Volume 12 // Number 3 & 4 // Article 9
PDF Comment on this article Issue contents Previous article Next article
Corporate Pricing Strategies with Heterogeneous Adopters: The Case of Herbicide-Resistant Sugar Beet
Katholieke Universiteit Leuven
Africa Rice Center (AfricaRice) and Katholieke Universiteit Leuven
Katholieke Universiteit Leuven
In ex-ante impact assessment of proprietary seed technologies, the assessor operates under scarce and imperfect data. No market has been established for the new technology and adoption has yet to take place. Recently, the scholarly literature has focused on the importance of accounting for heterogeneity among potential adopters to avoid homogeneity bias in the impact estimates. In this article, we argue that incorporation of heterogeneity in the corporate pricing strategy of the innovation is also needed to avoid a second bias in the welfare estimates—pricing bias. Therefore, a framework is developed which explicitly incorporates heterogeneity of proprietary seed technology valuation among adopters in both the pricing decision and the impact assessment. The results explain the tendency of innovators to engage in third-degree price discrimination if the market structure discourages arbitrage. Finally, the model is applied on the case study of herbicide-resistant sugar beet in the EU-27.
Key words: ex-ante impact assessment, GM crops, herbicide resistance, heterogeneity, parametric modeling, monopolistic pricing.
Introduction Since the commercial introduction of the first generation of genetically modified (GM) crops in agriculture, both the value creation and benefit sharing of these technologies have been of great interest. In contrast to earlier, publicly funded technologies in agriculture, most of the commercially available GM technologies are developed and commercialized by the private sector. The laws and enforcement of intellectual property rights (IPRs) have provided innovating firms with some monopoly power in the market for GM seeds, affecting the value-creation and benefit-sharing of these proprietary seed technologies (Acquaye & Traxler, 2005; Falck-Zepeda, Traxler, & Nelson, 2000; Moschini & Lapan, 1997; Moschini, Lapan, & Sobolevsky, 2000). The first generation of GM crops were introduced more than a decade ago and ex-post impact studies have uncovered the global value creation (Brookes & Barfoot, 2009) and benefit sharing (Demont, Dillen, Mathijs, & Tollens, 2007) of these technologies. Regardless of the variability of the impact evidence, Demont et al. (2007) found that, on average, two-thirds of the global benefits of first-generation GM technologies are shared among domestic and foreign farmers and consumers, while only one-third is extracted by the input suppliers (developers and seed suppliers).

Previous research has shown that the heterogeneous character of farmer populations (Weaver, 2004) results in a downward-sloping aggregate derived-demand curve for GM seed technologies (Falck-Zepeda et al., 2000). This suggests that, despite their high value, GM seed innovations can be considered non-drastic (Arrow, 1962) because the monopolist’s pricing decision is constrained by the threat of competition (Lemarié & Marette, 2003), leading to restricted monopoly pricing (Weaver & Wesseler, 2004) and reduced adoption. In this article, we argue that the observed benefit-sharing of first-generation GM technologies is a direct reflection of heterogeneity of farmers’ technology valuation, constraining pricing strategies of monopolistic technology providers. This is in contrast with earlier research stating that the pricing of biotechnology innovations has a clear strategic dimension (Fulton & Giannakas, 2001).

Oehmke and Wolf (2004) developed a model to assess to what extent observable farmer rents can be explained by adopter heterogeneity in the US Bt cotton market. Although heterogeneity is implicitly captured by ex-post adoption data, Demont et al. (2008a) show that ex-ante impact assessments need to explicitly account for it in order to avoid homogeneity bias. However, their impact estimates and the model application by Demont and Dillen (2008) are based on fixed, exogenous, technology pricing. In this article, we elaborate on their framework and relax the assumption of exogeneity by introducing alternative corporate pricing strategies. A parametric framework to fully incorporate farmer heterogeneity in ex-ante impact assessment of proprietary seed technologies under data scarcity is developed. Within this framework, alternative corporate pricing strategies are simulated to assess the effect of heterogeneity on classic adoption parameters such as profits and adoption. The output of this framework can be aggregated through a trade model—as has been done by Dillen, Demont, and Tollens (2009) in an earlier issue of this journal—to assess the effect of pricing strategies on total welfare distribution. The approach offers both a tool for policy makers to implement ex-ante socio-economic assessments with limited resources and for seed technology and gene developers to assess the value of proprietary seed technologies.

The article is organized as follows. In the next section, a theoretical framework for modeling heterogeneity among potential adopters of a new technology protected by IPRs is developed and some comparative statics are derived. In the third and fourth sections, an empirical application on herbicide-resistant (HR) sugar beet in the EU-27 is presented. A final section discusses the theoretical and empirical results of our framework.


We define a technology as a marketable good that allows farmers to surmount an agricultural constraint. Moreover, we introduce the concept of technology valuation to represent the willingness to pay (WTP) for the technology, which can include both pecuniary and non-pecuniary attributes (e.g., see Marra & Piggott, 2006). While some articles assess both the irreversible and reversible benefits and costs of the technology to determine the maximum incremental social tolerable irreversible costs (MISTIC) (Demont, Wesseler, & Tollens, 2004; Wesseler, Scatasta, & Nillesen, 2007), this article focuses on the reversible private benefits and costs, as these drive adoption among profit-maximizing farmers. It is important to recognize the fact that the innovation does not happen in a vacuum. Certain innovations alter the nature of production decisions, such as decisions about individual input (e.g., seeds and pesticides) decisions over bundles of complementary inputs (e.g., HR seed complimented with a broad-spectrum herbicide) (Alexander & Goodhue, 2002). Therefore, the value of such technology should be calculated as the value of the new production system as a whole. Previous research showed that the value of a technology is not uniformly distributed among farmers; some realize a profit from the technology and adopt it while others rationally choose not to adopt. In particular, GM seed technologies will pay off differentially depending on field conditions, pest densities, crop rotation and environmental conditions. Moreover, the technology valuation to any particular farm will depend on managerial expertise and local market conditions that condition the profitability of GM seed technologies relative to alternative technologies (Weaver, 2004). Furthermore, technology valuation is affected by the attitude towards risk of the potential adopter. Empirical evidence shows that most farmers are risk averse (e.g., Anderson & Hardaker, 2003). Farmers are averse to being exposed to unexpectedly low returns, and this affects their technology valuation and adoption behavior.

