IMPORTANT LEGAL NOTICE: The information on this site is subject to a disclaimer and a copyright notice.
 
European FlagEuropa
The European Commission
en
Agriculture

WORKING  DOCUMENT Rev. 2
Directorate-General for Agriculture

Economic Impacts of Genetically Modified Crops on the Agri-Food Sector

3. Farmers : strong profitability expectations, mixed outcome

The adoption of GM crops by farmers in the US, Canada and in Argentina has proceeded at an unprecedented rate compared to the uptake of conventional hybrids. The economic reasons for this rapid and massive adoption are analysed in section 3.1. Farmers had strong expectations on the profitability of GM crops, in particular as regards yield and/or cost savings. However, as shown in section 3.2, GM crops do not prove to be significantly more profitable than conventional counterparts. Other factors than profitability play role. They are reviewed in section 3.3.

The analysis is based on the available economic literature, which mainly concerns Northern America. It is limited to the two main GM crops under cultivation Herbicide-Tolerant (HT) soybeans and Insect-Resistant (Bt) corn. Two Canadian studies on HT Canola11 have also been taken into account.

3.1.     Adoption of biotechnology by farmers : strong expectations

    3.1.1.  Profitability expectations mainly based on yields
    Many surveys and studies have been carried out to assess reasons for adopting GM crops. They have confirmed that adoption of GM crops by farmers has been driven by profitability expectations.
    According to an USDA survey (1997), the majority of farmers (50 to 75%) cited increase in yield as first reason for adoption. Savings in costs appear to be the second reason, mentioned by 20 to 40% of the respondents. This survey was conducted in 1997, only one or two years after the introduction of the first GM seeds on the US market. Therefore, it addressed farmers' expectations.
    The quick rate of adoption in the first years is explained by the strong expectations of farmers as regards profitability. Whether they definitely adopt the new technology then depends on their degree of satisfaction, and in turn, on the effective profitability of the crop. Biotech firms have published encouraging results on the satisfaction rate of farmers having adopted GM crops (Monsanto, 1998).
    In practice, the most immediate and tangible ground for satisfaction appears to be the combined effect of performance (not necessarily measured by yields) and convenience of GM crops, in particular for herbicide tolerant varieties. These crops allow for a greater flexibility in growing practices and in given cases, for reduced or more flexible labour requirements. Where labour or time is a restriction, this convenience effect has an economic impact. In the medium term, it should translate into increased labour productivity and savings in labour costs. In the long run, it might have an impact on farm restructuring, alongside with many other factors which play a role in this process.
    The effective profitability of a GM crop can only be properly assessed on the basis of several years of cultivation and commercialisation. Several years have to be considered for two main reasons. First, many other factors have an impact on profitability. In particular, there are important yearly fluctuations in yields and prices. Second, effective profitability depends on developments on the supply and on the demand side.
    The first generation of GM crops is input-oriented. The primary effects of this new technology were expected and observed on the supply side.
    Bullock and Nitsi (1999) consider five possible effects of technical changes in the field of plant breeding:
    1. Increase in the maximum yield
    2. Increase of the economically optimal yield
    3. Input-switching technical change, lowering the cost but yield neutral
    4. Quality-enhancing technical change
    5. Risk-reducing technical change.
    According to these authors, Bt corn falls under category 2, while Herbicide-Tolerant soybeans rather have a type-3 effect. Both types of effects imply a shift in farmers supply functions. Under given prices, farmers produce more. If the demand function remains unchanged, prices drop. Only type-4 technological change induces a structural change in the demand function, and possible increases in prices. When assessing the profitability for farmers and the economic impact of biotechnology on agri-food markets developments in supply and in demand have to be considered together. However, it appears that this has not always been the case, neither for farmers, nor for the leading biotech firms. Their approach has been supply-oriented.

    3.1.2. The effect of agricultural policy: limiting price risk

    In the US as well as in the EU, GM and non-GM crops are not treated differently under the various support schemes, both are eligible. In the US, crops for which GM varieties have rapidly developed are all eligible for support under the flexibility payments, the marketing loan system, as well as for crop insurance.

