic and glucono delta lactone; the industrial and food thickener, xanthan gum; and the vitamins C and riboflavin [235]. It is also very involved in soya-based vegetarian foods (ADM invented textured vegetable protein, TVP). In the USA, ADM owns the major vegetarian food brand Worthington Foods. UK ADM subsidiary Haldane Food has 4 factories; Haldane claimed in spring 1997 that its products were non-GM, but testing by the Daily Mail found GM ingredients in Haldane's RealEat and So Good ranges [237]. ADM's protein enhancers are common in pet foods [238]. Cargill owns Sun Valley, the largest poultry producer in the UK, with interests also in France and the Netherlands [241]. Cargill also has major poultry production in Thailand. |