In ex-post adoption data of a technology, technology valuation is revealed by adopters through their adoption decisions and technology expenditures. From an ex-ante perspective, adoption has not yet occurred and the revealed preference information of a technology is quasi-unobservable to the researcher. Stated-preference information can be collected through contingent valuation (CV) analysis. However, this method requires costly survey data, as surveys need to be reproduced in different years and different regions in order to capture both the structural and stochastic foundations of heterogeneity. Moreover, farmer preferences are elicited directly based on hypothetical, rather than actual, scenarios. These constraints severely limit the use of CV analysis in large-scale, ex-ante impact assessment under time and resource constraints.

Therefore, we propose a framework that explicitly models heterogeneity of proprietary seed technology valuation among adopters under data scarcity. Just, Hueth, and Schmitz (2004) recommend the use of a probability density function (PDF) to model heterogeneity among producers. We adapt this modeling approach to the case of heterogeneous technology adopters. In ex-ante impact assessment, imperfect information is endogenous to the problem and parametric approaches can be used to complete scarce data with estimates based on assumptions, analogy, and theory.

Let x represent farmers’ valuation or WTP for a new proprietary seed technology and assume that the new technology is an innovation such that x > 0. Farmers’ technology valuation may include pecuniary benefits such as yield increases and cost reductions, as well as non-pecuniary benefits such as convenience and enhanced flexibility of farming operations induced by the technology (Marra & Piggott, 2006). Hence, it summarizes the total value of all attributes of the technology for which the farmer is willing to pay. Let f(x) represent the PDF of individual technology valuations in a heterogeneous population of farmers. From the ascending cumulative density function (CDF) of technology valuations, F(x), the descending CDF, Q(x), can be derived, which can be interpreted as a normalized demand curve, Q(x)∈[0,1], for the new technology:

Q(x) = 1 – F(x). (1)

The restricted monopoly in the market for privately developed proprietary seed technologies allows the innovator to set a monopolistic price, which is higher than the price of the conventional technology bundle. In most ex-ante impact studies, this technology fee (θ) is treated exogenously (Demont et al., 2008a, 2008b; Demont & Dillen, 2008; Demont & Tollens, 2004; Flannery, Thorne, Kelly, & Mullins, 2004; Hareau, Mills, & Norton, 2006; May, 2003). Alston, Hyde, Marra, and Mitchell (2002) endogenize θ based on first-order statistics, i.e., they assume that the technology would be competitively priced, which implies that the technology fee is set at the mean technology valuation. However, neglecting higher-order statistics does not account for heterogeneity and leads to homogeneity bias in the estimation of the impact of the technology on welfare and corporate revenue (Demont et al., 2008a). Throughout this article we assume that farmers act rationally, adopting a technology if valuation, net from technology price, is positive, i.e., xθ > 0. Assuming constant long-run marginal costs, c, the profit function of the monopolistic innovator is represented by:

π (θ) = (θc)Q(θ). (2)

The optimal price of the technology bundle, θ*, satisfies the following first-order condition:

[(θ) / ] = [dQ(θ) / ].(θc) + Q(θ) = 0. (3)

Alexander and Goodhue (2002) argue that if users are heterogeneous, q* may be below the technology valuation of a potential adopter, therefore leaving significant rents with the adopter. Lapan and Moschini (2000) alternatively interpret this pricing strategy as the monopolist choosing the profit-maximizing marginal adopter, m = q*, directly from a heterogeneous population of farmers and allowing the adoption to be incomplete. This marginal adopter is indifferent between adopting and a status quo. Therefore, all potential adopters with a higher technology valuation will adopt the new technology, leading to the adoption rate

ρ = f(x)dx.
        θ* (4)

We define fa(x) as the adopters’ density function of technology valuation:


The average net value (farmer surplus) of the new technology, , for all adopters can be measured by aggregating the farmer surplus, τ(x) = xθ*, and amounts to

= τ(x). fa(x)dx.

Similar to Oehmke and Wolf (2004), we derive some comparative statistics. Assume f(x) is characterized by a mean µ, standard deviation σ, and risk premium R. In order to derive comparative statistics analytically, independence between µ and σ is required. For R not to thwart this assumption, constant absolute risk aversion is assumed. Constant absolute risk aversion is often used to analyze farm decisions under risk (see Mitchell, Gray, & Steffey, 2004, for an application in GM crop literature). Assuming an increasing von Neumann-Morgenstern utility function (Babcock, Choi, & Feinerman, 1993), we can calculate the risk premium R as

R(A,σ) = [ln (0.5(e-Aσ + e)] / , (7)

where A represents the coefficient of absolute risk aversion.

The effect of a change in mean technology valuation over all farmers on θ* is determined by

(* / ) = [−Fμ(x) + (c−θ) fμ(x)] / [−F(x)θf(x) − θ fθ (x)], (8)

where subscripts denote partial differentiation. In order to determine the sign of the derivative, additional assumptions are needed. The denominator is negative under the assumption that f(x) is unimodal and θ* lies on the lower tail of the PDF. As we assume the innovation will be marketed, θ* > c and the denominator becomes positive. Therefore, as the mean technology valuation increases, the technology fee will follow.

The effect of a change in variance is determined by

(* / ) = [−F(x)σ + (c−θ) f(x)σ] / [−F(x)θf(x) − dF (x)θ]. (9)

Under the same assumptions, this derivative will be negative. This means the innovator will drop the price in order to maintain his customer base as the variance or heterogeneity increases. This price decrease originates both from the direct effect of a more heterogeneous population and the decrease of the risk premium, which acts as a shifter as [dR(A,σ) / ] < 0 for σ > 0. The effect of increased risk aversion is determined in a similar way by

(* / dA) = [−F(x)A + (c−θ) f(x)A] / [−F(x)θf(x) − dF (x)θ], (10)

where under the same assumptions, θ* decreases if absolute risk aversion increases as [dR(A,σ) / dA] < 0 for A > 0. The induced changes in θ* translate to changes in the welfare effects of the technology. However, the outcome of these parameters depends on the specific shape of the PDF as the general form used before leads to ambiguous results. Therefore, we discuss the results in the next section through an empirical case study.