    Soybeans became eligible for flexibility payments and under the marketing loan system in 1996, which is the year of first commercial sowings of GM varieties. Several analysts (FEDIOL, 1999) consider that existing support systems have favoured the development of soybeans sowings. In particular, the loan rate applied to soybeans makes this crop attractive compared to wheat and corn. The area under soybeans is expected to reach a record level in 2000, while prices are low. By mid-November 1999, the USDA estimated that 90% of the 1998 soybeans crop had received a marketing loan benefit, and that the average value of this benefit was worth around 0.44 US $/bushel (14.5 €/t). Oilseed producers are also eligible for the 1999/2000 emergency packages. A specific assistance programme was set up in early February for oilseeds producers, to offset record low market prices. Under this programme, payments for soybeans could average 0.141 US $/bushel (5.3 €/t), according to calculation by private consultants.

    Favourable support conditions for soybeans could have played a role in the rapid uptake of GM technology for this crop. In addition, in a low market price context, the expectation on cost savings is a further driving force for the adoption of the technology.

    Eligibility of GM crops under various support schemes limits the price risk of the productivity-enhancing technology. It accounts as another reason for the farmers to focus their planting decision on expected farm-level performance, on cost-efficiency of inputs. In other words, farmers also had an input-oriented approach.

    3.1.3. Comparing the profitability of GM and non-GM crops proves difficult

    Profitability is defined as the margin left over to farmers when costs have been deduced from receipts. The profitability of a GM crop is judged against corresponding conventional crops. Comparing the performance of both types of crops raises several methodological issues.

    3.1.3.1.On the cost side : the input-effect of GM crops

    Generally the cost comparison of GM crops and their conventional counterpart is limited to crop-specific costs, assuming that fixed costs are more or less the same.
    GM seeds are sold at a higher price than conventional ones. The price wedge is mainly attributable to the value of GM technology or to the "technological fee". According to a Monsanto communication (1998), the technology fee reflects "the insect, weed, disease control value of the inserted gene, and a significant part of the fee is used for further research".This difference also reflects the fact that markets for GM and conventional seeds are separate. Furman Selz (1998) reports about premia observed on the US market in 1998: US $ 30 per bag of seeds for GM corn and US $ 5 for GM soybeans seeds, which represents a 30% price-premium compared to non-GM seeds. They also give an indication on the average technological fee paid by seed companies to gene providers: US $ 27 (30% of GM seed price) per bag of corn seeds and US $ 4.25 (21% of GM seed price) for soybean seeds. Despite of the technological fee, GM seeds appear to be more profitable than conventional ones for seed companies.
    The above-mentioned convenience effect of GM crops allows for reduced or more flexible working requirements. However, the related savings in labour costs have not always been properly assessed. The valuation of family work is rarely broken down on a crop-specific basis. On the other hand, growing GM crops requires new management skills, growing practices and possible constraints. GM seeds are generally sold and sown in the context of contracts. These changes entail transaction and management costs, which are not easy to assess.
    GM crops are expected to allow for cost-savings through reduced insect and weed control and/or to achieve higher yields. Under the assumption that the price of non-GM and GM crops is the same12, the latter will become more profitable for farmers if the increased seed costs are offset by savings in weed and/or pest control costs and/or by higher yields.
    3.1.3.2. On the receipt side : yields and prices

    Yield is a key factor for profitability expectations and results. In fact, available figures on crop-specific costs are often broken down on an area basis, while prices are paid on a quantity basis. Based on yields, costs and prices are brought on a common basis, often per acre/ha. In other words, the effect of possible increase in yields is taken into account on the receipt side.

    Comparing yields of GM and non-GM crops is not a straightforward exercise. Yields depend on a large number of factors, and the inserted trait of GM crops is only one factor amongst others. It is worth recalling (OECD, 1999) that first generation genetic modifications address production conditions (pests, weeds), they do not increase the intrinsic yield capacity of the plant. In other words, referring to Bullock's classification, they do not induce a type 1 (maximum yield) technical change. Not surprisingly yield performance of GM crops against their non-GM counterparts depends on growing conditions, in particular on the degree of infestation in insects or in weeds. Data about yields of GM crops are widely available, however, often specifications on factors which influence yields are missing, such as temperature13, weed control applied etc.

    The USDA (1999) has examined different factors affecting the adoption of GM crops. These include farm size, education and experience, location of the farm, use of production or marketing contracts. In the case of herbicide resistant soybeans, the USDA has concluded that "larger operations and more educated operators are more likely to use herbicide tolerant soybean seeds". Such differences between adopters and non-adopters of biotechnology have to be taken into account when comparing yields and returns obtained on both types of farms. This study on factors of adoption served as a first step for assessing the impact of GM crop on farmers' returns and on the environment. It allowed for controlling statistically these exogenous factors and carrying out multivariate regressions for assessing aggregate impacts of GM crops on yields, profitability and the environment. Results of this USDA study are indicated below, for each type of GM crop.

       
    Another key factor on the receipt side is the market price of GM crops. In many profitability studies, prices of GM and non-GM crops are assumed to be equivalent. Most of the available studies are based on 1997 or 1998 data. In these first years of commercialisation of GM crops, their impact on commodity prices has not been manifest or is difficult to assess. Different pricing developments between GM and non-GM crops have only been observed in 1999. However, very few market reviews report on a regular basis about such developments. The question of price premiums/discounts will be addressed in chapter 5.

    A further issue would be the full assessment of costs and benefits of GM crops, including effects on welfare as well as non-market effects, particularly risk assessment and management. However, the studies reviewed below only cover on farm profitability in the short term.

3.2.    Costs and benefits for farmers for selected GM crops
The results of various North-American publications on the profitability of GM crops are summarised hereafter. The review is limited to the two main crops under cultivation, respectively Herbicide Tolerant (HT) soybeans and Insect Resistant (Bt) corn. In addition, some Canadian studies on rapeseed/Canola have also been included.
    3.2.1.Herbicide Tolerant Soybeans
    Three different types of GM soybeans have been authorised in the US. Two of them are tolerant to different herbicides. Soybeans tolerant to glyphosate, Monsanto's "Round up Ready"(RR) soybeans, have been on the market since 1996 and are the most widely grown (estimated 80% share in GM soybeans). The third one is a high oleic soybean variety.

    3.2.1.1.  Lower yields

    One of the reasons for the rapid adoption of GM soybeans has been the expectation of a higher yield than for non-GM soybeans. A number of US research projects have addressed this issue. Results seem to indicate the reverse: in most field trials the GM crop shows lower yields than the non-GM crop, as indicated in the table below, in the case of Roundup Ready (RR) soybeans.
     

      Table 3.1 Differences in yields between conventional and GM soybeans


            Source: Benbrook, 1998, based on Oplinger
    Similarly, according to Benbrook (1998) in South Minnesota, average performance of top yielding Roundup Ready soybean varieties was 3% lower than the top yielding conventional varieties, yet in Central Minnesota the yield drag was as much as 13% and in Southern Wisconsin 6%. While indicating lower yields in each case, these sub-regional results point to the great variability in yield performance.

    In Kansas, the yield drag varied between 2 and 11% in favour of non-GM soybeans, as indicated by Hofer et al. (1998):

Table 3.2 Differences in yields
between conventional and GM soybeans, Kansas
    Duffy & Ernst (1999) conducted a "cross sectional survey" among 800 farmers in Iowa, based on interviews and field observations. It was not a side by side observation of GM and non-GM crops and should provide reliable estimates at state level. The average yield reported was 3.43 t/ha for those farmers who grew non GMssoybeans versus 3.29 t/ha for those who grew GM soybeans.

    The USDA estimated, on the basis of the 1997 data, that the increased use of HT soybeans produced only a small global increase in yields.

    One of the explanations given for the lower yield of GM crops is that the GM-traits were initially not introduced in the top yielding varieties of soybean. Seed companies are now incorporating these traits in their yield-leading varieties. If this is indeed the case, then the yield drag should diminish in the coming years.

    3.2.1.2.  Reduced herbicide use and costs

    In the 1960s herbicide use started to replace tillage and cultivation practices as a primary means of weed control. At that time, these were mainly pre-emergence herbicides.

    The use of post-emergence herbicide in the production of soybean has been rising steadily since they became available in the 80s. In 1988, 44% of soybean acres were treated, by 1994 this share had risen to 72%. Quite often, they were used in combination with pre-emergence herbicides.

    However these classical herbicides had a number of drawbacks:
     
    • difficult management
    • risk of crop damaged
    • development of herbicide resistant weeds
    • some herbicides limit the possibility of crop rotation.
    The emergence of GM soybeans which are tolerant to glyphosate ("Roundup") has a significant impact on the use of other herbicides. For instance, the use of imazetaphyr ("Pursuit"), one of the most widely used post-emergence herbicides has declined from 44% of soybean acres in 1995 to 17% in 1998. The main advantages of using Roundup on HT soybeans are:
     
    • a wider window of application, both in terms of stage of growth of soybeans and effective control of larger weeds,
    • the easier management of weed control programs,
    • the fact that there is no carry over, thus giving growers more rotation options.