The case of herbicide resistant (HR) sugar beet is very appealing for EU agriculture as this crop is grown in most EU countries and economic sugar production is impossible without weed control. Moreover, the recent reform of the sugar regime (see Dillen, Demont, & Tollens, 2008) towards more market-driven and a booming market for raw materials (biofuels and bio-based chemistry) increases the need for cost reduction in the sector. We understood that the major impediment comes from the concentrated group of refiners, processors, and manufacturers of sugar and sugar-containing products. Processors face risks related to market acceptability of sugar and by-products (DeVuyst & Wachenheim, 2005). However, it seems that sugar processors recently opened their doors to biotechnology following the food and feed approval in Australia, New Zealand, Japan, and the European Union. The success of the commercial introduction of HR sugar beet in the United States in 2008—reaching an unprecedented adoption level of nearly 100% of the total sugar beet area in the second year (Kilburn, 2009)—proves the high value of the technology for farmers.

The appropriate PDF of the technology valuation among potential adopters is the centerpiece of the developed model. Oehmke and Wolf (2004), in their ex-post assessment, construct the PDF through a kernel density estimator on available adoption data. In ex-ante assessment, the assessor faces the endogenous problem of imperfect information that makes a similar approach impracticable. The amount of data available to the researcher can be situated in an information continuum. At the lower end of the continuum, no information is available and theoretical considerations dictate the parametric estimation procedures. Towards the higher end of the information continuum, in the direction of full information, empirical findings gradually replace theory and parametric procedures substitute for non-parametric procedures. Although the probability of selecting the correct parametric model is zero, a parametric approach does not necessarily lead to biased results. With a false model, parameters will converge such that the Kullback-Leibler distance between the true density and the best parametric estimate is minimized. Therefore, an incorrect parametric model may have greater efficiency than the correctly modeled density and a non-parametric model (Goodwin & Ker, 2002). In the case of data scarcity, expert opinions are often used to construct subjective probabilities for uncertain parameters. Although very popular, the triangular PDF might not be the best way to model expert opinions for modeling heterogeneity of technology valuation. Both tails are overemphasized and experts often have better knowledge on the central tendencies than on the extremes. Therefore, the PERT PDF, a special case of the Beta distribution, is preferred for modeling expert opinions. It can range from highly skewed to symmetrical distributions, has a close fit to normal and lognormal distributions, and attributes less weight to the extremes. The excessive reliance of the triangular PDF on the extremes also influences the variance introduced into the model. In Figure 1 we observe that for the same values of the extremes, the triangular distribution always has a higher standard deviation, therefore introducing unnecessary uncertainty in the model if we assume that we can rely on the expert’s knowledge. Demont et al. (2008a) use a re-parameterized version of the classic PERT specification to construct a PDF based on their specific context of limited data availability. In this article, we are one step further in the information continuum as we combine expert opinions with evidence from farmer surveys, which allows us to select the PDF based on empirical goodness-of-fit instead of forcing a pre-defined PDF to fit our expert data.

Figure 1. Effect of parameter specification on the standard deviation of the triangular and the PERT PDF.

For economic sugar-beet production, effective weed control is crucial. Yield losses can be up to 100%; such is the poor ability of beets to compete with the large range of weeds present in arable soils (Dewar, May, & Pidgeon, 2000). Expert opinions on the herbicide expenditures for most of the EU countries are reported by Hermann (1996; 1997; 2006), while detailed cross-sectional survey evidence is available for the Netherlands (Institute of Sugar Beet Research [IRS], 2004) and France (ITB, 2000; Lemarié et al., 2001). We use the first data source to obtain basic first- and second-order moments of the underlying data, while we use the second data source to determine higher-order moments in order to determine the appropriate functional form to be used in our parametric modeling. We fit alternative CDFs on the available cross-sectional datasets on herbicide expenditures in sugar-beet growing. The loglogistic CDF consistently provides the best fit based on the root mean square error (RMSE) criterion. It is therefore chosen to model heterogeneity among potential adopters (Table 1). We combine the empirical results with the expert opinions through the estimation of the three fractiles in the survey data corresponding with the minimum, maximum, and mean expert opinions. By assuming that the correspondence between both data sources is constant, we calibrate loglogistic CDFs of herbicide expenditures on expert opinions for major sugar-producing member states of the EU-27 in 2004 (Table 2). With this procedure, we avoid the difficulties of experts to estimate the extreme fractiles of a distribution (Miller, 1956). This yields the following PDF, h(z), and CDF, H(z), for our case study:

h(z) = (11)


H(z) = , (12)

where z represents herbicide expenditures, γ the scale parameter, and δ the shape parameter of the loglogistic PDF. In order to make the concept of technology valuation more dynamic, a shift parameter, φ > 0, is introduced. The parameter can be used to introduce constant absolute risk aversion premium, R(A, σ), as presented earlier, and various exogenous price and cost effects into our static model. For the assessment of a biotechnological innovation, the latter can include (i) the inclusion of coexistence measures (e.g., Devos et al., 2008), (ii) the existence of a market premium for non-GM crops, and (iii) changes in the pricing strategies of competing technologies. In the case of HR crops, history shows that conventional herbicide producers try to countervail the increased competition with price reductions. This effect took place with the introduction of HR soybeans in the United States. It can be observed with the introduction of generic products in the herbicide market as well (Just, 2006). Due to data limitations, we assume the cost implications of these price reactions to be homogeneous in the population and to shift the PDF without affecting relative heterogeneity measured by the ratio of the mean to the standard deviation as both statistics are multiplied by the same factor. Although spatial variation in some of these factors certainly exists, we assume that this heterogeneity is captured by the herbicide expenditures. In order to transform herbicide expenditures, z, to technology valuation, x, we subtract the total cost of the HR system, excluding the technology fee, from the total cost of the conventional system, including application and product costs. The country-specific PDF of technology valuation can then be calculated as

fi (x) = hi [x + (nci – ngi) ki – ], (13)

where ngi and nci represent the number of herbicide applications in the HR and conventional system respectively, ki the cost of a single herbicide application (assumed equal for glyphosate and conventional applications), gi the dosage, and pgli the price of glyphosate in country i. Note that (x + nciki) represents the total cost of the conventional system while (ngiki + pgligi) represents the total cost of the HR replacement system, excluding the technology fee. Again, due to data limitations we assume application costs and glyphosate expenditures to be homogeneous at the country level and to shift the PDF of technology valuation horizontally without affecting its shape.