    The use of this product has increased drastically. In 1990, about 10% of all soybean acreage were treated with Roundup (at that stage used only as "burndown" treatment). This figure has risen to 45% in 1998 (Carpenter & Gianessi, 1999). According to the USDA, the use of other synthetic herbicides have declined by a larger amount, and the net impact of increased cultivation of HT soybeans is a decrease in overall herbicide applications.

    The cost of a program of Roundup on HT soybeans was 14.7 €/acre (36.6 €/ha) in 1998, compared to 12€/acre (29.8 €/ha) for a conventional program with pre-plant treatment alone, or 22.3€/acre (55.2 €/ha) for programs using other combinations).

    However, due to emergence of resistance in the future additional treatments may be needed. From 1998 to 1999, an increase from 15 to 25% in terms of average pounds of Roundup/acre was observed. Benbrook reports an increase from 24 ounce/acre to 32-48 ounce/acre in the dose of Roundup Ultra required to gain adequate control of velvetleaf and ragweed species. This would clearly have an impact on the cost of GM crops.

    Nevertheless, in the short term, the cost saving effect seems to be dominant. In the Duffy report, farmers who used GM crops reported spending nearly 30% less than those who grew non-GM soybeans. Reduced herbicide costs was listed by 27% of farmers as one of the reasons for planting GM crops. Furman Selz reports a 33 to 35 €/ha lower herbicide cost for HT soybeans.

    Moreover, following the introduction of GM crops, there is a notable reduction in the price of weed control programs for non-GM crops. A University of Illinois study revealed that compared to 1995, the least expensive non-glyphosate herbicide program was between 4.5-6 €/acre (11-14.9 €/ha) cheaper in 1999. As indicated by Bullock et al. (1999), this means that non-adopters of HT crops might also benefit from an induced effect on cost savings.

    3.2.1.3.  Convenience effect

    It is difficult to quantify the convenience effect of choosing HT crops. However, there are some clear advantages. For example:
     

    • The ease of the glyphosate-herbicide use and the large time window for spraying, which increases flexibility.
    • HT crops make the adoption of no-till or conservation tillage easier. According to Monsanto, in 1997, nearly half of the acres planted in RR soybeans are not tilled anymore. The absence or limitation of tillage implies lower use of crop-specific resources (labour, fuel etc). It is also considered to be more environmental friendly, in particular as it reduces soil erosion.


    Indeed, in a survey by Duffy and Ernst, 12% of the farmers listed increased planting-flexibility as a reason for going for GM soybeans.

    3.2.1.4.  Increased seed price

    Because of the "technology fee", seed for GM soybeans is more expensive than conventional seed. The Duffy and Ernst study showed a seed cost of 57 €/ha for GMO soybeans, versus 42 €/ha for non-GMO soybeans. This difference corresponds to the technology fee of 15 €/ha reported by Carpenter & Gianessi (1998). Other sources report somewhat lower figures, but still in the same order of magnitude 13.5 €/ha (University of Illinois, 1999) and 14 €/ha (Furman Selz, 1999). This means that, in average, GM soybeans seeds are 35% more expensive than conventional seeds.

    3.2.1.5  No significant profitability effect?

    At this stage, there are two counterbalancing elements in the growing of GM soybeans. On the one hand, seed prices Of GM crops are higher while yields (and thus, in a hypothesis of the same price for both variants, income) are lower, on the other hand, input costs are lower as well.

    The Iowa survey (Duffy et al., 1999) showed that differences in costs and yields between GM and non-GM varieties do not result in significant differences in return on land and on labour (at price 5.27 US $/bu =172.9 €/t).

However, if HT soybeans allow for savings in labour through their convenience effect, the same return for less labour means an increase in income per working hour.
Table 3.3 Comparison of returns for GM and conventional soybeans
Crop
Yield (t/ha)
Seed Cost (€/ha)
Total cost (excluding land/labour) (€/ha)
Return on land/labour (€/ha)
GMO
3.295
57
254
320
Non-GMO
3.430
42
274
322
The costs (total costs excluding land and labour) for non-GM in the Duffy study are reported to be 8% higher than for GM crops. However, these higher costs are offset by the higher yields.

Similarly, in simulations of the University of Illinois, the variable costs/acre for non-GM crops were estimated to be 6 to 8% higher than for GM crops. However, the assumption of no-yield drag made in this study seems not to hold, taken into account the results of different studies as indicated above.