Table 1. Cumulative loglogistic distributions fitted on survey data on weed expenditures.
Model: H(z) = 1/(1+(z/γ)^(-δ)) (Equation 12, assuming ϕ = 0 before adoption of HR sugar beet)
France, 1997 France, 2000 Netherlands, 2004
Parameter Scale (γ) Shape (δ) Scale (γ) Shape (δ) Scale (γ) Shape (δ)
Estimate 103.3* (1.0) 4.8* (0.2) 160.2*
0.99852 0.99981 0.99917
Goodness-of-fit (RMSE) 0.003 0.006 0.005
Sources: IRS (2004), Institut Technique de la Betterave (ITB; 2000), and Lemarié et al. (2001).
Notes: Standard deviations are shown between brackets. * = significant at the 0.01 level

Based on the resulting PDF of technology valuation among potential adopters, the innovating sector (technology developers and seed suppliers) will set its price in order to maximize profits following the presented modeling framework. We differentiate between two distinct corporate pricing strategies. In the first pricing strategy, we use disaggregated data to calculate the price at the member-state level. This pricing strategy is known as spatial third degree price discrimination (Schmalensee, 1981). The second pricing strategy assumes a uniform price setting over the EU-27 sugar-beet producing countries. The resulting proprietary seed demand function for the proprietary seed is constructed based on the area-weighted average herbicide expenditures:

Q(x) = fi(x).χi, (14)

where n represents the number of countries taken into account and χi the share in area over the n countries. The two different pricing strategies enable investigating the effect of heterogeneity on the variables under research. Uniform pricing is a realistic pricing strategy for the European sugar sector, as it is common for sugar producers (processors) to purchase seed and distribute it upstream with contracted farmers. This practice, combined with the high concentration of the sugar sector through multinational companies (Smith, 2007), creates a situation of oligopsony at supranational levels, which would favor a uniform pricing strategy.

With these specifications, the impact of the different pricing strategies on adoption and value creation can be calculated following the system of Equations 3-6.


Table 2 presents the data used in this study. Similarly to Desquilbet and Bullock (2009), we assume for simplicity that the long-run marginal costs of supplying seed are zero, which comes down to assuming that the technology developer maximizes revenue instead of profit. All prices are deflated to the 2008 level by using GDP deflators (World Bank, 2008). To aggregate the revenues to the national level and for yields, we use data from F.O. Licht (2005). The shift parameter, φ, is calibrated on the expected reduction of the price level of competing conventional herbicides. Analogous to the effect of HR soybeans in the United States, we estimate this reduction to take the value of 20% (Just, 2006). Moreover, analogously to other studies (Demont et al., 2008a; Demont & Dillen, 2008; Dillen et al., 2009), we include a 5% yield increase for adopting farmers due to reduced toxicity of the glyphosate herbicide regime in HR sugar-beet cultivation. The price of glyphosate is assumed uniform among member states (€4.37/l), first because of data limitations and second to abstract from markets where glyphosate is not well established. In such markets, the price structure is likely to change with the introduction of HR sugar beet. To construct the demand function in the uniform pricing strategy we use the area-weighted average of fi(x) for the countries producing three-quarters of the European production (Equation 14). These regions can be considered the main market. They include the efficient production regions where innovation incentives are positively affected by the recent sugar reform. They therefore are more likely to adopt HR sugar beet (Dillen et al., 2008).

Table 2. Weed control costs under conventional and HR sugar beet technology in the EU-27.
Conventional weed control HR weed control
Member state Min (€/ha) Mean (€/ha) Max (€/ha) Scale parameter γ Shape parameter  γ Application cost k (€/ha) nc Total application costs (€/ha) ng Glyphosate dose g (l/ha) Glyphosate expenditures pgl (€/ha) Total application costs (€/ha)
Austria 156 311 467 261 5.4 41.5 2.5 103.8 2.5 6 26.2 103.8
Belgium 104 261 417 207 4.2 18.5 3.5 64.8 2.5 6 26.2 46.3
Germany 76 206 334 160 3.9 18.2 3 54.6 2.5 6 26.2 45.5
Spain 141 261 381 223 6.1 13.0 3 39.0 1 3 13.1 13.0
Czech Republic 138 198 276 180 10.0 7.3a 3b 21.9 2.5 6 26.2 18.3
France 103 150 206 136 9.7 20.0 3.8 76.0 2.5 6 26.2 50.0
Finland 154 220 297 201 10.0 16.4 3.8 62.3 2.5 6 26.2 41.0
Greece 94 132 202 121 10.5 17.5 1.5 26.5 1 3 13.1 17.5
Italy 95 169 253 145 6.4 15.3 2.5 38.3 2.5 6 26.2 38.3
Ireland 64 93 122 84 9.7 13.3 3 39.9 2.5 6 26.2 33.3
Netherlands 135 176 238 164 13.5 40.4 3.5 141.4 2.5 6 26.2 101.0
Poland 121 214 332 185 6.4 7.3 3b 21.9 2.5 6 26.2 18.3
Sweden 77 186 308 149 4.3 13.3 2.9 38.6 2.5 6 26.2 33.3
United Kingdom 78 149 225 124 5.9 14.4 4.6 66.2 2.5 6 26.2 36.0
Denmark 88 212 372 166 4.4 26.0 4 104.0 2.5 6 26.2 65.0
Portugal 141c 261c 381c 223 6.1 13.0 3 41.7 1 3 13.1 13.9
Hungary 64d 159d 211d 132 2.7 7.3a 3b 21.9 2.5 6 26.6 18.3
Sources: Hermann (2006) for most cases, complemented by Hermann (1996; 1997) if recent data is missing, and Schäufele (2000).
Notes: nc = number of conventional herbicide applications; ng = number of glyphosate applications
a no data available; used data from Poland
b Urban, Pulkrabek, Valenta, Beckova, and Kviz (2008)
c no data for Portugal; used data from Spain
d no data from Hermann (1996; 1997; 2006), data from Research and Information Institute for Agricultural Economics (AKII, 2004)

Results and Discussion

In Table 3, the results of our pricing framework are presented. Under third-degree price discrimination, the technology fee ranges from €50/ha to €147/ha. The area-weighted average technology fee amounts to €98/ha. These spatially adapted technology fees result in adoption potentials (ceilings) that range from 51% for Ireland to 97% for Spain. The estimated adoption rates are remarkably close to potential market-penetration rates estimated by experts of national sugar-beet institutes (Coyette, Tencalla, Brants, & Fichet, 2002). The uniform pricing strategy yields a European-wide price of €95/ha. This fee, less adapted to local conditions and demand, generates a dramatically different adoption pattern, from almost no adoption in Ireland to almost full adoption in Spain. At first sight, the endogenous technology fees seem rather high compared to commercially available GM seeds of other crops. However, they are in line with the current technology fee of HR sugar beet in the United States, i.e., €90-106/ha (KWS, personal communication, 2006), and can be explained by the high herbicide expenditures in conventional sugar-beet cultivation (Table 2). The results confirm the theoretical model as the increased farmer heterogeneity of herbicide expenditures, generated by applying a uniform pricing strategy over a larger area, leads to a lower technology fee at the EU-27 level.

Table 3. Technology fee, innovator revenue, farmer rent, and total value of HR sugar beet under two alternative pricing strategies in the EU-27.
Price discrimination Uniform pricing
Member state Technology fee
Farmer incremental profits
Adoption (%) Corporate revenue (million €) Total value (€/ha) Technology fee
Farmer incremental profits (€/ha) Adoption (%) Corporate revenue (million €) Total value (€/ha)
Austria 147 123 74 4.9 270 95 188 84 3.8 283
Belgium 123 131 76 8.1 254 95 174 98 7.2 269
Germany 94 99 69 26.5 193 95 108 63 26.3 203
Spain 145 135 97 12.9 280 95 192 99 9.0 287
Czech Republic 98 87 88 5.6 185 95 101 91 5.6 196
France 87 133 90 27.0 220 95 131 78 26.9 226
Finland 125 68 79 3.5 193 95 102 97 2.7 197
Greece 75 95 90 2.2 170 95 63 43 1.8 158
Italy 77 59 79 9.7 136 95 45 43 8.6 140
Ireland 50 32 51 0.7 82 95 1 5 0.003 96
Netherlands 121 92 92 11.2 239 95 151 97 8.6 246
Poland 102 79 80 22.4 181 95 96 85 22.0 191
Sweden 86 74 65 2.5 160 95 71 47 2.5 166
United Kingdom 78 117 86 8.8 195 95 99 59 8.5 194
Denmark 106 122 80 4.3 228 95 145 82 4.0 240
Portugal 145 135 97 1.2 280 95 192 99 0.7 287
Hungary 108 63 84 2.6 171 95 78 92 2.4 173
EU-27 area-weighted average (total) 98 99 (154) 204 95 116 (141) 218
Note: For comparison, the market potentials estimated earlier by experts of national sugar beet institutes in six selected countries are 75% for Belgium, 71% for Germany, 100% for Spain, 90% for France, 93% for the Netherlands, and 89% for the United Kingdom (Coyette et al., 2002).

Under price discrimination, the area-weighted average of the farmer profits amount to €99/ha, with the extremes in Spain and Ireland. Corporate revenue in the EU-27 is €154 million. If third-degree price discrimination is not possible, e.g., due to the market structure, the farmer profits amount to €116/ha, while the corporate revenue drops to €141 million. Under increased farmer heterogeneity of herbicide expenditures, farmers are able to capture a higher rent while the innovator loses revenue. The lower technology fee will relocate the position of the marginal adopter, but the increased adoption rate is not enough to compensate for the reduced optimal technology fee on a per-hectare base. Therefore, the corporate goal of increasing adoption might not be the revenue-maximizing strategy. Under market structures with strong IPRs (which reduce arbitrage), third-degree price discrimination is a profitable strategy for the innovator. By dividing the population into smaller, more homogenous segments, corporate revenue increases by progressively extracting farmer surplus from the total value of the proprietary seed technology. Heterogeneity is minimized in each sub-market fi(x), and revenue is maximized. Empirical evidence can be found for Bt cotton in the United States, Mexico, and South Africa (Frisvold, Reeves, & Tronstad, 2006; Gouse, Pray, & Schimmelpfennig, 2004; Traxler, Godoy-Avila, Falck-Zepeda, & Espinoza-Arellano, 2003). The separation is mainly introduced by different germplasm due to the sensitive reaction of upland cotton varieties to agro-climatic changes (Acquaye & Traxler, 2005) and preventing farmers to buy seeds in other districts (Traxler et al., 2003). With strong IPRs, the submarkets could be introduced through enforcement of a contract with a “no resale” clause, thereby strengthening monopoly power. In Europe, price discrimination can be found in Bt maize in Spain as a function of spatial pest-infestation levels (Gomez-Barbero, Berbel, & Rodriguez-Cerezo, 2008). These observations indicate that the innovator is aware that patent-based uniform pricing of a GM seed technology leaves substantial benefits with the producers, preventing full appropriation by the innovator due to heterogeneity among farmers. In developing countries with weak or no governance of IPRs, on the other hand, price discrimination of HR technologies can only be efficiently implemented in the case of hybrids as hybridization biologically strengthens IPR protection and acts as a substitute for weak IPR enforcement (Goldsmith, Ramos, & Steiger, 2003). This may explain the absence of price discrimination in the case of HR soybeans in developing countries (Qaim & Traxler, 2005).