The USDA has found no evidence of a significant change in variable profits in 1997, following the dramatic increase in GM soybeans sowings.

Before drawing definitive conclusions, the comparison of profitability between herbicide-tolerant and conventional soybeans systems deserves further analysis, in particular on the following elements:

  • Efficiency of different weed control systems: prices, herbicides quantities, management constraints versus convenience.
  • Will the yield drag close following the insertion of herbicide tolerant genes into top yielding varieties?
  • Are there divergent price developments between GM and non-GM soybeans?
    • 3.2.2.  Bt corn

      Profitability studies are mainly available for Bt-corn, which is the leading GM corn and has been grown on a wide basis for two or three years.

      3.2.2.1.  Evidence on yield gains

      By their stalk tunnelling action, corn borers are significantly damaging to corn crops. During one growing season, up to three generations of corn borers can affect a given crop. To be effective, insecticide applications have to be carried out at the appropriate stage of development. Hence they require scouting, or in other words, farmers have to follow developments regarding population and to make their applications decision on this basis, in due time. For cost/effectiveness reasons, uses of insecticide sprays against corn borers vary greatly from one production region to another, or even, from one grower to another.

      A soil bacterium, the Bacillus thuringiensis (Bt) produces toxins that kill the European Corn Borer. Bt corn includes gene material from the Bt bacteria, which allows own production of insecticide during the growth stage of the plant. Hence it is expected to have a yield and convenience advantage against non-Bt corn. A survey carried out in Iowa has shown that 80% of Bt-corn growers had chosen this option because of the expected yield advantage (Duffy, 1999).

      Several studies have found evidence on yield gains for Bt corn. Based on 1996-1998 data of the Agricultural Resources Management Data, the USDA has observed that adopters of Bt corn had obtained higher yields than non-adopters. This might however been partly explained by performance differences between these two groups of farmers. Gianessi and Carpenter (1999) report about average gains of 0.73 t/ha in 1997 and 0.26 t/ha in 1998, respectively, + 9% and +3% compared to 97/98 average yield for corn.

      The gap between 1997 and 1998 results can be explained by the difference in weather conditions and in insect pressure. Infestation was low in 1998. Other studies (like Alexander and Goodhue, Hyde and al., 1999) show the sensitivity of Bt performance to these two factors.

      3.2.2.2.  No clear savings in input costs

      According to an USDA case-study, insecticide treatments are significantly lower for Bt corn than for conventional corn. Globally, insecticide use for corn was lower in 1998 than in previous years. However, as previously mentioned 1998 had been a low infestation year. It is difficult to assess the role of Bt-technology in this reduction.

      Other studies (Rice, 1999) give more details on farmers' practices: an increasing percentage of farmers (13% in 1996, 26% in 1998) having adopted Bt corn indicate that they use less insecticide. Insecticides were not used at all by 50% of farmers. However, it is not clear whether the absence of applications results from Bt technology or if it was already the case with conventional varieties. Some farmers still spray insecticide on Bt corn, because its performance against second or third generation infestation is more limited. In addition, insecticide may still be needed against other pests.

      Considering that most of the farmers do not apply insecticide for controlling ECB, Furman Selz (1998) conclude that the value of Bt corn is not insecticide cost savings, but rather yield protection.

      The net effect regarding insecticide use and price is not clear-cut. Based on the 1998 Iowa survey14, Duffy (1999) reports reduced applications but increased insecticide costs: "Farmers applied insecticides in 12% of their Bt corn fields at an average cost of 17US $/acre. They applied insecticides to 18% of their non-Bt corn fields at an average cost of 15US $/acre". In this case, the advantage of Bt corn is not significant.

      In addition, Duffy observed that Bt fields required slightly higher weed control (+ 6 €/ha) and fertiliser (+11 €/ha) costs.

      3.2.2.3.  Refuges imply two-tiers crop management

      To prevent resistance in ECB populations, farmers planting Bt crops have been advised to keep "refuges" with non-Bt crops next to the Bt-fields. In early 2000, the US Environmental Protection Agency specified requirements which have to be observed in this respect. Refuges should cover at least 20% of the area planted in Bt corn. Where Bt corn is grown near Bt cotton, refuges have to cover an area equivalent to 50% of the Bt area. This should translate into increased cultivation constraints.

      It has been argued that resistance to Bt could raise problems for organic farming, which traditionally uses sprays or granulates of Bt preparations within pest control programmes.