Due to the antagonistic response of farmer rents and corporate revenues to changes in heterogeneity, the magnitude of total welfare depends on the shape of the parametric model. In the case of the loglogistic PDF, the upwards effect on farmer rents following an increase in heterogeneity more than compensates the downwards effect on corporate revenue on a per-hectare basis such that the total value increases. The total value under uniform pricing amounts to €218/ha, while it only reaches €204/ha in the case of third-degree price discrimination. Full welfare effects under uniform pricing are calculated using a partial equilibrium model, EUWABSIM (Dillen et al. 2008; Dillen et al. 2009).

Demont et al. (2008a) demonstrate the homogeneity bias that emerges from not incorporating second-order statistics in ex-ante impact assessment. They conclude that the bias is a decreasing function of the mean and an increasing function of the variance of f(x). This result contrasts with earlier statements in the literature that the distribution is of minor importance as the lower and higher tails of the distribution will compensate each other (Breustedt, Muller-Scheessel, & Latacz-Lohmann, 2008). However, by using an exogenous technology fee, , they potentially introduce a second source of bias, i.e., pricing bias. In the literature on HR sugar beet, different assumptions and estimates have been reported for the technology fee, varying from €25/ha (Flannery et al., 2004), €30-40/ha (Märländer, 2005), €40/ha (Demont, 2006), €32-48/ha (May, 2003), €38/ha (Gianessi, Sankula, & Reigner, 2003), and €77/ha (Lemarié et al., 2001) in Europe to €128/ha (Gianessi, Silvers, Sankula, & Carpenter, 2002), €133/ha (Burgener, Feuz, & Wilson, 2000), €157/ha (Rice, Mesbah, & Miller, 2001), and €164/ha (Kniss, Wilson, Martin, Burgener, & Feuz, 2004) in the United States. We illustrate the magnitude of the exogenous pricing bias for France by comparing farmer surplus generated under the assumption of an exogenous technology fee of €40/ha with surplus generated under the assumption of third-degree price discrimination. The assumption of exogeneity would lead the researcher to overestimate the potential adoption ceiling by 11% (100% instead of 89%) and the value captured by farmers by 39% (€185/ha instead of €133/ha). If the technology developer commercializes HR sugar beet seed into the European sugar industry through uniform pricing, endogeneizing the technology fee but assuming third-degree price discrimination instead of uniform pricing would generate a second source of bias, i.e., farmer benefits would still be overestimated by 1.5% (€133/ha instead of €131/ha). The small bias is explained by the fact that France, the largest sugar producer, would play a determinant role in EU uniform pricing. The total pricing bias on farmer surplus would then be the sum of both biases and would amount to 41% (€185/ha instead of €131/ha).

Our results underline the importance of incorporating heterogeneity in ex-ante impact assessment of proprietary seed innovations, especially if the impact assessor is operating under data scarcity and imperfect expert information. Not accounting for heterogeneity introduces both homogeneity bias (Demont et al., 2008a) and pricing bias into the impact estimates. Moreover, the presence of heterogeneity explains the strategy of the innovating seed sector to engage in third-degree price discrimination, as the latter maximizes corporate revenue but possibly reduces total welfare.

Finally, our proposed parametric modeling framework provides a practical tool for breeders, technology developers, agricultural economists, crop protectionists, and biosafety regulators to estimate the value of future proprietary seed technologies under data scarcity. In particular, it can be useful for large-scale, ex-ante impact assessment under time and resource constraints, such as the socio-economic assessment of new technologies under the Cartagena Protocol. Falck-Zepeda (2009) argues that inclusion of socio-economic considerations may become an obstacle to potentially valuable technologies, particularly for developing countries facing higher barriers in terms of biosafety regulatory compliance due to resource constraints.


Acquaye, A.K.A., & Traxler, G. (2005). Monopoly power, price discrimination, and access to biotechnology innovations. AgBioForum, 8(2 & 3), 127-133. Available on the World Wide Web:

Alexander, C., & Goodhue, R. (2002). The pricing of innovations: An application to specialized corn traits. Agribusiness, 18(3), 333-348.

Alston, J.M., Hyde, J., Marra, M.C., & Mitchell, P.D. (2002). An ex ante analysis of the benefits from the adoption of corn rootworm resistant transgenic corn technology. AgBioForum, 5(3), 71-84. Available on the World Wide Web:

Anderson, J.R., & Hardaker, B.J. (2003). Risk aversion in economic decision making: Pragmatic guides for consistent choice by natural resource managers. In J. Wesseler, H.P. Weikard, & R.D. Weaver (Eds.), Risk and uncertainty in environmental economics (pp. 171-187). Cheltenham, UK: Edward Elgar Publishing.

Arrow, K.J. (1962). Economic welfare and the allocation of resources for invention. In R.R. Nelson (Ed.), The rate and direction of inventive activity: Economic and social factors. Princeton, NJ: Princeton University Press.

Babcock, B., Choi, K., & Feinerman, E. (1993). Risk and probability premiums for CARA utility functions. Journal of Agricultural and Resource Economics, 18(1), 17-24.

Breustedt, G., Muller-Scheessel, J., & Latacz-Lohmann, U. (2008). Forecasting the adoption of GM oilseed rape: Evidence from a discrete choice experiment in Germany. Journal of Agricultural Economics, 59(2), 237-257.

Brookes, G., & Barfoot, P. (2009). Global impact of biotech crops: Income and production effects 1996-2007, AgBioForum, 12(2), 184-208. Available on the World Wide Web:

Burgener, P.A., Feuz, D.M., & Wilson, R.G. (2000, June). The economics of transgenic sugarbeet production. Paper presented at the Western Agricultural Economics Association Annual Meetings, Vancouver, Canada.

Coyette, B., Tencalla, F., Brants, I., & Fichet, Y. (2002). Effect of introducing glyphosate-tolerant sugar beet on pesticide usage in Europe. Pesticide Outlook, 13(5), 219-223.

Demont, M. (2006). Economic impact of agricultural biotechnology in the European Union: Transgenic sugar beet and maize (Dissertationes de Agricultura No. 713). Leuven: Katholieke Universiteit Leuven.

Demont, M., Cerovska, M., Daems, W., Dillen, K., Fogarasi, J., Mathijs, E., et al. (2008a). Ex ante impact assessment under imperfect information: Biotechnology in New Member States of the EU. Journal of Agricultural Economics, 59(3), 463-486.

Demont, M., Daems, W., Dillen, K., Mathijs, E., Sausse, C., & Tollens, E. (2008b). Regulating coexistence in Europe: Beware of the domino-effect! Ecological Economics, 64(4), 683-689.

Demont, M., & Dillen, K. (2008). Herbicide tolerant sugar beet: The most promising first-generation GM crop? International Sugar Journal, 110(1318), 613-617.

Demont, M., Dillen, K., Mathijs, E., & Tollens, E. (2007). GM crops in Europe: How much value and for whom? EuroChoices, 6(3), 46-53.

Demont, M., & Tollens, E. (2004). Ex ante welfare effects of agricultural biotechnology in the European Union: The case of transgenic herbicide tolerant sugarbeet. In R.E. Evenson, & V. Santaniello (Eds.), The regulation of agricultural biotechnology (pp. 239-255). Wallingford, UK: CAB International.

Demont, M., Wesseler, J., & Tollens, E. (2004). Biodiversity versus transgenic sugar beet: The one Euro question. European Review of Agricultural Economics, 31(1), 1-18

Desquilbet, M., & Bullock, D. (2009). Who pays the costs of non-GMO segregation and identity preservation? American Journal of Agricultural Economics, 91(3), 656-672.

Devos, Y., Demont, M., Dillen, K., Reheul, D., Kaiser, M., & Sanvido, O. (2008). Coexistence of genetically modified (GM) crops and non-GM crops in the European Union. A review. Agriculture for Sustainable Development, 29(1), 11-30.

DeVuyst, C.S., & Wachenheim, C.J. (2005). American crystal sugar: Genetically enhanced sugar beets? Review of Agricultural Economics, 27(1), 105-116.

Dewar, A.M., May, M.J., & Pidgeon, J. (2000). GM sugar beet—The present situation. British Sugar Beet Review, 68(2), 22-27.

Dillen, K., Demont, M., & Tollens, E. (2008). European sugar policy reform and agricultural innovation. Canadian Journal of Agricultural Economics, 56(4), 533-553.

Dillen, K., Demont, M., & Tollens, E. (2009) Global welfare effects of GM sugar beet under changing EU sugar policies. AgBioForum, 12(1), 119-129. Available on the World Wide Web:

F.O. Licht. (2005). World sugar yearbook 2006. Kent, UK: Author.

Falck-Zepeda, J. (2009). Socio-economic considerations, Article 26.1 of the Cartagena Protocol on Biosafety: What are the issues and what is at stake? AgBioForum, 12(1), 90-107. Available on the World Wide Web:

Falck-Zepeda, J.B., Traxler, G., & Nelson, R.G. (2000). Surplus distribution from the introduction of a biotechnology innovation. American Journal of Agricultural Economics, 82(2), 360-369.

Flannery, M.L., Thorne, F.S., Kelly, P.W., & Mullins, E. (2004). An economic cost-benefit analysis of GM crop cultivation: An Irish case study. AgBioForum, 7(4), 149-157. Available on the World Wide Web:

Frisvold, G.B., Reeves, J.M., & Tronstad, R. (2006). Bt cotton adoption in the United States and China: International trade and welfare effects. AgBioForum, 9(2), 69-78. Available on the World Wide Web:

Fulton, M., & Giannakas, K. (2001). Agricultural biotechnology and industry structure. AgBioForum, 4(2), 137-151. Available on the World Wide Web:

Gianessi, L.P., Sankula, S., & Reigner, N. (2003). Plant biotechnology: Potential impact for improving pest management in European agriculture, sugarbeet case study. Washington, DC: National Center for Food and Agricultural Policy (NCFAP).

Gianessi, L.P., Silvers, C.S., Sankula, S., & Carpenter, J.E. (2002). Plant biotechnology current and potential impact for improving pest management in U.S. agriculture: An analysis of 40 case studies. Washington, DC: NCFAP.

Goldsmith, P.D., Ramos, G., & Steiger, C. (2003). A tale of two businesses: Intellectual property rights and the marketing of agricultural biotechnology. Choices, 3rd Quarter, 25-30.

Gomez-Barbero, M., Berbel, J., & Rodriguez-Cerezo, E. (2008). Bt corn in Spain―The performance of the EU’s first GM crop. Nature Biotechnology, 26(4), 384-386.

Goodwin, B., & Ker, A.P. (2002). Modeling price and yield risk. In R. Just & R. Pope (Eds.), A comprehensive assessment of the role of risk in U.S. agriculture (pp. 290-323). Norwell, MA: Springer.

Gouse, M., Pray, C.E., & Schimmelpfennig, D.E. (2004). The distribution of benefits from Bt cotton adoption in South Africa. AgBioForum, 7(4), 187-194. Available on the World Wide Web:

Hareau, G.G., Mills, B.F., & Norton, G.W. (2006). The potential benefits of herbicide-resistant transgenic rice in Uruguay: Lessons for small developing countries. Food Policy, 31(2), 162-179.

Hermann, O. (1996). Enquête sur les parasites et sur le coût des traitements phytosanitaires dans la culture de la betterave sucrière en Europe [Survey of parasites and the cost of pesticide treatments in growing sugar beet in Europe]. International Confederation of European Beet Growers (CIBE) XXXVIth Congress Proceedings (pp. 1-8). Paris: CIBE.

Hermann, O. (1997). Enquête sur le coût de la lutte phytosanitaire dans la culture de la betterave sucrière en Europe [Survey on the cost of the pest in the cultivation of sugar been in Europe]. IIRB 60th Congress Proceedings (pp. 469-475). Brussels: International Institute for Beet Research (IIRB).

Hermann, O. (2006). Survey on pests and diseases in sugar beet growing and the cost of plant protection products in Europe. CIBE congress 2006, Cracow. Brussels: CIBE.

Institut Technique de la Betterave (ITB). (2000). Unpublished document. Paris: Author.

Institute of Sugar Beet Research (IRS). (2004). Bietenstatistiek 2004 [Beet statistics 2004]. Bergen Op Zoom, Netherlands: Author.

Just, R. (2006). Anticompetitive impacts of laws that regulate commercial use of agricultural biotechnologies in the United States. In R. Just, J.M. Alston, & D. Zilberman (Eds.), Regulating agricultural biotechnololy: Economics and policy (pp. 353-396). New York: Springer.