      Furthermore, since findings on sensitivity of the Monarch Butterfly to Bt toxin have been published and debated, the effect of this toxin on insects other than corn borers has become an issue.

      3.2.2.4.   Increased seed price

      GM seeds are more expensive than conventional ones. This reflects both the technological fee charged by some biotech firms and the fact that GM and conventional seeds are sold on different markets. Alexander and Goodhue (1999) report on GM seed premiums for 20 GM corn varieties ranking from 3 €/ha for high yield varieties to 35 €/ha for some Bt varieties. The figure of 22 €/ha can be found in the Furman Selz paper (1998) as well as in the Gianessi and Carpenter publication.

      3.2.2.5.  Contrasted results on profitability

      As explained by Hyde et al (1999), the profitability of Insect resistant crop will depend on whether the "value of the protection" is less or more than the highest seed price. Results obtained by this research team for Indiana suggest that this is generally not the case. However, results depend on the level of infestation. Hyde and al have found that "when the probability of infestation increases from 25 to 40%, Bt corn value increases by about 69%". Therefore, Hyde considers that in areas where infestation is more likely or where average yields are higher, Bt corn should be profitable.

      Several other studies show that profitability of Bt is higher where infestation is high. The calculations carried out by Furman Selz (1998) are summarised in the following table.

    Table 3.4 Farmers economics for Bt corn, various infestation scenarios

    Source: Furman and Selz
    Compared with other studies, Furman and Selz calculations on income gain appear over-estimated, in particular, the relative high yield gains under the medium and heavy infestation scenario.

    Different results are outlined in table 3.5 Gianessi and Carpenter have assessed net gains/losses for the years 1997 and 1998. They have assumed that there was no cost-saving effect for lower insecticide applications. Results obtained by Duffy for Iowa are also summarised in the table.

    Table 3.5 Net gains and losses for Bt corn
    The results of the two studies are not directly comparable. As already mentioned, Duffy has estimated the insecticide, weed and fertiliser effects, while Gianessi and Carpenter have not.
    In the Gianessi and Carpenter study, the combined effect of lower yield gain and corn prices in 1998 resulted in net losses for Bt-corn growers. These first results show that profitability of Bt corn is highly dependent on the extent of yield gains and on prevailing market prices for corn. This also explains the gap in the results of different types of calculations.

    Taking into account differences in variable costs, Duffy concludes that there have been no cost savings. However, as a result of yield gains, Bt-corn has been slightly more profitable than conventional corn. Duffy nevertheless considers that the 9 €/ha gain is not significant.

    The cost of GM seeds is also a key factor in the relative profitability of GM crops. Alexander and Goodhue have examined the relationship between seed price and profitability, as well as the likely breakdown of profitability between firms and farmers for various types of GM corn in Iowa. They found that the ranking of net revenue performance matched the ranking of seed costs. Under their simulations, Bt corn appears to be the type of GM corn most likely to allow profits for farmers. A possible factor of explanation might be the number of Bt Corn types on the market (7 transformation events have been authorised in the US). There is a competition between these types of Bt, which are later incorporated into various hybrids. Hence, the authors consider that the likelihood of monopolistic pricing of the technology appears more limited.

    However, as explained in chapter 2, biotech companies are considered to form an oligopoly on the input-side of the farm sector, furthermore after having acquired seed companies or concluded agreements with them. Their margin of manoeuvre as far as prices of GM seeds and associated agro-chemical products are concerned is a key factor in the breakdown of profitability of GM crops. Farm-gate profitability of GM crops is very sensitive to input prices.

      To quote again Alexander and Goodhue, "analysis provides suggestive rather than conclusive evidence". There is evidence on yield gains of Bt corn, compared to conventional varieties, which are exposed to corn borers. The extent of the gain and hence, the cost-effectiveness of Bt technology, depends on the degree of infestation. The decision to plant Bt corn or conventional is a complex one, as it has to take into account the likelihood of infestation and various adjustments in crop management.

      3.2.3.  Herbicide Tolerant Canola

      Canola is a type of rapeseed which has been developed and is grown in Canada. It is a registered trademark, corresponding to specified low contents in erucic acid in oil and in glucosinolates in meals. It has initially been obtained through conventional breeding, but in recent years, GM herbicide tolerant varieties have been developed.

      The importance of Canola has increased drastically: barely grown twenty years ago, it became the third most important crop in Canada in 1994, its value representing 29% of all grains and oilseed receipts (Agricultural Institute of Canada, 1999). Canola production in Canada is mainly limited to the provinces of Alberta, Manitoba and Saskatchewan. These three provinces produce more then 98% of the Canadian Canola output.

      The production of GM Canola has risen spectacularly over the last years: In 1996, it represented only 4% of the output, in 1999 it was estimated by Fulton & Keyowski at 69%.

      3.2.3.1.  Contrasted results on yields

      Canola yields have gone up throughout the 1980s and 1990s, for example, in the province of Ontario yield has doubled between 1983 and 1996.

      Yield data comparing herbicide tolerant (GM) Canola to conventional Canola does not prove to be convergent. Estimations in Alberta vary between 15% lower to 15% higher yields for GM crops than for conventional crops, depending on region and variety. Manitoba figures show higher yields (up to 15%) in most cases.

      3.2.3.2.  A convenience effect

      Typically, the production of Canola requires two herbicide applications: one pre-emergent and the other post-emergent, the latter controlling only for a limited spectrum of weeds. The characteristic of herbicide resistance offered by GM Canola therefore improves potential in two ways:
       

      • removing competition for moisture and nutrients between Canola and weed.
      • eliminating costs for additional machine movements over the field (Fulton & Keyowski, 1999).


      3.2.3.3. Unclear results on costs and profitability

      Comparing costs and margins of conventional and GM canola is not a straightforward exercise. Based on 1998 accountancy data, the production economics and statistics branch of Alberta Province carried out a comparison between different Canola varieties grown on two types of soil, black and brown ones. There are two species of conventional Canola with different agronomic characteristics: "Argentine" Canola provides good performance under frost-free conditions, while "Polish" Canola is more resistant to frost and drought, but more vulnerable to diseases. The yield of Polish Canola is generally lower than for Argentine Canola. The result of the comparison between these varieties and GM ones on the two types of soils are outlined in table 3.6.

    Table 3.6 Costs and returns of different Canola varieties (in €/ha)

    Source: Alberta Simulations, 1998

    This table illustrates the difficulty of comparing profitability of these varieties, due to the number of factors which might have an effect.

    • Black soil areas allow for higher yields.
    • In the dark brown zone, there is no significant difference in yields and in total costs between the Argentine and the HT variety. Differences appear on the receipt side, lower sale receipts for HT crops, but higher insurance revenue.
    • In the black soil area, there are significant differences in yields, costs and receipts. The "Argentine" variety achieves the highest yield, with the HT variety coming close to that level. Total costs for HT Canola are lower than for conventional varieties, and this is mainly the result of reduced capital costs. However, due to higher receipts, Argentine Canola turns out to be more profitable.
    Although variable costs of HT and conventional Canola are broadly equivalent, even from one zone to another, the breakdown is different. While costs are higher for GM seeds, those for fertiliser and herbicides are lower for HT than for "Argentine" Canola. The convenience effect of HT Canola is reflected in the lower labour and fuel costs.

    As illustrated in table 3.7, the Fulton & Keyowski study seems a bit more prudent and stresses the fact that whether or not it is economically advisable to grow GM Canola varies from farm to farm. This points to a possible source of bias in the Alberta study: the average size of plots sown in HT Canola is higher than for conventional varieties. The Fulton & Keyowski assumption that HT Canola has lower costs and lower yields than conventional varieties appears to be confirmed by the Alberta data.

    Table 3.7 Conventional and GM canola production systems
    Canola product line: a system comparison, 1999
     
    Roundup Ready
    Smart Open Pol
    Liberty Hybrid
    Conventional Open Pol
    Costs        
    Total system costs
    84
    97
    102
    94
    Of which: Seed €/ha
    40
    40
    53
    29
    Herbicide
    11
    57
    49
    65
    Yield (bu/ha)
    82
    78
    88
    88
    Commodity price
    18
    18
    18
    18
    Expected Gross
    571
    545
    618
    618
    Less System Costs
    -84
    -83
    -102
    -95
    Gross Return
    487
    462
    515
    523
          Source: Fulton & Keyowski - the producer benefits of herbicide resistant canola
      There again, no clear-cut conclusion regarding the effects of the use GM canola can be drawn. There is only limited availability of data and all simulations start from different premises. Results depend on varieties compared, on growing and marketing conditions. However, the rapid adoption of GM Canola indicates that the variety is very attractive to the farmer.
    3.3.    Mixed outcome, many factors, longer-term assessment needed
    The results of different studies on profitability of the main GM crops can be summarised as follows:
    • Herbicide Tolerant soybeans allow for cost savings thanks to reduced use and cost of herbicides. This could offset the higher seed price. However, the yield of GM soybeans is still lower than for conventional varieties. When comparing returns per ha or per labour unit, no significant difference appears between the two types of crop. In this context, the convenience effect of HT crops appears to be the main driving force.
    • For Bt-corn, significant yield gains have been observed. However the cost-effectiveness of Bt corn depends on growing conditions, in particular on the degree of infestation in corn borers. Applications of insecticide have decreased globally. Some studies show increased total costs for Bt-technology, first for seeds but also for weed control and fertiliser. Results regarding profitability are contrasted, none can be considered as significant.
    • There are no clear-cut results for comparing the profitability of Herbicide Tolerant Canola with non-GM crops.
    These rather contrasted and unclear results indicate that short-term profitability is not the only driving force for the adoption of GM crops by farmers.

    Other factors have played a significant role in the rapid extension of GM sowings.

    The convenience effect seems to be a significant advantage, in particular for herbicide tolerant crops. This benefit does not directly translate in terms of profitability, but rather in terms of attractiveness of GM crops for efficiency purposes. This convenience effect has to be further assessed in particular, the valuation of the labour effect. In the longer run, it should imply increased labour productivity and savings in crop-specific labour costs. Further efficiency assessments, including price and use of herbicide over a longer time frame, would also be useful.

    The profile of adopters of the new technology also plays a role. First adopters were mainly young, educated and well-performing farmers, established on large holdings. The adoption of biotech crops is not size neutral. The higher than average farm size of adopters might be a factor explaining, amongst others, the dramatic increase in areas sown to GM crops. Theoretically, more benefits are accruing to early starters. Those already having adopted the technology are likely to have gained from it. In the case of HT crops, gains in efficiency should translate into improved labour productivity. In the case of Bt corn, yield gains mean enhanced productivity of land. Under given prices, enhanced productivity leads to an increase in supply. While more and more producers are adopting biotech crops, thus contributing to the increase in supply, on the demand side, concerns about GM food are emerging. This may lead to a drop in prices. Hence, gains for late adopters are expected to be lower than for early adopters. In the long run, enhanced productivity will have an impact on farm restructuring, alongside with other factors playing a role in this process.

    The reviewed studies only compare farm-level and short-term profitability. Profitability of GM crops should be analysed over a longer timeframe. First, there are important yearly fluctuations in yields and prices, and it is difficult to isolate the possible effect of biotechnology. Results are very sensitive to the price of seeds and agro-chemical products on the one hand and to commodity prices on the other hand. In most profitability studies, prices for GM and conventional crops are assumed to be equivalent.

    Developments on the supply and on the demand side of the food chain have to be considered together, and this is another reason for assessing profitability over several years. As a result of consumer concerns and preferences, segregation between GM and non-GM crops is developing, which implies differentiation in costs and prices. The economic implications of segregation and identity preservation are analysed in chapter 5. They might change the outset as regards profitability of GM versus non-GM crops.

    Policy measures, in particular the eligibility of GM crops under various support schemes, have reduced the price risk of the new technology. Until recently, no significant differences in prices between GM and non-GM crops have been systematically recorded, expect on niche markets. Hence, growers have mainly based their planting decisions on expected farm-gate performance, on cost-efficiency of inputs. In other words, they had an input-oriented approach.

    The marketing strategy developed by biotech firms must also be considered. It has been focused on farmers, the first customers interested in input traits. In the case of herbicide tolerant crops, the marketing strategy was based on the concept of "technological package" (the GM seed and the product to which it is resistant), which allows for "combined pricing". Benefits of GM crops have been extensively advertised throughout key production areas (Corn Belt). Biotech firms have been present up to the field, providing commercial and technical assistance to farmers, whether directly or through their subsidiaries. They have shaped famers'expectations.

    The supply-oriented approach of both biotech companies and farmers has been quickly confronted with reactions stemming from the downstream side of the food chain. Consumer concerns have been echoed and amplified by NGOs and retailers, and they had a cascading effect on the upstream side. These reactions are analysed in the next chapters.

    [Index] - [Previous chapter] - [Next chapter]

    Top

    [ Agriculture ] - [ Publications

    What's New ?
    Key Issues

    Index

    Mail-Box

    Search

    Information

    Feedback