Just, R., Hueth, D., & Schmitz, A. (2004). The welfare economics of public policy. Cheltenham, UK: Edward Elgar.

Kilburn, D. (2009, February). US growers tackle weeds with herbicide-tolerant beet. Farmers Guardian.

Kniss, A.R., Wilson, R.G., Martin, A.R., Burgener, P.A., & Feuz, D.M. (2004). Economic evaluation of glyphosate-resistant and conventional sugar beet. Weed Technology, 18(2), 388-396.

Lapan, H.E., & Moschini, G. (2000). Incomplete adoption of a superior innovation. Economica, 67(268), 525-542.

Lemarié, S., Desquilbet, M., Diemer, A., Marette, S., Levert, F., Carrère, M., & Bullock, D.S. (2001). Les répartitions possibles entre les acteurs de la filière agro-alimentaire des gains éventuels tirés des plantes transgéniques en France [The possible divisions between the actors of the agri-food potential gains from transgenic plants in France] (Rapport au Commissariat Général du Plan [Report to the Commissioner General Plan]). Rennes, France: Institut National de la Recherche Agronomique (INRA).

Lemarié, S., & Marette, S. (2003). Substitution and complementarities in the biotechnology and pesticide markets: A theoretical framework. In N.G. Kalaitzandonakes (Ed.), The economic and environmental impacts of agbiotech: A global perspective (pp. 287-305). New York: Kluwer Academic/Plenum Publishers.

Märländer, B. (2005). Weed control in sugar beet using genetically modified herbicide-tolerant varieties—A review of the economics for cultivation in Europe. Journal of Agronomy & Crop Science, 191(1), 64-74.

Marra, M.C., & Piggott, N.E. (2006). The value of non-pecuniary characteristics of crop biotechnologies: A new look at the evidence. In R.E. Just, J.M. Alston, & D. Zilberman (Eds.), Regulating agricultural biotechnology: Economics and policy (pp. 145-177). New York: Springer.

May, M.J. (2003). Economic consequences for UK farmers of growing GM herbicide tolerant sugar beet. Annals of Applied Biology, 142(1), 41-48.

Miller, G. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63(2), 81-97.

Mitchell, P.D., Gray, M.E., & Steffey, K.L. (2004). A composed-error model for estimating pest-damage functions and the impact of the Western corn rootworm soybean variant in Illinois. American Journal of Agricultural Economics, 86(2), 332-344.

Moschini, G., & Lapan, H.E. (1997). Intellectual property rights and the welfare effects of agricultural R&D. American Journal of Agricultural Economics, 79(4), 1229-1242.

Moschini, G., Lapan, H.E., & Sobolevsky, A. (2000). Roundup Ready soybeans and welfare effects in the soybean complex. Agribusiness, 16(1), 33-55.

Oehmke, J., & Wolf, C. (2004). Is Monsanto leaving money on the table? Monopoly pricing and Bt cotton value with heterogeneous adopters. Journal of Agricultural & Applied Economics, 36(3), 705-718.

Qaim, M., & Traxler, G. (2005). Roundup Ready soybeans in Argentina: Farm level and aggregate welfare effects. Agricultural Economics, 32(1), 73-86.

Research and Information Institute for Agricultural Economics (AKII). (2004). Hungarian food and agricultural statistics: 2003. Budapest: Author.

Rice, C.A., Mesbah, A., & Miller, S.D. (2001). Economic evaluation of weed management systems in sugarbeets. Proceedings of the American Society of Sugar Beet Technologists, 31, 64.

Schäufele, W.R. (2000). Chemische Unkrautbekämpfung in Zuckerrüben im Wandel—Ergebnisse einer Befragung in der IIRB-Arbeitsgruppe “Unkrautregulierung” [Chemical weed control in sugar beet in transition—Results of a survey in the IIRB working group on weed control]. International Institute for Beet Research (IIRB) 63rd Congress Proceedings (pp. 93-109). Brussels: IIRB.

Schmalensee, R. (1981). Output and welfare implications of monopolistic third-degree price discrimination. American Economic Review, 71(1), 242-247.

Smith, G.L. (2007). Sugar policy reform in the European Union and in world sugar markets. Paris, France: Organisation for Economic Co-operation and Development (OECD).

Traxler, G., Godoy-Avila, S., Falck-Zepeda, J.B., & Espinoza-Arellano, J.d.J. (2003). Transgenic cotton in Mexico: A case study of the Comarca Lagunera. In N.G. Kalaitzandonakes (Ed.), The economic and environmental impacts of agbiotech: A global perspective (pp. 183-202). New York: Kluwer Academic/Plenum Publishers.

Urban, J., Pulkrabek, J., Valenta, J., Beckova, L., & Kviz, Z. (2008). Influence of lower and more frequent doses of herbicides on yield and technological quality of sugar beet. 43rd Croatian and 3rd International Symposium on Agriculture Proceedings (pp. 646-649). Opatija, Croatia.

Weaver, R.D. (2004). R&D incentives for GM seeds: Restricted monopoly, non-market effects, and regulation. In R.E. Evenson, & V. Santaniello (Eds.), The regulation of agricultural biotechnology (pp. 143-151). Wallingford, UK: CAB International.

Weaver, R.D., & Wesseler, J. (2004). Monopolistic pricing power for transgenic crops when technology adopters face irreversible benefits and costs. Applied Economics Letters, 11(15), 969-973.

Wesseler, J., Scatasta, S., & Nillesen, E. (2007). The Maximum Incremental Social Tolerable Irreversible Costs (MISTICs) and other benefits and costs of introducing transgenic maize in the EU-15. Pedobiologia, 51(3), 261-269.

World Bank. (2008). World development indicators [database]. Available on the World Wide Web:

Suggested citation: Dillen, K., Demont, M., & Tollens, E. (2009). Corporate pricing strategies with heterogeneous adopters: The case of herbicide-resistant sugar beet. AgBioForum, 12(3&4), 334-345. Available on the World Wide Web:
© 2010 AgBioForum | Design and support provided by Express Academic Services | Contact ABF: