Katholieke Universiteit Leuven
Department
SOCIO-ECONOMIC ASSESSMENT
DIABROTICA VIRGIFERA
Koen, DILLEN, Tinne, VAN LOOY, Eric, TOL
This paper (pdf) can be downloaded following the link:
http://www.
Agricultural and Food Economics Section
Celenstijnenlaan 200E, B
Katholieke Universiteit Leuven
Department of Earth and Environmental Sciences
Working paper 102/2009
ECONOMIC ASSESSMENT OF CONTROLLING THE INVASIVE SPECIES
DIABROTICA VIRGIFERA VIRGIFERA IN CENTRAL EUROPE
Koen, DILLEN, Tinne, VAN LOOY, Eric, TOLLENS
May 2009
This paper (pdf) can be downloaded following the link:
http://www.biw.kuleuven.be/aee/clo/wp/Dillen2009.pdf
Agricultural and Food Economics Section Katholieke Universiteit Leuven
Celenstijnenlaan 200E, B-3001 Leuven-Belgium
NVASIVE SPECIES
EUROPE
Dillen, K., T. Van Looy and E. Tollens. "Socio economic assessment of controlling the invasive species Diabrotica virgifera virgifera in Central Europe" Working Paper, n°102, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, 2009.
Koen Dillen, Tinne Van Looy & Eric Tollens
Centre for Agricultural and Food Economics,
Department of Earth and Environmental Sciences
Celestijnenlaan 200E, bus 2411,
3001 Heverlee
Tel.: +32 16 32 23 97
Email: [email protected]
Copyright 2009 by Koen Dillen, Tinne Van Looy and Eric Tollens. All rights reserved. Readers
may make verbatim copies of this document for non-commercial purposes by any means,
provided that this copyright notice appears on all such copies.
1
Abstract
Diabrotica virgifera virgifera or western corn rootworm (WCR), is a major pest of cultivated
maize, Zea mays L.. It was recently introduced into Europe, where it was first observed near
Belgrade, Serbia in 1992. The beetle spread through Central Europe since, leading to a
continuous population in 11 countries, from Austria to the Ukraine and from southern Poland to
Serbia. In this paper a socio- economic assessment of the possible control options is presented in
eight of these countries. A farm level model is designed which, based on several Monte Carlo
simulations, estimates the created rents for different control options. The simulation approach
allows to account for the ex ante setting of the study, the limited data availability, the
heterogeneity among potential adopters and a high uncertainty about potential damage. The latter,
stemming from the low correlation between the pest’s population pressure and yield loss, creates
a risk for maize producers. Therefore risk behavior of farmers is incorporated in the model
through the use of a utility function introducing constant absolute risk aversion, and risk neutral
behavior. Furthermore, the model exploits the heterogeneity among farmers to endogenize the
technology premium of the proprietary transgenic innovation and its potential adoption rate.
Therefore an aggregation can be made to the welfare effects at national level. For Hungary, the
modeled results are complemented by survey results giving insight in the production constraints
and the willingness to pay for the transgenic control option.
2
Table of Contents
Abstract ............................................................................................................................................... 1
Introduction ......................................................................................................................................... 4
The nature of the pest .......................................................................................................................... 5
Available control options .................................................................................................................... 8
Cultural control ............................................................................................................................... 8
Chemical control ............................................................................................................................. 9
Transgenic crops ........................................................................................................................... 12
Other control options .................................................................................................................... 13
Explorative characteristics of the invasive pest in Hungary ............................................................. 14
Economic assessment of the invasive species WCR ........................................................................ 19
Methodology at the farm level ...................................................................................................... 19
Data ............................................................................................................................................... 25
Simulations and results ................................................................................................................. 28
Farmers willingness to pay for Bt maize ...................................................................................... 36
Aggregation to the national level .................................................................................................. 43
Control of WCR in Czech Republic ................................................................................................. 46
Control of WCR in Slovakia ............................................................................................................. 52
Control of WCR in Austria ............................................................................................................... 57
Control of WCR in Poland ................................................................................................................ 61
Control of WCR in the Republic of Serbia ....................................................................................... 65
Control of WCRin Romania ............................................................................................................. 72
3
Control of WCR in the Ukraine ........................................................................................................ 77
Conclusions ....................................................................................................................................... 82
References ......................................................................................................................................... 86
Figures and Tables ............................................................................................................................ 94
4
Introduction
Invasive species may present a major threat to agriculture (Pimentel et al., 2001). The newly
introduced alien species can induce significant economic losses, present a public health risk
(directly or indirectly) or affect biodiversity. Especially in agriculture, where output is stochastic
in nature, an extra constraint may cause a burden for farmers exposing them to additional risk in
the production function. Therefore the introduction and development of an invasive pest should
be understood and managed using strategies that are not only suitable to specific agricultural and
economic conditions, but also in the socio-economic context of crop production (Boriani et al.,
2006). In this study, a socio-economic assessment of a new threat to European agriculture,
Diabrotica virgifera virgifera or Western Corn Rootworm (WCR), is conducted. The scope of
such a study could be very broad. The invasive species is still spreading, economic damages are
not yet stabilized, uncertainty about the efficacy of different control measures is not yet resolved
and future trends are difficult to predict. In the USA, a study with a similar research question
(Alston et al., 2002) investigated the potential value of a specific control option, biotechnology,
based on the USA experience, where the pest is well established. This study aims at a better
understanding of the different control options for WCR in the European setting. As the pest is
still spreading and the effects in European cultivation patterns are not yet clear, risk and
uncertainty play a key role in the assessment. Moreover, the ex ante setting of the assessment
endogenously leads to imperfect information. Therefore, the incorporation of heterogeneity in the
assessment is crucial (Demont et al., 2008, Dillen et al., 2008). The heterogeneous properties of
the affected farmers lead a multitude of possible outcomes. Therefore the adoption decision plays
a central role in the assessment. Therefore, the study is centered on a farm level assessment
leading to rational adoption decisions and altered farm profits. The modeling approach is
complemented by a survey in Hungary to get insight in the farmer’s perceptions surrounding
5
WCR and its control options and their willingness to pay for some non pecuniary benefits of a
particular control option. These results are aggregated to give some insight in the economy wide
effects. The study starts with an introduction on the nature of the invasive alien pest, its
introduction and development in Europe and the available control options. In a second part an in
depth analysis of the situation of both pest and maize production in Hungary is conducted. In this
section the modeling framework is developed. The results the model initiate the development of
a multcriteria adoption decision tool. In the remainder of the study, seven other countries under
threat: Czech Republic, Slovakia, Austria, Poland, Serbia, Romania and Ukraine are assessed for
the regional optimal control options against WCR. Finally a general conclusion which can aid
both policy makers and farmers to decide which control options should be taken in order to
protect the maize production in the countries under research, and by extension European maize
production, from WCR.
The nature of the pest
Diabrotica vigifera virgifera or Western Corn Rootworm (WCR), is a major pest of
cultivated maize, Zea mays L. In the Midwestern USA, it is the most important maize pest with
an estimated $1 billion loss in yields and control expenditures annually (Metcalf, 1986).
Nowadays these costs may be considerably higher as resistance to some control options has
evolved since, making control more costly. Moreover, WCR is the insect pest which causes the
highest insecticide use in the world (Baufeld & Enzian, 2005c). Most of the damage to the crop is
caused by larvae feeding on the root system of maize although some economic damage may
occur through adult beetles feeding on the silk. Adult feeding is especially the case in the
production of sweet and seed corn due to its high value. Predicting and assessing the damage
caused by the species is very difficult as exogenous factors as drought, soil conditions affect the
6
resulting damages and yields to a high extent (i.e. Rice & Oleson, 2005). Larvae damage may be
offset by the regrowth of the root system if water is available during the appropriate time window
of the growing process (Simic et al., 2008). This leads to difficulties or even impossibilites in
determining the correlation between population pressure and damages accordign to Urias-Lopez
and Meinke (2001). As the damage caused by WCR is highly stochastic and related to exogenous
factors, it is also very difficult to assess the resulting damage ex post as no control group is
available. Some papers present (Mitchell, 2002;2004) quantitative relations between root ratings
in the USA, but overall a high uncertainty surrounds the damage of the pest. This uncertainty
inflicts problems in determining the optimal control options in the field. In particular conditions it
can lead to a total loss of production due to lodging or goose necking, in other conditions no
economic damage can be confirmed despite a high population pressure.
The species probably originated in Central America but has spread over the United States
corn belt, reaching the east coast of North America (Gray et al., 2009). It was recently introduced
into Europe, where it was first observed near Belgrade, Serbia in 1992 (Bacca, 1993). Since, at
least four other introductions from the USA into Europe took place, presumably travelling with
airplanes (Ciosi et al., 2008). Two types of infested area have been identified: (i) areas of
continuous spread (in Central and southeastern Europe and northwestern Italy), and (ii) several
disconnected outbreaks that did not persist over time or did not spread due to enforced
eradication measures in place. The Central European outbreak now extends over 11 countries,
from Austria to the Ukraine and from southern Poland to Serbia. The most recent information
available on the spread can be found in Figure 1. WCR is listed in Annex IAI of the EC Directive
2000/29/EC. Organisms listed on Annex IAI are harmful organisms whose introduction into, and
spread within all Member States shall be banned. However, WCR has an established population
7
in parts of the EU inclusion in Annex IAII (harmful organisms known to occur in the community
and relevant for the entire community) may be reasonable in the future to ensure the availability
of effective control options for farmers. Especially since studies indicate that the spread of WCR
in Europe is only a matter of time and cannot be avoided (e.g. Baufeld & Enzian, 2005a). Note
that the Directive does not distinguish between the two subspecies of D. virgifera, i.e. between D.
virgifera virgifera (WCR) and D.virgifera zeae (Mexican corn rootworm). This distinction should
be recognised so that D. virgifera zeae becomes listed in Annex IAI and WCR is listed within
AnnexIAII. Extra decisions of the European Union including some information on WCER: EC
Decision 2003/766/EC2, which prescribes an annual survey and phytosanitary treatments, and
2006/564/EC3, which prescribes measures around airports which also apply to WCR.
The steady spread of this invasive species causes a threat to maize production in times with
generally good prospects for European maize production. Among the structural causes of this
expected appreciation are (i) the steady rise in global commodity demand driven by record
economic growth rates, urbanization and changes in dietary patterns (notably for meat) in many
parts of the world (in particular India, China and Latin America); (ii) the emergence of new
market outlets such as the biofuels market (mainly in the US where this market is estimated to
absorb around 25% of US maize production in 2007/08, whereas EU biofuels production would
only use less than 1% of domestic cereals production); (iii) the significant slowdown in cereal
yield growth in the EU (unlike many other producing regions) (European Commission, 2008).
The continuously high and increasing maize demand in the USA could lead to a lasting change of
relative prices in favor of coarse grains. Furthermore, the abolishment of intervention prices
should increase fluidity in Eastern European markets despite permanent high transportation costs
and therefore offer a bigger marketing potential for European maize. These prospects increase the
8
importance of suitable pest control strategies in European maize production in order to exploit
these opportunities.
Available control options
Cultural control
A first control option, which has been succesfully used for a long time in the USA, is directly
related to the nature of the pest. The univoltine beetle lays his eggs during late summer, mainly in
maize fields where they overwinter and hatch the next spring. The larvae cannot differentiate
between the roots of plant species (Krysan & Miller, 1986) and have limited mobility, therefore
feeding on the roots in its vicinity. However, the potential of crops besides maize to act as a
feeding source is low (Branson and Ortman, 1967). Therefore rotating maize with a non host crop
offers a practical solution to limit population growth and overcome damage. Up to recently crop
rotation with soybean, a dicotyledonous on which no larvae feeding occurs, was the dominant
control option in the USA. However, since the early nineties damage in first year corn occurs
despite rotating (Levine and Oloumi-Sadeghi, 1996). In a behavioral study the only difference
found between the rotation resistant and non resistant variety was the level of mobility (Knolhoff
et al., 2006). Indeed, adults can disperse widely (perhaps up to 100 km per year) (Coates,
Tollefson, & Mutchmor, 1986), are highly fecund (Elliot, Gustin, & Hanson, 1990) and crop
rotation might have favored this behavior (Levine et al., 2002). This means there is no preference
for soybean by the resistant variety, only lack of fidelity to maize as an oviposition site for adult
females (detailed discussion in Gray et al., 2009). European maize production differs regionally
but is generally different from the USA due to a more diverse array of cropping patterns (Kiss et
al.2005). This diminishes the pressure towards resistant populations with the same
charactheristics. However, the same high agrobiodiversity increases the potential presence of
9
other host crops in the production system. Moeser and Vidal (2004; 2005) studying European
populations of WCR, found that feeding on some grasses was significant and adults emerged.
Gloyna and Thieme (2007) tested WCR development on barley, oats, spelt, triticale and wheat.
On all of these crops, possible to use as a rotational crops except oats, adult WCR development
took place. Unfortunately none of these studies assessed whether these adults were fecund and if
so where egg deposition takes place. These results show that a rotation with solely
monocotyledonous crops may not be a sustainable solution as adaptation could occur in the WCR
population. Currently crop rotation is the main management strategy controlling WCR in regions
where crop rotation is economical and as a tool to eradicate the quarantine pest.
Besides crop rotation, delayed planting of maize offers the possibility to overcome WCR
damage. In Italy, delayed planting of maize until late May or early June resulted in acceptable
levels of silage maize and prevented adults emerging (Furlan, 2006), presumably due to the lack
of feed for the larvae in the appropriate development state. However, Baca et al. (2003) found
higher damage in second year plantings after late planting in Serbian trials, presumably because
of the higher attraction of imagoes. Therefore even in suitable climatical conditions, significant
reduced yields (Hyde et al., 1999) and the lagged effect on second year corn delayed planting
offers limited scope for a rational control option.
Chemical control
Both in the USA and in selected European countries, chemical control strategies are applied
to both reduce the WCR population and prevent damage. These chemical options include soil
insecticides and seed treatment as a protection against larvae damage and foliar insecticides to
suppress adult populations and protect high value cobs from silk feeding. Determining economic
thresholds for chemical control options is a sensitive matter. Standard procedures as scouting and
10
trapping are of low use due to the low predictability of the resulting damage.. However, a multi-
criteria decision tool, taking into account other exogenous factors in the production function may
be more suitable to decide upon a rational control option.
Soil insecticides require an at-planting choice of the farmer as application takes place either
as a granular or in furrow during the planting process. This leads to a decision making under
imperfect information. Because larval damage was poorly predicted by adult counts, a study by
Foster et al. (1986) argues that applying soil insecticides prophylactically every year to
continuous cornfields is more cost effective than sampling for adults. However, the efficacy and
consistency of soil insecticides depends upon a number of factors: the applied active ingredient,
timing of the application because of limited persistence in the soil (about 6 weeks), leaching,
physical and chemical composition of soil, mechanical and operational aspects,… (for a detailed
review see Gerber (2003)). These factors result in a highly heterogeneous outcome of the soil
insecticide application. Therefore prophylactic application may not be the rational choice for each
farmer as more detailed knowledge about these constraints is available to them.
Seed treatment requires an at-buying decision of the farmer, increasing the imperfect nature
of the information available in the adoption process compared to soil insecticides. On the other
hand as the insecticides are placed directly onto the seeds via a coating, both spatial and temporal
application is optimized and management is facilitated. Although these products provide
adequate protection under low population pressure, they tend to more variable in protection than
traditional soil insecticides under high pressure (Cox, Shields, & Cherney, 2007, Horak et al.,
2008, Ma, Meloche, & Wei, 2009). Important to note is that some seed treatments offer
protection against wireworms at the same time and therefore might have a higher attraction to
farmers (Muska, 2008). Furthermore, nor soil insecticides, nor seed treatment are an efficient way
11
to control population dynamics. Both offer the ability to farmers to reduce damage but not
necessarily the presence of WCR in the field (Furlan et al., 2006) introducing a potential for
lagged damages.
Adult control can be achieved through foliar application of insecticides. The decision whether
or not to adopt this control strategy takes place later in the production process. Although adult
control might be rational in high value seed and sweet maize to reduce silk feeding it does not
directly protect from larvae feeding. By reducing the population, egg laying is prevented which
reduces the larvae damage in the next season. First-year preliminary results indicate that the
economic threshold for WCR adults is between 3-6 adults/ear for inbred lines and above 9
adults/ear for commercial corn (Tuska et al., 2001). These estimates are more reliable than for
root damage as there is a direct relation between adult counts and damages. Foliar spraying is
done by either high clearance tractors or aerial application which not only increases application
costs but also depends on the availability of this machinery to the farmer.
Based on the experience gained in the US where insecticide resistance for some active
ingredients developed in WCR populations (Meinke et al., 1998), cautious resistant management
is a key factor in pertaining the efficiency of the different control options. Especially the limited
and reducing availability of active ingredients allowed for use in the EU due to environmental
concerns could create a genetic bottleneck. In Hungary, the country with the highest WCR
infestation, only one active ingredient was available as a soil insecticide, tefluthrin, in 2008
(Ripka, 2008). This combined with the findings of Ciosi (2008) proofing multiple introductions
of WCR in Europe and the likely increase in the probability of adaptations to management
strategies points towards the need of a diversified approach towards WCR in Europe.
12
Transgenic crops
Biotechnology offers a potential control option against WCR. In the USA a genetically
modified maize resistant to WCR damage has been commercially introduced in 2003. The
genome contains a coleopteran specific insecticidal toxin form the soil bacterium Bacillus
thuringiensis (Bt). The adoption of these Bt varieties was succesfull at the expense of soil
insecticides. In 2005 already about 1.8 million hectares were planted with WCR resistant maize.
In 2006 it was expected that the area would increase beyond the number treated with soil
insecticide because of the introduction of stacked plants which include traits against several pests
at the same time. In 2008 58% of the maize area in the USA had a Bt gene implanted1 (NASS,
2008). The Bt technology gives a systemic protection to WCR and, as it is incorporated in the
root, its performance is unlikely to be affected by environmental conditions nor of planting time,
soil conditions, calibration of machinery (Mitchell, 2002). The results in the USA indeed show
high efficiency of the trait to avoid economic loss and high efficacy (Ward et al., 2005). Similar
to seed treatment, the application of Bt maize is a at buying decision for the farmer. In Europe
regulatory approval of GM technologies has been slow and for the moment no WCR resistant
variety is allowed for cultivation. The single event MONA863 is however allowed for for food
and feed use in the EU (Agbios, 2008).
1 This also includes the area of Bt maize that only carries resistance against the European Corn Borer but
disaggregated data are not available.
13
Other control options
Some other options are being developed to prevent damages from WCR in maize production,
the so-called biological control options. Within this group we find biological chemicals, use of
natural enemies (including fungi, nematodes,…), natural habitat changes and so on. From a
societal point of view these option may seem favorable as the social reversible and irreversible
costs are low. However, at present the private costs for farmers are high and therefore adoption is
discouraged. Fall & Wesseler (2007a) show that biological control options are not economically
viable in grain and silage production and in high value production (sweet corn, organic corn and
seed production) where crop rotation is often compulsory. Another protection option that is
currently developed is the resistance to WCR damage via conventional breeding techniques.
However at the moment no variety of this kind is commercially available and therefore it is not
included in our analysis. If these varieties come to commercialization, an assessment would be
necessary.
14
Explorative characteristics of the invasive pest in Hungary
In the summer of 1995 WCR appeared in Hungary, via the southern county Csongrad and in
2000 it crossed the north border. In 2000 the first economic damage by larvae to maize roots was
observed and by 2003 no uninfested plots were left by the end of 2003 (Boriani et al., 2006).
Table 1 displays some key figures on the maize sector in Hungary. Maize is the most important
crop in Hungary with about 1.2 million hectares sown each year, closely followed by wheat. The
sown area has been stable despite changes in farm gate prices and the significant decrease in
livestock since 2000. Cattle decreased with 12%, pigs with even 30% (AKII, 2008a). On average
about 40% of the maize area is cultivated without annual rotation, in monoculture. The
percentage differs in each county and depends on farm size and profitability of competing crops.
In absence of detailed data one can assume a high percentage under monoculture in smaller fields
of individual farmers (Toth, 2005).
To get some detailed insight in the Hungarian maize sector after accession to the EU in 2004
and the perception towards WCR, a computer assisted telephone survey of maize farmers was
undertaken by Kleffmann &Partner Kft Hungary in October 2008. The survey sample was
randomly taken from their database of maize producers. The sample was weighted towards the
cultivation of maize in the region, Danube valley (200), North (30) and the Great Plain (220). The
farmers grew maize on 34% (σ=19%) of their cultivated area (of which only 15% is owned) or
253ha (σ=394). If farmers were cultivating maize in monoculture in 2008, it was on average 36%
(σ= 28%) for silage maize and 44% (σ=31%) for grain maize. Farmers in the Great Plain have
significantly more area under monoculture which coincides with the figures of AKII (2008)
stating that gross margins are the largest in this region. The high variance signifies a high
heterogeneity among production systems as pointed out before. If maize is rotated, the prominent
15
rotation schemes include winter wheat, barley, sunflower and rape seed. Most farmers have
limited rotation systems, with 62% stating winter wheat as their primary pre crop and 44% stating
it as the primary post crop. In a secondary rotation scheme we have winter barley and sunflower
and to less extent rape seed. The maize variety used is endogenous to the production system. In
monoculture or in rotation with sunflower or rape seed, varieties with medium FAO number
(300-380) are used while shorter maturing (lower FAO number) varieties are used in a rotation
scheme with winter wheat.
The presence of livestock on the farm influences the choice of maize type and the
participation in markets. Almost half of the farmers (43%) have some kind of livestock on the
farm, 51% of these have pigs and 49% have cattle which is not complementary. This herd can be
complemented with some other animals (poultry 12%, sheep 10%). Farmers in a mixed structure
cultivate a significant higher acreage with silage maize and in the case of cattle the total maize
area significant higher. Silage maize is mainly produced for the spot market or for farm use
(35%), only a limited amount (11%) is planted on a contract base for which the expected price is
lower than on the spot market. The supply of own fodder is almost sufficient since only 9% of the
mixed farms buys extra fodder. Grain maize on the other hand is mainly used on farm (47%) or
produced on a contract basis (34%) with limited sells on the spot market (19%). Interesting to
note is the significant correlation between expected yield and contract production. Yhe most
efficient farmers seem favored to produce on contract.
The respondents were asked about their control measures against pests in their maize fields.
The most important pest2, for which 82% of the respondents applied some control measure, was
2 Limited amount of respondents (56%) had knowledge about all the pests within their field. We only used these
farmers to determine importance of different pests and control options.
16
the invasive species Diabrotica virgifera virgifera. 11% of the farmers controlled for Aphids and
on a third place came the European Corn Borer with 10% closely followed by the other European
pests as cut and wireworms. These results are in line with previous research on pest pressure in
Hungarian agriculture, except for the invasive pest WCR which is a much higher constraint for
Hungarian farmers than all other pests combined (Nagy, Szentkirályi, & Vörös, 1999; Szõke et
al., 2002). Asked about the first detection of WCR in the field, 30% had no idea or did not detect
the beetle yet. If we look at the time series of detection in their own field (Figure 2), it seems
there is a lag of detection within the field. Official figures (Boriani et al., 2006) indicate total
coverage by the end of 2003 while at that time only 40% of the respondents detected WCR in
their field. This limited knowledge with farmers could indicate a problem in designing the
appropriate control strategy by farmers. However, as argued before, the adoption decision should
be based on a multi-criteria decision in which damage is only one criteria and therefore the
decision may be reasonable even with this information constraint present.
The most widely used protection against WCR among farmers is soil decontamination with
59% of the maize farmers using the technique on an average area of 203 ha. Crop rotation and
seed protection follow closely with respectively 47% on 228 ha and 44% on 196 ha. Interesting to
note is the fact that farmers see crop rotation as a control option on some part of their arable land
and not as a part of the normal production system. Foliar treatment has a smaller application
density with only 20% of the farmers using it on 91 ha on average with application rates between
1 and 4 (average 1.28). As mentioned before adult treatment is only economical in sweet maize,
seed production or under high population pressure which explains the lower figures. All the
measures have a very high coefficient of variation (~1) which is endogenous to the heterogeneous
maize production system in Hungary.
17
Some interesting properties of the control options can be extracted from the data. Crop
rotation has a significant negative dependence with soil insecticides but positive with seed and
foliar treatment. This can be explained by the broad control spectrum of the seed treatment as
indicated before. As it protects maize from other pests it might still be used even if rotation is
sufficient to reduce WCR damage. Seed treatment and foliar application are combined again
confirming the specific range of usage for foliar treatment. In monoculture soil insecticides,
negatively dependent with other control options, are the preferred sole option as they have a high
efficiency in this setting. Respondents were also asked about their change in production system
after detecting WCR in their field. 52% of the farmers decreased their acreage under monoculture
with on average 59% (σ=30%). Compared to the respondents rotation as a control option (47%)
this indicates that 5% of the farmers introduced crop protection as a precautionary measure and
don not account for it as a control measure. This is presumably a result from the fact they don’t
have limited land resources and can still produce exactly the same output mix by rotating.
In the survey farmers where sounded out their decision influencing characteristics in the
adoption of a certain pest control measure. Preferences were elicited through a five point Likert
scale. Although all factors mentioned are important to farmers (even the least important,
application equipment, averaged 3.2), a significant ranking could be constructed through paired t-
tests. The direct monetary value of increased yields was stated as the most important (4.2)
criteria. Management costs, personal health and environmental effects are regarded on a second
level (4.1) followed by insurance value (3.9) and application equipment.
18
Expected yield loss stemming from reduced cobs and lodging without treatment is estimated
by the farmers at 46%. In Figure 3 Epanechnikov kernel function with bandwith 20 is used to
estimate the non parametric probability density function (PDF). The resulting PDF is positively
skewed with about 50% of the farmers stating they do not fear any damage from WCR. The
estimated yield loss is strongly correlated with the amount of monoculture a respondent has.
Asked about their actual loss in the last 2 cropping seasons we see that there is important
temporal variation, in 2007 20% of the farmers had economic damage with an average yield loss
of 20% (σ=19%). In 2008 only 12% of the farmers had damage with 15% yield loss (σ=14%).
Interesting to note is that 8.5% of the farmers had an economic loss in both years, meaning their
pest protection was not adequate, they miscalculated the severity of the pest or they valued the
damage from WCR lower than the price of the control measure. Farmers using soil insecticides
perceive the potential damage significantly higher than farmers using any of the other techniques
once again confirming its status as the best solution with high population pressure in
monoculture.
19
Economic assessment of the invasive species WCR
Methodology at the farm level
As the pest is still spreading and some of the investigated control options are not yet
commercialized in the countries under research the assessment takes place in an ex ante setting.
This has two important consequences. At first a counterfactual baseline has to be determined. As
Baufeld and Enzian (2005a) argue that the spread of WCR in Europe is only a matter of time, our
counterfactual assumes this is already the case. By doing the assessment in a hypothetical year x
the time dimension disappears. Imperfect information is endogenous to the ex ante approach as
no real adoption data are available (Dillen et al., 2008). Demont et al. (2008) and Demont and
Dillen (2008) developed a model to incorporate the risk induced by the imperfect information
through explicitly incorporating the resulting heterogeneity in the population of farmers. This
heterogeneity stems from the differences in soil, managerial capacities, climate, pest pressure,
credit availability, market developments… It is important to make a distinction between risk and
uncertainty in this respect. If an event is unpredictable but can be estimated through some
statistical modeling we call it a risk factor and it is incorporated in the model. Uncertainty on the
other hand is totally unpredictable and as such has no probability and can therefore not be
modeled. We base our model on the aforementioned studies and adapt it to the specific needs of
this assessment. A simulation approach is followed which transforms distributions on primary
believes to stochastic outcomes. This allows for confidence intervals on the outcomes, detailed
sensitivity analysis on the parameters and revealing the determining factors in the farmer’s multi-
criteria decision process. This is done in a rational choice framework which assumes farmers
maximize utility. At first we assume profitability is the solely determinant of utility. We
investigate the three groups of control options; chemical treatment (chem), transgenic plants (Bt)
20
and crop rotation (rot). The annual per hectare profit under each control option can be calculated
as, respectively:
chem chem chemP Y Cπ = ∆ − (1)
(1 )( ) ( )Bt Bt Bt Bt Bt baser P Y Y P C r P Yπ = − ∆ − ∆ − + (2)
2( )
3maize other
rot
w wπ
−= (3)
With
• P the price for identity preserved (IP) maize and ∆P the price discount for Bt maize or the
negative IP premium
• Ci are the costs of the different control measures on a hectare base.
• r is the percentage area not planted with GM maize in the GM control measure. The value
of r includes the value of refuge areas and the area planted with ex ante coexistence
measures (see Devos et al. (2008) for a review of the European coexistence rules).
• Y is the average yield with the application of the different control measures, ∆Y gives the
yield surplus compared to no control (Ybase).
• wmaize is the gross margin of maize and wother of the complementary crops in the 3-year
rotational plan.
In the assessment of an invasive species, the damage abatement function is crucial. As a
starting point we model the base yield, assuming no damage. Even without the presence of WCR,
the maize yield is not known at the beginning of the cropping season. Yield is an output from
germplasm and environmental influences (soil, climate, pests,…). The central limit theorem is
21
not applicable due to correlation of yields between space and time following from the systemic
nature of crop production. Furthermore, crop yield distributions are most of the time assumed to
be skewed due to the biological constraints at the higher yielding tail of the distribution
(Goodwin & Ker, 2002). Therefore the Beta PDF is commonly used to model the potential yields
in a given situation. In this study the distribution is calibrated on the trend adjusted yield. We
assume deviations from the trend (êt) to be proportional to the trend itself. Thus we regress yields
in time and then recenter the yields. This allows us to stabilize yields and projections for the yield
in the future.
^(1 )t
t
ê
t zy
y y= + (4)
With yt the recenterd yield and ��t the estimated yield and yz the unadjusted yield. This
adjusted yield is used as an input to the damage abatement function. We assume the countries
under research have no significant loss on the aggregate level at the moment (due to efficient
control or not infested)3. The yield under different control option can be modeled by:
(1 )base t untreatedy y s= − (5)
(1 . )soil base soil chemy y c e= + (6)
(1 . )Bt base Bt Bty y c e= + (7)
(1 . )seed base seed chemy y c e= + (8)
3 This might be a conservative assumption in the countries where the population of WCR is well established.
Average yields could be influenced by the adoption but determining the effect of WCR on the average yield is difficult because of the nature of the pest presented before.
22
ybase represents the yield under WCR pressure without control, suntreated represents the
percentage loss caused by WCR with no control measures present. This ybase is used to estimate
the yield under different control options. The performance of a certain control measure depends
on two factors. The theoretic yield protection by a certain technology i, ei, allows for differences
in different technologies. A second parameter, ci, denotes the consistency of the control. This
consistency can be correlated with other parameters in the model to allow for underlying
causalities. This is particularly important for seed treatment where the level of protection in
negatively correlated with the pest pressure and the overall effect of insecticide use may vary due
to bad timing and use of wrong product (Qaim, Pray, & Zilberman, 2008). Moreover we assume
there is no yield drag in the case of Bt maize.
Beside yields, different prices play a role in the profitability and therefore in the adoption
decsion. The first price is the farm gate price for the end product as this determines to a high
extent the value of the potential damage. In that respect, differentiation can be made between
farmers producing on a contract base which reduces the price risk due to the fixed price, and
those selling on the spot market or using it on the farm4. We assume the price at the farm gate is
negatively correlated with the average yield in order to capture some market effect (Goodwin,
2008). The price on the spot market we take the classical approach of a lognormal distribution
based on the variation in historic price series (Goodwin & Ker, 2002). The IP premium bears
both a risk aspect, the temporal variation, and an uncertainty aspect as because no market exists
for the product at the moment. Therefore the spread of this parameter will be rather big in our
model and theoretically cannot be assessed. The determining price in the adoption process is the
price to be paid for the different technologies. For chemical control options this price is known
4 On farm use shall be valued by the opportunity cost and as such relate to the spot market price.
23
from the actual market price in (some of) the countries under research. However as Bt maize is
not yet introduced in Europe, the technology fee to be paid to the innovator is unknown.
Therefore we base our calculation on the methodology introduced by Dillen, Demont and Tollens
(2008). The non drastic nature of the innovation and the heterogeneity in both space and time of
potential adopters creates a situation where only restricted monopoly power exists for the owner
of the proprietary right (Weaver, 2004). Each potential adopter has a valuation for the Bt
technology depending on his expectations about all feasible control options. The technology
valuation can be calculated by the by setting the initial price of the technology at zero. The result
of this simulation is a probability density function f(x) of the technology valuation of the whole
population of potential adopters. Farmers at the right tail will have a high valuation and will
likely adopt the technology while farmers at the negative left hand side would not even adopt at a
zero price. From this PDF a normalized demand curve for the technology can be constructed
(Q(θ)=1-F(x) with θ the price of the technology). We assume development costs are sunk and not
incorporated in the pricing decision of the innovating firm. Assuming constant marginal costs, c,
the profit function of the monopolistic innovator is represented by:
( ) ( ) ( )c Qπ θ θ θ= − (9)
The optimal price of the technology bundle,θ*, satisfies the following first-order condition:
( ) ( )
.( ) ( ) 0d dQ
c Qd d
π θ θθ θ
θ θ= − + = (10)
From this condition the profit maximizing technology fee, θ*, can be calculated. However, as
we don’t have access to the cost structure of the innovator for development, regulatory affairs
marketing and so on, we assume c to be zero which transforms equation 9 to a revenue
24
maximizing equation. Alexander & Goodhue (2002) argue that if users are heterogenous, θ*, may
be below the technology valuation of a potential adopter therefore leaving significant rents with
the adopter. Lapan and Moschini (2004) alternatively interpret this pricing strategy as the
monopolist choosing the marginal adopter 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 technology valuation higher than
θ* will adopt the new technology. The resulting adoption ceiling is,
*
( )f x dx
θ
ρ∞
= ∫ (11)
With the assumption of rational farmers adoption will take place if the technology valuation of a
certain farmer is bigger than θ, the marginal adopter (indifferent between accepting and refusal)
can be determined. The marginal adopter allows calculating the adoption rate through its position
in the PDF and allows calculating the average benefits created by the adopters:
*
( ). ( )x f x dx
θ
α π∞
= ∫ (12)
The results of the survey indicated that not only the direct monetary value is of importance as
a farmer’s decision criteria, but also the insurance value. This means farmers are risk averse. We
specify a negative exponential Von Neumann-Morgenstern utility function. This specification
allows introducing constant absolute risk aversion (CARA). The assumption of CARA
preferences is often used to analyze farm decision under risk (Just & Pope, 2002). As the absolute
risk aversion levels does not convey sufficient information to indicate whether the implied level
of risk aversion is reasonable, Babcock, Kwan Choi & Feinerman (1993) calculate risk premium
25
as a percentage of the wealth at risk. Following Mitchell (2004) for the case of WCR in the USA,
20% of the standard deviation of the returns as a measure for moderately risk averse farmers. The
resulting certainty equivalent (CE) might lead to an ordering of control options different from the
risk neutral case.
Data
Table 2 presents the data used as an input for the simulation model. As the starting point of
the assessment creates the problem of imperfect information, data with endogenous risk factors
are introduced as PDFs calibrated on the prior beliefs and will be transformed by the model to
stochastic results. We will describe the calibration in this section.
The lognormal PDF on price data is constructed on the mean and standard deviation of a time
series from 2000 till 2008. To allow for market effects induced by the variance in yield, this price
PDF is negatively correlated with the yield following Goodwin (2008). As discussed before, yield
distributions are often modeled through Betageneral PDFs. Under the constraints of scarce data
we opt for a special case, the PERT distribution, to model the base yield (see Dillen et al.,
2009).By parameterizing the distribution on the centered yield with zero and 150% of yt as the
extremes. This specification leads to a distribution with a constant coefficient of variation of 0.3.
This is an interesting property as it introduces a similar heterogeneity in the simulation for all
values of yt, therefore allowing the comparison of different case studies. The specification of the
base yield on the yt is conservative as it assumes that the used yield data are not yet affected by
WCR which is debatable in those countries already suffering from a high population pressure.
The high dependence of damage by WCR on stochastic exogenous makes the assessment of
yield losses a delicate matter. We base our model on the results by Mitchell (2002) presenting the
26
damage abatement of different control options from USA fields. The author presents a three
parameter estimation of the yield gain compared to the control group for each root rating. This
allows the construction of a PERT distribution for each root rating. Ideally these damage specific
yield gains should be weighted with the distribution of actual root ratings in the field. However,
as this data is not available for the European settings on a scale large enough to incorporate the
spatial and temporal variation in the pest pressure. Therefore we calibrate the PDF on USA field
trial data (Pilcher 2001). We believe the resulting (fitted) PDF (see Figure 4) captures the
possible occurrence of damage in a variety of fields, from first year corn to monoculture. The
average damage under this approach amounts to 14,1%. Comparison of this data with other
studies in the EU let us belief this construction is reasonable. Schaafma, Baufeld en Ellis (1999)
estimate the damage for Germany around 10%, Baufeld and Enzian use 10-13% (2005b) while
Macleod et al. (2007) use 25-30% for the UK. The same data and methodology is used to assess
the protection of chemical control, echem. This results in a PDF an average protection by chemical
control options of 11.7%. In this specification we assume no difference in efficiency between the
two types of chemical control, soil insecticides and seed treatment only in consistency.
Consistency is a key factor in the valuation of the different control measures. Data on
consistency is spare and mostly stemming from the US where other agro-environmental
conditions determine the consistency (i.e. Ma, 2009). We based our assumptions on field trials
(Horak et al., 2008) and compared it with data by Draper (2007) for the UK who states a
difference between GM and seed treatment of 5% damage on average. Exact data on the
triangular PDF’s can be found in Table 2 and results are in line with earlier findings as the
average difference in comparable environments between seed treatment and GM takes the value
27
of 4.5%. To account for the lower consistency of seed treatment under higher population
pressure, we include a negative correlation between these two parameters.
The costs of the different control measures are essential to investigate the competitiveness of
the different control options. For seed treatment and soil insecticide data from the Hungarian
State Phytosanitary institute are used. These prices are in line with the €40-60/ha for insecticides
according to Takacs, Balogh, and Nadasy (2007) and data we received from wholesalers in
Hungary. For seed treatment we use a uniform function within the appropriate ranges while we
assume a fixed price for soil insecticides as only teflutrin was allowed for use in the 2008 season.
In the case of Bt maize costs arise from different sources as reflected in equation 2, besides the
endogenized technology fee. The refuge area needed to ensure an effective insect resistant
management is assumed to be equal to the requirements in the USA, where Bt maize is
commercially grown, at 20% of the planted area. We assume this refuge area at the same time
allows farmers to comply with the European legislation on ex ante spatial coexistences of
production systems (see Devos et al., 2008). For simplicity we assume the farmer does not
control for WCR in the refuge area and therefore encounters some yield losses while having to
sell the maize as Bt maize. The price for Bt maize may differ from the IP maize. European
markets in GM crops are still premature but for now this IP premium did not develop in the
maize market (Gómez-Barbero et al., 2007, Skevas et al., 2009).However, to be conservative and
to allow for the negative perceptions of consumers towards the Bt technology we introduce a
triangular PDF ranging for no IP premium to 3%.
Cultural control options are evaluated based on the alternative cultivation options. As the
gross margin of other crops is highly uncertain at the time of decision making the cost connected
to this control option is also uncertain. AKII (2008b) documents gross margins of commodities
28
with their spatial and temporal variation. Using the standard methodology of the PERT PDF we
estimated the appropriate distributions based on the most likely value and the extremes. As winter
wheat and barley are the main crops rotated with maize, we use this rotation scheme for the
calculations, however local variations may exist. We believe the broad variance introduced is
enough to cover these local differences in crop rotation. Important in the assessment of crop
rotation as a control measure is the introduction of a land constrained farm. Crop rotation only
induces a cost to the farmer if the output mix from the farm is changed by the introduction of
crop rotation. As a consequence, farmers cultivating continuous maize before the infestation by
WCR may be able to change the location of the maize plot each year without any cost if his
arable area permits it. However, other farmers may to have to reduce their maize output as not
enough land is available to accommodate his intended production system. We call these farmers
land constrained. Land constrained farmers often be characterized by a small cultivated area or
the presence of livestock as these farmers need the maize feed. This is in line with the finding in
the US that small farmers lack flexibility and apply more control options against WCR (Wilson,
2005).
Simulations and results
Before the actual calculations can be done to obtain insight in the competitiveness of the
different control options, a primary simulation has to been done to determine the technology
valuation f(x) in order to endogenize θ. This can be done by calculating the benefits for farmers
from the adoption of Bt maize at a zero price level. By assuming the farmers are rational profit
maximizers this yields their theoretic willingness to pay for the variety on which the innovator
will base his corporate pricing strategy (Dillen et al., 2008). The resulting PDF from a Monte
Carlo simulation of 10 000 iterations in Excel with the @RISK add-in by Palisade Corporation
29
can be seen in Figure 55. While the average valuation for the Bt maize amounts to €37.3/ha, there
is 3% of the population with a negative valuation for the technology. These farmers would not
even adopt if the technology were freely available. This means the yield protection is not high
enough to cover the cost from the refuge area and the IP premium. For a population of CARA
farmers the technology valuation is a little higher as can be seen in Figure 5. At first sight this
might seem strange as the certainty equivalent (CE) is calculated by a deduction of the risk
premium. However, the technology valuation is a function of the CE for the other control options.
As these control options seem to have a higher variance in their potential profits, the surplus
value for Bt maize increases. The sensitivity analysis presented later will shed further light on
this effect. To use equation 5 we need to parameterize the PDF. Using the Kolmogorov-Smirnov
test, the (shifted) Weibull PDF seems to have the closed fit to the resulting distribution (Figure
6). The Weibull PDF is common in estimating continuous data as it is very flexible and can take
symmetric or asymmetric shapes (see review by Kim & Yum, 2008). Following equation 4 and 5
we calculate a technology fee of 32€/ha for a risk neutral population and €33/ha for CARA
farmers. Following the same approach for silage maize leads to a higher technology fees of
38€/ha under both risk behaviors. Dillen et al. (2008) show that a differential price in submarkets
is ideal for the innovator as third degree price discrimination leads to higher revenues. However,
in the remainder of the study we assume no price discrimination between silage and grain
varieties. The choice made between the two at planting time can still be reviewed if exogenous
factors favor the one of the other system (e.g. cob diseases, price changes…) as contracted silage
production is low. Moreover the area under silage maize is small and will therefore only have a
limited influence on the corporate pricing strategy. We opt for an area weighted average
5 The figure shows the valuation for grain maize producers. The same procedure can be followed to determine
the value for silage grain producers.
30
technology fee leading to a €33/ha for risk neutral and CARA adopters. To assess the effect of
the technology fee on the created rents we introduce the technology fee in the model as a
triangular PDF (see table Table 2). Our estimates of the technology fee lies within the range of
magnitude found in other papers. Alston et al. (2002) price the technology competitive to soil
insecticides, at about $30/ha. Fall & Wesseler (2007b) design different scenario’s with a
technology fee ranging from €27.65/ha till €102.7/ha.
With all the inputs known or estimated, the competitiveness of the different control options
can be assessed. Again a Monte Carlo simulation of 10 000 iterations is run and the results are
shown in Table 3. Looking at the individual profits of the different control options gives an
insight in the expected turnout for farmers if they apply the considered control option. First have
a look at the applied control options in Hungary for a population of risk neutral farmers. Soil
insecticides creates an average benefit of €6.0/ha. Compared to the other control options this is
low but due to the positive skew of the distribution this means 47% of the cases yields a
significant positive benefit while the 53% of negative cases is of a rather low value. Seed
treatment offers a higher average profit for farmers of €29.8/ha with a negative outcome in 24%
of the cases. In cases were crop rotation is feasible because farmers are not land constrained if
offers a high average profit for farmers of almost €48/ha with a negative value in 23% of the
cases. Finally we assess the effect of the hypothetical adoption of Bt maize on the farmers profit.
On average this would yield an extra profit of €63/ha compared to no control, with a potential
loss to Hungarian farmers of 3%. However, the size of the average profit does not mean this
control option is optimal under all given situations. Therefore, for each iteration in the model we
determine the best control option. Bt maize offers the best control in 77% of the cases with crop
rotation (22%) and seed treatment (1%) in the remainder of the cases. In the absence of Bt maize
31
this result is somewhat different with 10% of the cases for soil insecticides, 60% crop rotation
and 30% seed treatment. In forced monoculture no crop rotation is feasible and the model shows
that in this case seed treatment would be optimal in 60% of the cases and soil insecticide in 40%
of the cases. This is in contrast with the survey results from the actual control in Hungarian maize
indicating that 59% of the farmers choose soil insecticide. This hinges farmers in Hungary are
turning to the more expensive soil insecticides to often as their only goal is to protect yield. As Bt
maize seems to be the most value control option available we can calculated the loss to farmers of
not having the possibility of cultivating GM crops. In the table we see this extra value for land
constrained farmers. The adoption of Bt maize brings an added rent of €18/ha from adopting the
technology. In the case the farmer is forced to cultivate maize in monoculture this extra profit
amounts to €34/ha as the possibility of controlling the invasive species with crop rotation is
absent.
For CARA farmers the valuation for the specific technology decreases due to the
incorporation of a risk premium in the resulting CE. The coefficient of variation of the different
resulting profits gives some insight in the variability of the output. It seems Bt maize, due to its
high efficiency in controlling the pest, has the least variable result for the farmer while soil
insecticides bears the highest variability. As this variability determines the height of the risk
premium a simple results is that Bt maize will be favored by CARA farmers. The ranking of
different control options does not change on average but the option chosen may differ for CARA
farmers. The resulting best options can be found in the last column of Table 3. Bt maize is now
the optimal control in about 80% of the cases, crop rotation in 18% and seed treatment in 2%.
Although the CE is lower, the extra value created by adopting Bt maize remains about the same.
The CE of all control options can be found in Table 3.
32
The results for silage maize follow the same line although all benefits are somewhat lower
due to the different price and yield levels. As the technology fee is mainly targeted on the grain
maize producers, it is perceived by silage maize producers as being low and therefore the
potential adoption of Bt maize is higher with Bt maize as the best control option in 93% of the
cases.
The average results presented give an indication on the competitiveness of different control
options within the whole population of maize farmers in Hungary. However, the full results of the
model incorporate the aggregate risk and heterogeneity surrounding the parameters in the model
and therefore create PDF’s of the rents from different control options. Dillen, Demont & Tollens
(2008) argue that the heterogeneity in technology valuation among farmers has both a temporal
and a spatial dimension. The temporal dimension contains constraints like weather, pest pressure
and price volatility while the spatial dimension includes social capital, soil, water availability,
machinery and much more. The former is highly stochastic and difficult to assess for the
individual decision maker. The latter however is at least partly observable at the time the
adoption decision has to be taken. An individual farmer has some private information on how he
differs from the average which he can exploit. Therefore it is useful to determine which are the
determining factors favoring one control option. Through insight in these factors a multi-criteria
decision framework can be developed to optimize damage abatement strategies and reduce the
dependence on scouting and anticipated yield loss.
The determinants of the rents for a certain control option are straightforward as they are clear
from the model structure. In the second column of Table 4 the results for the Bt maize rents are
presented as an example. High potential yields and high prices augment the value of the
technology because the higher value of the percentage damage in the counterfactual no control
33
option. The technology fee has a significant negative influence on the rents while higher potential
damage increases the value of protection. As mentioned before, the average value of a technology
give an indication on whether the control option is the most suitable as there is competition of
other control options. Therefore it is of higher interest to determine what factors influence the
difference in rents between two control options. In column three of Table 4 the determinants of
the difference between Bt technology and the next best solution, is presented. It seems a high
potential yield decreases the advantage for Bt technology. As yields are higher, the potential loss
is higher, both in refuge area and in the GM field. As the next best control option is crop rotation
in a majority of cases, which does not suffer from yield losses and is therefore not directly
affected by increases in yield, the competitiveness of Bt maize decreases. Indeed, if we look at
the result for those farmers without the ability to engage in crop rotation, then the position of Bt
maize significantly augments with increasing yields. It is important to separate this effect of the
gross margin as this is not correlated to the yield in our model. An increase in the gross margin
leads to an increase for Bt maize competitiveness as crop rotation becomes more costly. As the
price of maize also affects the value of the yield loss the same reasoning can be developed and
has a significant negative effect.
As the cost for competing technologies increases, Bt maize becomes more profitable as could
be expected. The effect is bigger for the seed treatment as it is more variable in output and
therefore extra investment in the technology may have significant lower turnouts. The effect of
price increases is even stronger if crop rotation is not feasible. If the consistency of the chemical
treatment is on the low side, Bt maize will gain competitiveness. This effect is even stronger
under continuous corn as the competition with crop rotation disappears. The cost of crop rotation
is influenced by the gross margin on competing crops. As expected the there is a significant
34
negative effect on the competitiveness of Bt maize if a farmer can generate higher than average
benefits on other crops, in this case winter wheat and barley. An increase in the price for the Bt
technology through the technology fee would decrease the profitability (and the adoption) of the
technology. The application cost and the IP premium have no significant impact on the
competitiveness as there effect is variable.
Finally we discuss the classic decision factor damage. The second column of Table 4 shows
the anticipated effect of uncontrolled damage on the rent created by Bt maize. As potential
damage is bigger, potential damage abatement is bigger and therefore the rent increases
significantly. However in relation with other control options the effect is less straightforward. As
mentioned before crop rotation is not affected by the size of the potential damage and therefore a
higher potential damage erodes the preference for Bt maize. If monoculture is favorable on the
other hand Bt maize increases its competitiveness as the lower efficacity of the chemical control
options increases the yield loss.
The analysis shows that several parameters play a key role in the adoption decision besides
yield protection. Therefore a farmer should exploit his private information to locate himself in the
presented distributions. As an example let us have a look at a farmer with a land constraint but
not confined to continuous maize. With only aggregate information available Bt maize would be
optimal in 80% of the cases. However this farmer is aware of the fact that the plot he intends to
assign to maize has a high infestation of WCR due to previous cultivation with maize and
therefore a higher anticipated damage, this percentage reduces and crop rotation might be
favored. This procedure is what we called the multi-criteria decision making before.
35
This sensitivity analysis also allows predicting the competitiveness of different control
options if structural changes take place in the future, or some of our data assumptions, for
example on damage, proof to be wrong. Let us assume prices go up in the near future due to the
increased demand of coarse grains. On monoculture the position of Bt maize would increase will
it would decrease on the area of land constraint farmers. As the model structure does not change,
this sensitivity analysis can be extrapolated to the other countries under research. Only the size of
the regression coefficients and as such the extent to which the competitiveness is affected differs.
36
Farmers willingness to pay for Bt maize
Besides the direct monetary gains as accounted for in the model presented earlier, WCR
resistant Bt maize has a set of non-pecuniary benefits. Non-pecuniary benefits are those benefits
that are not traded in the market and therefore do not have a direct monetary value. Among these
benefits for WCR resistant maize breeds are; increased farmer health from reduced handling of
chemicals, safer for the environment, reduced management and scouting activity (time savings);
certainty of protection (reduced risk), easier application of the protection,… (Alston et al., 2002).
Several papers assessed these benefits in different GM crops via different approaches, contingent
valuation as the most prevalent (i.e.Alston et al., 2002; Marra & Piggott, 2006; Marra, Piggott, &
Carlson, 2004). Although these benefits are not represented in the standard cost benefit analysis,
they influence the adoption outcome significantly by increasing the technology valuation among
adopters. Because most of these non-pecuniary benefits are credence or experience based, the
question rises to which extent they are accounted for in the initial adoption phase. This will
depend on the previous knowledge and beliefs of the farmer about both the GM technology and
the information he has on the pest and competing technologies. Papers by Krishna & Qaim
(2007) and Marra, Hubbel & Carlson (2001) demonstrate the importance of information in the
valuation and adoption of Bt technology. Pigott & Marra (2008) show the effects non-pecuniary
factors embodied in new technologies have on the demand for these technologies over time. In a
first phase the demand shifts out due to the pecuniary benefits and in a second phase demand
becomes more price inelastic if the technology becomes known and the non-pecuniary benefits
valued more. This means disadoption will decrease in time and price changes for the technology
will have less effect on the adoption decision.
37
For WCR control in the countries under research, we assume these non-pecuniary benefits do
not play a role in the initial adoption phase. The maize sector in Hungary is highly competitive
and profit driven and therefore monetary values are the key decision factor in the control option
choice (see also results of survey). In some countries where this assumption might not hold this
aspect is discussed. A second reason for not incorporating the benefits in the model is the limited
knowledge about the pest by farmers. This limited knowledge is demonstrated in Figure 2 as the
perceived presence of WCR in Hungarian fields compared to the official data on infestation. The
uncertainty about yield combined with the limited knowledge results in a big range surrounding
the perceived damage of WCR by farmers both with and without control. This lack of
information is even bigger in those countries not yet having economic damage. We therefore
assume the non-pecuniary benefits only play a role in the prevention of disadoption of GM crops
induced by exogenous factors (i.e. technology fee, price fluctuations)
In order to get some insight in the non-pecuniary benefits we elicit the potential willingness to
pay (WTP) among Hungarian farmer for WCR damage resistant varieties. We used a
dichotomous model as this is generally superior to an open-ended estimation as it creates a more
market-like situation (Bateman et al., 2002). Respondents were asked to evaluate the premium
they were willing to pay compared to their actual conventional hybrid. Prices were randomly
varied from 1000 till 10000 HUF/ha based on the actual price of maize and the findings in
previous valuation studies of Bt maize. Conditional on whether the respondent was willing to pay
the proposed value, a second value was given being 30% higher if the initial answer was “yes”,
30% lower if “No”. Hanemann, Loomis, & Kanninen (1991) provide both theoretical and
empirical arguments as why such a dichotomous bounded choice model is statistically more
efficient than a single bounded approach. In order to get realistic WTP measures we started the
38
question with a clear description of each attribute of the damage resistant crop, especially the
more difficult to understand non-pecuniary benefits. We presented the variety as a hybrid
resistant to WCR damage but specified this could also be through conventional breeding. This
setup may increase some of the perceived benefits (especially towards the environmental
benefits) but we believe it gives more reliable estimates in the hostile Hungarian climate
surrounding agricultural biotechnology and to avoid strategic answers. As a consequence we did
not describe the potential IP premium in the set of attributes thereby. However, a price decrease
would be gradual, following the gradual adoption. Furthermore, due to the increase of demand for
maize in the future, this price decrease might not be visible in absolute terms.
We use the model initially developed by Cameron & Quiggin (1994). If we define y1j=1 if the
response to the first question was “yes” and o if “no”, y2j coding for the second answer, with d1j=
2 y1j -1, d2j= 2 y2j -1, ti the value of the bid, µ i the mean, σi the standard deviation and εi the error
terms, the jth contribution to the bivariate probit likelihood function becomes
1 21 2
1 2 1 2 1 21 2
( ) ( ( ), ( ), )j j j j j
t tL t d d d dε ε
µ µµ ρ
σ σ
− −= Φ (14)
This model is very general and flexible by allowing for the possibility of different means,
dispersion, and non perfect correlation across the two responses. We used a Wald test to assess
whether µ1= µ2 and σ1= σ2. Independent from which parameters are introduced in the model we
can reject the hypothesis. This can be explained by the fact that respondents may consult a
different set of preferences for the follow-up question (Haab & McConnel, 2002). A respondent
may feel exploited if the price goes up in the second question, he may perceive the value of the
product as decreased if the price goes down or he answers strategically on his expectation of a
39
third question. Although the preference set can be different between the two questions, Haab &
McConnel (2002) argue that the bivariate probit model still offers higher efficiency if the
correlation between the two independent probits significantly differs from zero. As this is the
case in our data we choose to adopt a bivariate probit model based on the first equation
parameters as the preference set in the first bid can be assumed to be the original unaltered set. In
theory this could lead to a starting point bias: responses to the second question depend on the
price offered in the first question but as the starting price is random this effect will be limited
(Herris & Shogren, 1996). The mean willingness to pay is estimated by evaluating at variable
mean values. Because of the high variability stemming from different sources in the whole survey
population the WTP analysis only takes into account those farmers indicating they detected WCR
in their field before. We assume these farmers have a better understanding of the pest’s properties
and will therefore have a more reliable valuation for the resistant hybrid variety.
The results of the first equation in the estimated bivariate probit can be seen in Table 5. Since
we introduce the new variety as a hybrid, this may decrease the value of the variety for some
farmers as they are used to farm saved seed. However, in Hungary the use of hybrid seeds is
common and therefore this characteristic is unlikely to influence the bidding process. Of the
decision parameters in the choice for a control option only equipment has as significant positive
coefficient. Farmers who choose the control option according to the available machinery value
the new variety higher. This can be explained by the fact that soil insecticides and certainly aerial
application need specialized machinery or contracted services which are not needed with the
systemic protection. This is more likely to happen in non specialized smaller farms. However, the
insignificant coefficients on the percentage of off-farm income and maize area cannot confirm (or
reject) this hypothesis. It is worthwhile to note the big coefficient on the latter variable. Although
40
not significant, the percentage of maize will have an important increasing effect on the mean
WTP. The location of the farm seems to matter to a high extent. Farmers in the Great Plain are
willing to pay significantly more than farmers in either the North or the Danube valley. In this
region monoculture is more present than in other regions which limits the opportunity to control
the WCR with a rotation strategy. Moreover, as this is the main maize producing area with the
highest gross margins, the cost of crop rotation will be high (AKII, 2008b). Consistent with this
result is the significant positive coefficient for the percentage of maize in monoculture, leading to
a higher valuation. These findings are in line with the results of the economic model presented
before. A logic result from the model is the fact that farmers anticipating problems with WCR in
the future, not trusting the performance of their actual control option, value the new technology
significantly higher. The coefficients for both grain and silage prices have a positive sign. At first
sight, this is not consistent with the prediction of the economic model saying this only increases
the value under monoculture. However, in the economic model yields are not correlated with the
gross margin as this is introduced separately in the model. In the bivariate probit model we can
assume there is positive correlation between yields and gross margin and therefore higher prices
increase the WTP. The significant negative effect by the expected yields in 2008 is explained by
the fact that these farmers (by selection having a population of WCR on their field) manage the
pest well with the present technology and are therefore less willing to pay for a new technology.
Farmers with lower yields, partially explained by the WCR damage, are willing to pay more for a
protecting technology.
An interesting result is the significant and large effect of the experience of the farmer with
Integrated Pest Management (IPM). IPM is an environmentally sensitive approach to pest
management. IPM programs use information on the life cycles of pests and their interaction with
41
the environment. This information, in combination with available pest control methods, is used to
manage pest damage by the most economical means, and with the least possible hazard to people,
property, and the environment. The positive effect means that farmers having an interest in a
sustainable cultivation method value the variety higher. Therefore, the non-pecuniary
characteristics of the new variety are perceived as high value by the farmer having the
characteristics of the IPM in his preference set. The demographic variables in the model do not
produce significant coefficients. This could be induced by two reasons. All farmers might have
the same information on WCR and control options, or information is scattered and based on
personal experiences not captured by the demographic variables. The latter option is not
convincing in the setting of Hungary with highly educated farmers (70% of the sample has a
higher education training of which 70% a university degree in agriculture) and information
sourcing. The survey shows that (Table 6) farmers stating own experience as their major
information source regarding insect control is limited to 9.1% of the population. With 49%,
almost half of the farmers state an independent source (advisors, agricultural magazines, lectures
and professional literature) as the determining source of information. Private companies are only
in 16% of the cases the major source. Therefore we can assume the extension services are
efficient and well used and information is uniform among farmers (although not necessarily
correct).
Finally, the specified bivariate probit model leads to an average WTP of €70.3/ha. However,
by looking at the confidence intervals of the coefficients (Table 7) calculated according to the
Delta method (Oehlert, 1992) we see a highly heterogeneous technology valuation. This stems
from the high perceived variability of the pest damage and the effect of the available control
options. We therefore tried to use the bootstrap methodology introduced by Krinsky & Robb
42
(1986) in order to construct a confidence interval for the average WTP but failed to reach
convergence. The average technology valuation based on the economic model gave an area
weighted technology valuation of €58/ha for a risk neutral population. This means the excess
value of €12/ha can be considered as a premium for all the non-pecuniary aspects present in the
technology bundle. Alston et al. (2002) in their assessment get to a lower estimation of 12$/ha in
the USA setting. The only study the authors know of indicating a measure of WTP for non-
pecuniary benefits of insect control options in Hungary (or neighboring countries) reports values
€59-99/ha (Demont, Tollens, & Fogarasi, 2005).
From the presented regional farm level analysis we learned that with an investment of €33/ha,
the Hungarian grain farmer gains an average direct monetary gain compared to no control of
€63/ha. This means the investment yields a benefit of about 200%. Chemical protection other
hand has a much lower return on investment lower than 100%. Compared to the other available
control measures the extra rent for Bt maize amounts to €18/ha for land constraint farmers and
€34/ha for farmers with continuous corn which can be complemented with €12/ha to incorporate
the non-pecuniary benefits. From these results the value of Bt maize for Hungarian farmers is
clear. What this means on a national level is assessed in the next section.
43
Aggregation to the national level
According to the previous farm level analysis, Bt maize is a high value control option for
Hungarian farmers. In order to assess the social surplus created by the technology the technology
induced supply shift needs to be estimated. In a standard market model, three factors influence
this shift: the yield boost compared to the next best technology, β, induces a horizontal shift,
while the per hectare cost effect, γ, acts as a vertical shifter and finally the adoption rate, ρ, is
used as a scaling factor. As Oehmke & Crawford (2002) show, the commonly used Alston-
Norton-Pardey approach to model the shift is very sensitive to the supply elasticity and therefore,
does not give very robust results. Therefore we opt for the approach taken by Qaim (2003) which
uses the change in total factor productivity as a shifter which according to Demont, Oehmke, &
Tollens (2006), reduces to the following K-shift:
1
Kγ β
ρβ
+= +
(15)
Next we model the technology as a spill-in into Hungary. This seems consistent as the
technology is already introduced in other parts of the world and the decision whether or not to
allow the Bt variety is taken at the European level which does not leave a rent for first movers. In
a global market, the production volume of Hungary can be considered as small which suggest
modeling the Hungarian maize sector as a small open economy. This assumption leads to an
infinite elastic demand curve and the change in producer surplus as (Alston, Norton, & Pardey,
1995, p.227):
(1 0.5 )basePS py LK Kε∆ = + (16)
44
with ε the supply elasticity (0.2819 from the ESIM model (Nolte, 2007)) and L the area under
maize cultivation. The revenue extracted by the seed suppliers is calculated through:
LθρΠ = (17)
It is important to note that Π does not include the extra cost of marketing, production or
distribution costs and has to be shared among all stakeholders in the seed supply.
To calculate the total welfare effects, the consumer effect should be included. As we introduced
an IP premium in the model, buyers that are indifferent between GM maize and conventional
maize could extract a consumer surplus from the reduction in prices. However, this is not within
the scope of the study as such an assessment should incorporate horizontal differentiated supply
chains, consumer perceptions, labeling and so on.
The variables needed to assess the welfare effects can be extracted from the farm level analysis
with the exception of the adoption rate ρ. In order to estimate the adoption rate we would need to
know how much farmers do indeed have a land constraint, which is not available. Some
indication however can be used. From the survey it is known that 52% of the farmers reduced
their area under monoculture since the infestation by WCR. Of these farmers, 10% does not
recognize the reduction as a control option and therefore we assume these farmers have no land
constrain. The actual use of soil insecticides and seed treatment does not give substantial
guidance as these technologies are inferior and therefore don’t reach the same adoption rate as Bt
maize. From the fact that in 2007 200 000ha was treated with soil insecticides and another 70 000
ha with seed treatment, we can only conclude that the minimal adoption rate for rational farmers
is 23%. Based on this scarce information we assume and adoption rate between 30% and 50%
with an average adoption equal to the amount of monoculture in Hungary (ρ~triangular
45
(0.3;0.4;0.5)). As the high yielding grain maize producers are most likely to produce on contract,
and therefore are able to reduce risk and uncertainty, they will be the most likely to adopt.
The results of this preliminary analysis show a gross revenue accruing to the seed industry of
€15.8 million on average annually with a coefficient of variation 0.16. The farmers on the other
hand generate a surplus compared to the technology not being available of €17.1 million as an
annual average. The spread around this average is very big as it is an outcome incorporating all
the uncertainty and assumptions of both the aggregated and the farm level analysis. The
coefficient of variation amounts to 2.3 because of this. When more detailed information on the
parameters in the model becomes available in the future, both the uncertainty for the assessment
as the uncertainty for the farmer himself is reduced and estimation will become more robust.
Interesting to note is the fact that despite the high uncertainty, the results follow the rule of thumb
regarding the societal diffusion of welfare created by biotechnology innovation in agriculture as
described in Demont et al. (2007). About 1/3 of the benefits accrues to the upstream sector with
37% while the rest accrues to the downstream sector.
46
Control of WCR in Czech Republic
Maize production in the Czech Republic is very different from Hungary (Table 8). The area
sown with grain maize is small compared to the area under silage, which represents 71% of the
sown area. Maize is with a total area of 287 900ha the fourth crop in the country (13% of the
arable land) after soft wheat, barley and rape seed, the latter is gaining importance the last few
years. The share of silage and grain maize has been changing the last few years. Grain maize is
gaining importance as the amount of livestock held is decreasing and the traded share of
production increases. This trend towards increase in grain maize (at the expense of other cereals)
is expected to continue in the next years due to climatic change favoring its production (Trnka et
al., 2007). At present, grain maize is grown in all regions (kraj) of the Czech Republic, but
production is merely concentrated in two regions (Central Czech and South-Moravia),
i.e.Stredoceský and Jihomoravský, where roughly 50% of total grain maize acreage is cultivated.
Another 25% is grown in the South-East, i.e. Olomoucký, Zlínský, Moravskoslezský and
Pardubický (Figure 8).
The major pest in Czech maize production is the European corn borer (ECB). The borer
penetrates the stalk and excavates large tunnels that result in important yield losses. Although the
pest has a long term tradition in the Czech Republic with its first detection in hop in the 19th
century, it took up to 2001 to be found in all regions where maize is grown. According to data
available from the State Phytosanitary Administration (State Phytosanitary Administration, 2009)
average use of insecticides of €1.8/ha to control for ECB is negligable compared to the total
expenditure on crop protection in silage (€59/ha) and grain maize (€75/ha), which mainly exists
out of herbicides. However, as will be argued later, the technology used against ECB could offer
47
possibilities in the protection of WCR, especially since the adult control measures work on both
insects and seed treatment is similar (Muska, 2008).
WCR was introduced into the Czech Republic in 2002 via the south eastern border.
Pheromone traps reveal an annual increase of the WCR population. In Figure 7 the steady spread
of WCR through the country up to 2008 can be seen. In 2008 the beetle spread through all the
major maize producing areas. As the highest density of maize cultivation coincides with the
oldest population of WCR in the Czech Republic in South Moravia, this area is under the most
acute danger of economic loss which has not been detected up to now. We assess the control
options under a scenario where economic damages occur.
The regional farm level model presented in an earlier section is applied on the Czech setting.
The necessary data to feed the model, prices, gross margins and yields, can be found in Table 23.
The price of chemicals is difficult to assess as the main soil insecticide (tefluthrin) is not
registered for use in maize production in the Czech Republic, so we rely on an expert opinion in
Fall & Wesseler (2007b) stating a price of around €80/ha. Although some seed treatments are
available in the market (i.e. thiamethoxam which costs around €37/ha) others as clothianidin are
not yet approved for use. Therefore the price of seed treatment is assumed to be similar to the
price of the seed treatment in Hungary.
As has been expound before, running the model with a technology fee of zero allows the
construction of a PDF at to assess the technology valuation for Bt maize in the Czech Republic.
The results seem to indicate that the valuation of the Bt technology on its own is about as high as
in Hungary. However, as the efficiency of crop rotation increased, the extra valuation is lower
than in Hungary. This effect is explained by the lower gross margin for maize relative to the
48
other crops grown within the Czech Republic leading to a lower cost of rotating. This is of course
the same reason which makes maize less dominant in Czech agriculture. Therefore the average
technology valuation compared to crop rotation for Bt maize over all modeled cases is low with
€-4/ha and €-1/ha for grain maize for risk neutral and CARA farmers respectively. For silage
maize the values are somewhat higher €8/ha and €11/ha under the same situations. For farmers in
a monoculture setting the technology valuation is much higher. Grain maize in monoculture has a
technology valuation of €60/ha and €58/ha while for silage maize this would amount to €59/ha
and €58/ha. The seed industry will only target those cases where there is some rent for their
technology. Using the revenue maximizing pricing strategy and taking into account the area
planted with grain and silage maize, this leads to a technology fee of €17/ha. In order to capture
the uncertainty of the estimation and the preliminary data we model θ as a triangular PDF
(7;17;27).
The results for the Czech Republic are given in Table 9. In grain maize, all of the
technologies offer an average benefit compared no control; the costs outweigh the potential yield
loss. Crop rotation offers the highest value with €107/ha while soil insecticides on average have
the lowest performance, €17/ha, due to the high cost of the product compared to the potential
damage. Detailed analysis of the PDF shows that despite the low average benefit soil insecticides
both create a positive rent in 66% of the cases. By constructing the CE, the simulation model
shows that for CARA farmers Bt maize is the optimal control strategy in 50% of the cases,
closely followed by crop rotation and a very small application window of 0.3% for seed treatment
in land constraint farmers. From this it is clear that the welfare created from the introduction of
Bt maize in Czech Republic will be considerably smaller than for the previous case study of
Hungary. In fact, the average welfare for a farmer with no private information to exploit the extra
49
value would be negative (€-10/ha). To get some insight in the potential however we tabulate the
maximal rents achievable from the Bt technology, assuming full information and perfect rational
adoption. Therefore it differs from the benefits given in table Table 3 which represent the benefits
for the total population given all the uncertainty surrounding the decision. Even with this
specification the extra value created by adoption of Bt maize is limited at €11-15/ha depending
on which scenario one considers. Under monoculture are calculated with the traditional approach
of incorporating uncertainty as the competition from crop rotation resides and the created welfare
is clear. Under continuous grain maize farmers from €48 till €46/ha for risk neutral and CARA
farmers respectively.
In silage maize, the largest share of the area sown with maize the benefit from applying
different control options is somewhat lower as the value of the potential yield loss is smaller in
this market. The order of competitiveness remains the same, but the average return from applying
soil insecticides in silage is now even negative with only in 31% of the cases a positive outcome.
CARA farmers crop rotation increases its competitiveness compared to Bt maize as it would be
the optimal control strategy in 55% of the cases compared to the 50% in the grain maize market.
With the same assumptions as made in the alinea above, the benefit for farmer from introducing
Bt maize in monoculture amounts to €47/ha.
Assessing the social welfare effects is highly influenced by the assessed adoption rate. We
saw that with perfect information about half of the land constraint farmers would opt for the Bt
maize variety. Once the pest creates damages in the Czech Republic in the magnitude we
assumed, we assess the potential adoption at 5% in the most land constraint and maize profitable
grain maize area and about 18% in the silage maize area. These figures include the 10% of maize
that is now planted in monoculture around the Czech Republic, mainly by those farmers having
50
significant livestock. To have some variability in the model we choose to model the adoption as a
triangular PDF with for grain (0.0;0.05;0.1) and for silage (0.10;0.20;0.25). Because of the very
preliminary data and uncertain data in the Czech Republic we chose not to use the adapted
Alston-Norton and Pardey formula for the Czech Republic (and the remaining countries in this
study) but the change in revenue method. We therefore keep the uncertainty as small as possible
by not introducing a supply elasticity.
R p y C∆ = ∆ + ∆ (18)
This methodology assumes that the changes in returns to land are equal to the change in
revenue less the change in cost. The change in producer surplus can be calculated as
PS RLρ∆ = ∆ (19)
while the change in innovator’s surplus remains equal to equation 17. The results show a
benefit for grain maize of €66 640 (σ=55232) annually and €607 755 (σ=312000) for the silage
maize producers. The seed industry is able to extract on average 48% of the generated welfare or
€652 715.
The results show that once the pest pressure in Czech Republic is as high as in Hungary,
causing the same potential losses, crop protection offers great benefits in most of the Czech
Republic due to the high gross margin of other crops within the region. In monoculture however,
Bt maize offers high benefits to farmers as it is much more efficient than the available chemical
protection options. If in the future due to market powers, climate change or other reasons (see
(Trnka et al., 2007)) the demand or gross margin for maize and in particular grain maize would
increase, we know from the sensitivity analysis of the model that Bt maize would have a bigger
51
group of potential adopters. Therefore the model should be adjusted once these changes take
place.
52
Control of WCR in Slovakia
Maize is the third commodity grown with 150 000ha grain maize and 80 000ha silage maize
(see Table 10) preceded by winter wheat and barley. This leads to about 13% of the arable land
under maize production. According to the Slovakian ministry of agriculture about 51% of the
arable land is denoted as being suited for maize production. Silage maize is mainly used for on
farm purposes. Slovakia has about 1700 farms with more than 20 heads of cattle of which only
20% use only grass silage. This means the other farms rely on silage maize as fodder crop. The
average share of silage maize in the fodder production in the two year period 2006-2007 was
88%. The main grain maize producing region of Slovakia is located in the Southern parts,
bordering Hungary. Within this region, maize cultivation is intensive, often under irrigation
practices. Because of this spatial aggregation a significant part of maize cultivation is under
monoculture, despite the larger area suited for maize production. This area fluctuates annually but
over the four year period 2004-2007, 16% of the maize was cultivated under monoculture (Table
10). If maize is rotated, the rotation scheme depends on soil conditions, pest pressures and the
availability of irrigation. As a standard practice wheat is rotated with maize and on a regular base
the introduction of legumes or alfalfa.
The WCR was first detected in Slovakia in August 2000. The pest had been reported from
three districts bordering Hungary: Lucenec, Komárno and Vel'ký Krtíš. In 2003 it already
reached the districts of Bánovce nad Bebravou, Topoľčany, Senica, Skalica, Michalovce,
Trebišov and Vranov nad Topľou (see Figure 9). The first economic damage occurred in 2004 on
only about 340ha despite high population of beetles (significantly above the economic threshold)
in a much larger area. In 2005 7% of the maize area had an economic population of beetles and
2.66% (6419ha) of the area had economic larval damage. The area with economic damage kept
53
increasing and despite the control options applied, in 2007 economic larval damage occurred on
4.86% of the area (Cagan, 2008). In Figure 10 the results of a capturing plan with yellow traps
are shown. The results indicate an established adult population in 5.97% (14391ha) of the maize
fields. If we look at the spread we see that indeed the south western part of the country has the
highest populations, which coincides with the area where monoculture is dominant but that the
beetle is found all over the arable lowlands. The fact that the pest can have a drastic impact on
the profitability and yield of maize production is confirmed by the average yield loss in
monoculture as indicated by Cagan (2008) of 30% in 2006 and 60% in 2007.
Different control options are available in Slovakia. In 2007 one soil insecticide, tefluthrin,
was registered for use. However, the law allows the application of unregistered applications if
local and controlled for max 120 days when no possibility to control with other tools for selected
pesticides. These include some foliar application and seed treatments. Despite these chemical
options a lot of farmers apply crop rotation as a control measure. To indicate the importance of
this practice Prof Cagan estimates the losses due to crop rotation at €1 000 000 in 2008.
In order to assess the technology valuation we apply the simulation model described before.
The input data can be found in Table 23. The simulation shows that crop rotation is a non
competitive control option in Slovakia as the average rent, both in grain and silage maize, is
negative on average and only positive in 5% of the cases. This means that for those farmers
cultivating maize is very costly due to the high gross margin they generate from the crop and not
applying any control at all would be often more rational than crop rotation. Of course as
mentioned before, with the necessary private information crop rotation could become rational for
some farmers. The low suitability of crop rotation as a control strategy increases the demand and
54
technology valuation for Bt maize in Slovakia. This increases the technology fee stemming from
the optimal corporate pricing strategy to €49.5 on average.
The rents created under the different control options in Slovakia as an average for the whole
population are given in Table 11. Bt maize creates an average rent of €36/ha and €30/ha for risk
neutral and CARA grain maize farmers and is therefore the most profitable technology. This is
reflected by the number of cases in which Bt maize, despite the rather high technology fee, is the
best option, being 90%. The second technology, the best option in 10% of the cases in grain
maize, is the seed treatment. Crop rotation and soil insecticides are never the preferred option
although for risk neutral farmers soil insecticides have a positive revenue in 40% of the cases
leading to an average rent of €-5/ha. This means in the absence of Bt maize soil insecticides
would be applied but with the availability of the Bt technology this is preferred in those cases.
Crop rotation is highly inefficient for most cases with a technology induced rent of €-75/ha and
€-85/ha for risk neutral and CARA farmers respectively. The extra value created from
introducing Bt maize in Slovakia amounts to about €20/ha on average. As crop rotation is never a
rational choice, the value is the same for monoculture areas and is therefore not separately
presented in Table 11. The necessity for chemical or GM control measures in the Slovakian
settings is confirmed by the estimated use of chemical treatment in Slovakia the last few years:
20% in 2006, 30% in 2007 and 70% in 2008 (Cagan, 2008).
In silage maize the results are similar with the exception that only Bt maize offers an average
positive rent compared to no control. The next best technology, seed treatment offers a positive
rent in 37% of the simulated cases. However only in 3% of the cases it creates a higher rent than
Bt maize. The extra value created by Bt maize is a little lower than in the case of grain maize
55
with €16/ha and €17.6/ha because of the technology fee that is dominantly priced for the grain
maize market.
In order to assess the welfare effects of introducing Bt maize in Slovakia the crucial
parameter is the adoption rate as presented above. With full information more than 90% of the
land constrained farmers would adopt Bt maize. However, as imperfect information is
endogenous to the WCR pest due to the low correlation between pest population and damage and
the other exogenous parameters as yield and price, this adoption rate will not be reached. As
mentioned above, about 70% of the maize area now uses chemical control against WCR in
Slovakia. As our results show that Bt maize is superior to chemical control in most of the cases
we assume this area would be replaced with Bt maize and we therefore model adoption as a
triangular PDF~(60;70;80).
The total welfare created for maize farmers amounts to €3.0 million annually, divided
between grain maize farmers (€2.0 million) and silage maize growers (€0.9 million). The seed
sector extracts about €8 million. The share accruing to the innovators is much larger as in the
other countries stemming from the shape from the technology valuation which allows the
innovator to extract almost all the (pecuniary) benefits. For a discussion of the effect of the PDF
on the rent creation and distribution see Dillen, Demont & Tollens (2008).
As arbitrage between the Slovakian maize seed markets and the neighboring countries can not
be excluded the effects of introducing the Bt technology at the price rate of the Hungarian market
may be useful. The average profit from adopting Bt maize now amounts to €49/ha and €43/ha for
risk neutral and CARA farmers respectively. Moreover, Bt maize would be the best control
option in 98% of the cases. If we aggregate these new results to the national level, the welfare
56
captured by the maize farmers would grow to €4.8 million annually and the benefit for the seed
sector decrease to 5.3 million, leading to an almost equal distribution of the created welfare.
The results show that Bt maize offers a high induced rent due to the high efficiency of maize
compared to other crops in the south western part of the country. This high profitability of maize
is reflected in the high percentage of monoculture of 16% on a country level increasing to more
than 30% in the irrigated areas. Soil insecticides and seed treatment also offer protection against
WCR but the efficiency depends on the information available to the farmer at the time of the
decision making as they have a significant chance creating lower rents compared to no control.
57
Control of WCR in Austria
On a hectare base, grain maize is the third most important crop cultivated in Austria, preceded
by barley and soft wheat. The area planted with grain maize has been increasing steadily over the
last 5 years, leading to an all time high of 194 000ha in 2008. Besides the importance of grain
maize in Austrian agriculture there is also a considerable acreage of silage maize of 81 000ha in
2008 (Table 12). This means maize is cultivated on about 22% of the available arable land.
Yields are consistently higher in Austria than in the other countries discussed in this report. A
grain maize yield of 11.1t/ha (AGES, 2008) on average and even 12t/ha in the region of
Steiermark (Statistik Austria, 2008) indicates an intensive cultivation of maize. Indeed, because
of spatial constraints such as elevation, the cultivation of maize is predominantly concentrated in
the south eastern part of Steiermark bordering Slovenia and Hungary. Figure 11 represents the
maize density in the crop rotation on a community level. Three areas, indicated in black, have
more than 50% maize and are therefore classified as highly under risk of WCR damage by
Baufeld & Enzian (2005b). As the density is higher than 50% there needs to be at least some
continuous maize in this area. According to information from the Austrian Agency for Health and
Food Safety (AGES, 2008), the area under monoculture ranges from 0% in the low density areas
to 25% in the high density areas. Two types of farms are likely to have monoculture; small farms
(max 20ha) with animal husbandry and farms in eastern Austria using irrigation techniques. If
maize is planted in a rotation scheme AGES (2008) reports 50% is rotated with winter wheat and
in smaller amounts also with summer grain (barley and oats), durum wheat, peas and pumpkin.
WCR was found for the first time in Austria in the summer of 2002 near the towns of
Jahrndorf and Andau in Burgenland (near the borders with Hungary and Slovakia). It was later
captured at several locations in Burgenland and Niederösterreich. (EPPO, 2008). Figure 12 shows
58
the spread of WCR in Austria in 2008. The beetle is established in the intensive maize growing
region of Steiermark. It also spread towards the north into the area that is characterized by a low
maize concentration. The beetle has now about 65% of the Austrian maize area infested
(AGES,2008). Despite the presence of WCR in most maize fields no economic damage has yet
been detected. As no economic damage is yet recorded, control options against WCR are not
widespread in Austria. Although some soil insecticides (i.e. tefluthrin) are registered for use
within Austria the only control option used is clothianidin at a very small scale. However, with
the intensive maize growing practices and the established population, damage may only be a
matter of time.
Using the data presented in Table 23 and under the assumption of revenue maximizing
technology suppliers the technology fee for Austria would amount to €29.8/ha. Table 13 presents
the results of the stochastic model. It seems Bt maize is has the highest average technology
induced rent in both grain and silage maize with €112/ha and €99/ha for risk neutral farmers
respectively. Chemical control is on average preferred to crop rotation as a control measure as the
average benefit of crop rotation is negative. Soil insecticide offer a positive rent in 77% of the
modeled cases for risk neutral grain and maize producers, averaging €46/ha and €30/ha. Seed
treatment has a somewhat higher average with. The relative high value from soil insecticides
compared to the other countries under research can be explained by the high value of potential
damage stemming from the high yields combined with the high prices. As the potential loss if
significant, the high price of the soil insecticide bears less risk for the farmer (as can be seen from
the sensitivity analysis in Table 4). Seed treatment even creates positive rents for 88% of the
cases for risk neutral farmers with €62/ha and €47/ha for risk neutral grain and silage maize
farmers. Crop rotation has a negative average rent but nevertheless in the risk neutral case, it
59
offers positive rents in 34% of the cases. For CARA farmers the results are similar. The only
difference is the relative decline in the competitiveness of soil insecticides and crop rotation as
they have the highest coefficient of variation. Especially the results for crop rotation have a high
variance which is reflected by the fact that despite the high negative average value farmers would
still gain in 1/3 of the cases.
As argued before the average value is only one parameter in the decision making process.
First have a look at the actual situation where Bt maize is not allowed in the Austrian maize
sector. In that case seed treatment would be optimal in 74% of the cases, soil insecticides in 19%
of the cases and crop rotation in only 7% of the cases for grain maize under a land constraint.
From this we can conclude there is a demand by Austrian farmers for control options besides
crop rotation. Bt maize, due to the low price compared to soil insecticides and the higher
efficiency changes this pattern drastically. If Bt maize would be commercially available it would
be optimal in 98% of the modeled cases, seed treatment being optimal in the other cases. This
explains why the extra value created by introducing Bt maize in Austria to control for WCR
amounts to around €50/ha on average depending on the different sort of farmers and maize
cultivation. This value in monoculture is the same as crop rotation is never a preferred option for
the farmers if BT maize is available.
The impact of the different technologies on the aggregate country is of high interest for policy
makers and industry. Baufeld & Enzian (2005b) quantify the area under high risk of WCR
damage in different European countries. They analyze 86 communities and find that 48 108ha
under grain maize is under high risk and 6 236ha silage maize in 2001. This means 28% of the
grain maize and 8% of the silage maize are at risk of damage. If we translate this figures to the
area planted in 2008 this means about 54 000ha of grain maize and 6 500ha of silage maize
60
would be in the high risk area. If we take into account the optimal choice for the farmers model
the adoption rate in grain maize with a triangular PDF~(0.25;0.28;0.31) and for silage maize
~(0.05;0.08;0.13). Applying the CIR method discussed before yields an annual induced rent from
adopting Bt maize of €2.6 million for the grain maize farmers, €3.48 million for the silage
farmers and about €1.9 million for the seed industry. This leads to a distribution of benefits of
39% for the seed industry and 61% for the Austrian maize farmers. This analysis does not include
the consumer surplus created through the potential price decrease. The results show that Bt maize
offers a high value to Austrian agriculture creating €4.8 million annually distributed between
seed suppliers, gene developers and farmers.
61
Control of WCR in Poland
Grain maize is a relatively small commodity in Polish agriculture. Looking at the sown area
in 2007 of 262 000ha it is only the 9th crop (Table 14). The dominant crops are soft wheat, barley,
rye and triticale. According to data from Eurostat (2009) silage maize is grown on another
368 000ha making the area under maize cultivation about 5% of the total arable land6. This area
was not enough to cover the needs of Polish livestock farmers as the total imports of grain maize
are substantially. However, despite the low importance within Poland, it still is the 6th producer of
grain maize in Europe with 3.5% of the total sown area. The main maize producing areas within
Poland are centrally located in the provinces of Dolnoslaskie, Wielkopolski and Podlaski,
together representing 38% of the sown maize area and even 49% of the sown grain maize (Polish
Statistics, 2008). Despite the low average of maize in the arable land, an intensive maize growing
area is found in the region of Dolnoslaskie where 9% of the arable land is sown with maize, the
majority in monoculture. On a national level about 23% of all maize grown is monoculture
(European Commission, 2006). The relative high percentage of monoculture despite the low
share of arable land planted with maize can be explained by the Polish farm structure. In 2007
52% of the land and 56% of the livestock was located within farms with less than 20ha cultivated
(Eurostat, 2008). Moreover, 59% of the grain maize cultivation is located on farms with
livestock. Therefore the monoculture in maize is predominantly found on small farms that need
maize as a source of feed and low flexibility due to a land constraint. If maize is rotated, it is
rotated in a 3 or 5 year plan with rape seed, wheat, barley or root crops (European Commission,
2006).
6 The data from Eurostat (2009) differs from the data from the Polish Statistics Office (2009)
indicating that in 2007 206 639ha was grown with grain maize and 310 00ha of silage maize. To keep consistency with the calculations for the other countries we use the aggregated Eurostat data for the welfare calculations later on but use the Polish data to look at the community level
62
WCR was first detected on the 22nd of August 2005 in a trap at Dukla (south Poland), near an
international road leading to Slovakia. The nearest maize field was situated 6 km away. On the
25th of August, another finding was made in a maize field near the airport of Jasionka (EPPO,
2008). In 2006 and 2007 the beetle spread to the whole south of Poland (Figure 13) with findings
in 9 provinces. The provinces of Podkarpackie and Maloposkie have the highest population. Up
to 2007, only beetles where found with the trapping devices and nor larvae nor economic damage
were detected. Because of the high percentage of monoculture in the area of infestation, 20%
(European commission, 2006), economic damage can be expected in the next few years. As the
beetle is spreading towards the central regions of Poland, the most important maize growing
regions might also be infested within the next couple of years. To control the spread of WCR in
order to delay the establishment in the rest of Poland, chlotianidine is temporary registered for
use together with some foliar treatments. However, the application of foliar treatment is difficult
in the Polish settings as very few high clearance sprayers are available (Piorin, 2008).
At first the revenue maximizing technology fee for Bt maize is calculated from the
technology valuation of maize farmers. To incorporate the possibility of rotation with rape seed
as this is an important crop in Poland, the cost of rotation was now assessed with 4 different
possible rotational crops, each counting for 2/3 in the rotation scheme. Estimated in the way
presented earlier, the framework presents a technology fee of €37.6/ha. To allow for the
uncertainty of our data we model θ as a triangular PDF~(27.6;37.6;47.6). It seems from the
comparison of gross margins (Table 23) that for those farmers planting maize in Poland, the gross
margin of maize is significantly higher than for the other crops in rotation. As a result crop
rotation is not a favorable control option for maize farmers against WCR. With an average rent of
€-107/ha for risk neutral grain maize farmers and similar outcomes for silage maize and CARA
63
farmers, no control option is favored in most of the cases in the absence of other control options
(Table 15). In only 2% of the cases crop rotation is better than no control. This indicates there is a
strong demand by the Polish maize farmers for alternative control options. Chemical control, both
seed treatment and soil insecticides offer a significant better result, being preferred to crop
rotation in 62% and 22% of the cases respectively for risk neutral grain maize farmers. Despite
the induced rents of chemical control, the Bt technology offers higher rents in all of the modeled
cases leading to 100% optimal choice. The average rent of Bt maize is €34/ha and €29/ha risk
neutral and CARA grain maize farmers. For silage the rent is a little bit higher with €36/ha and
€31/ha. We do not consider a separate case of monoculture as crop rotation is never a preferred
solutions for those farmers growing maize.
In order to assess the welfare created by introducing Bt maize in Poland we need the extra
value created by the technology. This value is presented in Table 15. For risk neutral farmers the
benefits from introducing the technology amount to €26/ha for grain maize and €31/ha for silage
maize on average per adopted hectare. To assess the aggregate benefits the adoption rate of the
technology is essential. As we know 23% of the maize is cultivated in monoculture,
predominantly by small farmers with livestock, we can assume these are the farmers with a land
constraint and as such willing to adopt. This is in line with findings in the USA where small
farmers use more WCR control due to the reduced flexibility (Wilson et al.,2005). As for farmers
with a land constraint Bt maize is always the preferred option, we model the adoption as a
triangular distribution surrounding the 23% (triangular PDF~(0.18;0.23;0.28) both in silage and
grain maize. Applying the CIR method to aggregate the results, we see that the introduction of Bt
maize in Poland, in case WCR would be established in the maize growing area, generates a
welfare of €9.5 million annually. These benefits are almost equally shared between the seed
64
industry (€5.4 million) and the farmers with €1.5 million for the grain maize producers and €2.5
million for the silage maize producers.
The results show that despite the relative small importance of maize in the Polish agriculture
protection against WCR offers a substantial benefit. The main reason is the presence of 23% of
the maize in monoculture predominantly in small farms. Moreover, the gross margin of maize
makes crop rotation not preferable, increasing the potential of WCR control in farms with a land
constraint despite not having monoculture.
65
Control of WCR in the Republic of Serbia
Maize is the most important crop in the Serbian agricultural sector. Over the long term, maize
has been the crop with the largest output value, ranging from 10% to 20% of the total Serbian
agricultural output. With an average area sown with grain maize of 1.2 million hectares annually
it occupies 38% of the total area sown in Serbia (Statistical Office of Serbia, 2008). The key
figures on Serbian agriculture are presented in Table 16. Besides the dominant area of grain
maize there is a small area of silage maize. The main maize growing areas are located along the
valleys and in the north of the country. The further south less maize is planted and the production
system becomes less intensive. The district northern Vojvodina, covering 24% of Serbian
territory produces about 52% of the grain maize (Statistical Office of Serbia, 2008). Within this
region, grain maize covers 41% of the sown area. No official data is available on the percentage
of monoculture within this region. A recent study by Sivcev et al. (2009) indicates that in their
sample maize was the pre-crop in 48.53% of the fields studied. Before the introduction of WCR,
monoculture was a common practice and differed a lot between communities. In some
communities the share of maize reached 80% (Sivcev, 2008). This practice has been changed
since but it seems that continuous corn is again gaining importance as in the same sample by
Sivcev (2009) the maize fields under first year corn decreased from 93.2% in 2002 to 57.5% in
2006. If maize is grown in crop rotation, wheat (26,2%), sunflower (8.3%) and soybean (5.3%)
where to most common rotational crops (Sivcev et al., 2009). Due to agricultural reform in 2000
the share of cereals fell from 48% to approximately 45% of the total cultivable area (i.e. by
around 90,000 hectares). The reduction has been mainly in the area planted to wheat, the second
major crop, whilst the area of maize has remained fairly constant due to it’s the lower
profitability of wheat (Agripolicy, 2006).
66
As presented earlier, Serbia was the first country in Europe to be infested with WCR with
detection in 1992 around the airport of Belgrade (Baca, 1993). It spread at a high speed from the
Surcin Plateau to the northwest into Lower Srem crossing both the Sava and Danube rivers. The
first economic damage also took place in 1992 on 0.5ha. Therefore it is assumed the actual
introduction took place somewhat earlier but was not documented earlier. The high percentage of
maize under monoculture created the ideal conditions for population growth and damage. In
Figure 14 the registered damages until 2000 are shown. Sivcev (1997) reports average damages
of 20%, even registering damages of 80%. Ever since, the registered damage has been on less
than 1000ha annually with the exception of 2003 with about 3000ha of damage. Sivcev et al.
(2009) report that about 12% of the fields under research had more than 6 beetles/trap/day which
is considered the economic threshold value. As sampling was not representative we cannot
extrapolate this for the rest of the country. In 2008 WCR is established on the whole teritory but
the registered damage is scattered and depends on the individual decision of a farmer to cultivate
in monoculture.
Serbian farms are small with on average 2.49ha of cultivated land per holding (Statistical
Office of Serbia, 2008) allowing little flexibility in the production system. Moreover, especially
maize production is based in small farms as maize only accounts for 10% of arable land on the
larger farms (European Commission, 2006). Analogous to the situation in the US (Wilson, 2005)
one would expect chemical control options to be a rational choice for these small farms.
However, only a few farmers are using chemical control to fight WCR damage. If applied the
chemical control takes the form of a seed treatment also protecting the maize yield against
wireworms. No farmer is applying chemical control solely against WCR (Sivcev, 2008). The
preferred control option of Serbian farmers seems to be crop rotation, as it is promoted by most
67
extension services. This change away from monoculture introduced a drop of sown area of
236 000ha from 1991, just before the first detection, to 2008 (Statistical Office of Serbia, 2008).
The fact that cultivating maize for more than a year on the same plot combined with the
established population of WCR creates a situation in which more damages would be expected
than registered in the last years. Therefore we plot the yield trend in Serbia from 1991 on in
Figure 15. To do a full analysis requires more data on the input use, climatical conditions,
policies and so on but it seem yield has been at a low level since the first economic damage in
1992. It took up to 2005 to reach the same national average yield as in 1991. This could indicate
there is an aggregated yield reduction due to WCR. As mentioned before, detection of reduces
yields is difficult and often only lodging is considered damage. Therefore the damage in Serbia
might be underestimated as even Vojdovina the percentage of maize in rotation is still 41% on
average with regional variation. This means a significant part of the area has more than 50% of
the arable land planted with maize, by Baufeld & Enzian (2005) assessed regions under high risk
of damage. If there is indeed a national yield effect due to WCR in the official data, the results
from our analysis will be conservative as we use the official data as a base yield.
As only 2% of the area is silage maize which is not traded, only used on farm, we consider
only grain maize in the assessment of the control options in Serbia. Prices supplied by the Serbian
Statistical office from 2000 on were used to construct the appropriate distribution found in Table
23. These prices are an annual average but because of the lack of storage capacity on the small
farms, most farmers cannot sell at the optimal point in time. Therefore the variance of prices can
be higher than recorded in our model. As application rates for insecticides are the same as in the
68
other countries and international companies are present in the Serbian input market7, prices are
assumed to be the same as everywhere else in Europe. Data on gross margins has not been
collected in Serbia for a long time by official institutes. We therefore rely on the estimations
using the engineering approach to assess the gross margins in Serbian agriculture (FAO, 2004).
As no heterogeneity in the gross margins is acknowledged we assume boundaries of the PDF to
be 50% of the average. No gross margin for wheat is available but as the area has been decreasing
since the policy reforms due to the low gross margins, we believe that our estimate with only
soya and sunflower is a conservative one. Gross margins may seem high considering the state of
Serbian agriculture but gross margins in Serbian agriculture are known to be high even up to 39%
compared to the neighboring countries with an average gross margin of 15% of the revenue
(European Commission, 2006).
Because of the increased uncertainty in the Serbian data compared to the other countries
discussed, we opt to set the technology fee at the level of Hungary instead of calculating it via the
framework. We believe the case of Hungary is similar because of the similar competition with
crop rotation and because of the established status of the pest causing real economic damage. The
average technology fee used was therefore €33/ha.
The results of the farm model are presented in Table 17. Looking at the rents created for risk
neutral maize producers, we see that both crop rotation and Bt maize offer a solution on average
preferred to no control at all. Crop rotation creates a profit of €11/ha on average with positive
rents in 56% of the modeled cases. This positive outcome, compared with the fact that crop
7 The only input sector where foreign firms are underrepresented is the seed sector with a share of only 5% of
the sold seeds (European Commission, 2006). This market structure may have its effect on the introduction of Bt technology in Serbia as either licensing with domestic firms has to take place or acquiring a larger share of the market by foreign firms.
69
control does not increase input expenses may offer the explanation for the adoption of crop
rotation in a Serbian setting. Moreover, it also explains why Sivcev et al. (2009) find that farmers
are going back to continuous maize as no control offers a better result in 44% of the cases due to
the low value of the yield loss. This result combined with the findings of Baca et al. (2007)
advising new tools for WCR control indicate that crop rotation might not be the economically
optimal solution after all. The reason for not adopting other control options might be the cash
flow of farmers combined with too high prices in the Serbian setting. With an average income of
about €150/month for persons employed in the agricultural sector (Statistical Office of Serbia,
2008) input use is low. This can be seen in the fact that small farmers only apply pesticides in
cash crops (Agripolicy, 2006). This behavior can be explained by the risk aversion of credit
constrained farmers. As maize is almost completely used for feed on farm, pesticide use might
not be favored by farmers although potentially offering benefits. This fact should be taken into
account when assessing the potential adoption of different control options in Serbia.
Competing technologies now available in the market create negative average rents. Seed
treatment offers positive revenues in 42% of the cases while soil insecticides are only profitable
in 14% of the cases. This may explain the finding that farmers are only willing to adopt chemical
treatment if it protects at the same time for other pest such as wireworms. Looking at the full
sample of modeled cases we see that even with the introduction of Bt maize in 33.3% of the
cases, crop rotation is preferred by land constrained farmers. Seed treatment offers the best
revenue in very few cases and Bt maize takes the edge in 66.5% of the cases. The situation for the
CARA grain maize farmer is similar but because of the high coefficient of variation of crop
rotation the option of crop rotation looses a significant part of its rents. However, as we modeled
the spread in the gross margins in an arbitrary way the results for CARA farmers have to be
70
interpreted with care. Therefore we focus on the risk neutral farmers in the Serbian case. If crop
rotation is not feasible at all due to livestock production or contracting, the extra value of
adopting of Bt maize increase slightly to 26€/ha and would be preferred in 99.7% of the cases
and seed treatment in the other 0.3% of the cases.
Assessing social welfare induced by the adoption of Bt maize in Serbia is difficult for the
reasons described earlier. The strictly rational case assumes farmers will adopt and apply a
certain technology if it potentially generates a rent. In this scenario we can rely on the knowledge
gained from the risk neutral case in the farm level analysis. As we assume the full area of
Vojvodina has some kind of land constraint as maize still represents 41% of the arable land
despite the presence of WCR, 66.5% of the maize area or 412 300ha in 2007 would be planted
with Bt maize in Vojvodina. The rest of Serbia has a less densely planted maize area with 35% of
the sown area under maize but we can still assume that about half of the area has some kind of
land constraint as planting density is still high compared to other countries. This would lead to
another 183 540ha planted with Bt maize in 2007. The total percentage of adoption in 2007
amounts to 50% of the Serbian maize area. To allow some flexibility we model adoption to range
between 40 and 60% in a triangular PDF. The results show an annual welfare creation of
€27.5million for the grain producers and €19.3 million for the seed suppliers. As the revenue for
the seed industry is that high, Serbian maize seed producers have a high incentive to engage in
licensing strategies with the technology owners.
However as argued before, the adoption decision may not be purely based on the average
change in revenue. Input costs can be an important decision factor as cash flow is low and
income of farmers low. With the adoption of Bt maize there is a 15% change of making a lower
profit than not applying any control and a 33.3% change that crop rotation, which does not
71
necessarily increases input costs, may have been a better solution. Therefore in order to allow Bt
maize to be adopted to its full potential credit provision might be a key. This phenomena is
already observed in the case of machinery where mechanization took place in the last years due
to the development of leasing and commercial credit markets. State subsidies of 30% of the
purchase price of certain kinds of new machinery coupled with measures to facilitate short-term
and long-term credit have also encouraged machinery purchases. As there is such a potential high
value for both Serbian seed suppliers and maize farmers a policy inducing the same effect in the
seed market would have the potential of increasing nationwide welfare and certainly of those
engaged in maize farming.
72
Control of WCRin Romania
The transition process in Romania was focused on the privatization of land, aiming at
changing collective agriculture to individual agriculture, as well as on the downsizing of the
farms (Lerman 1999). The majority of farmers chose individual farming and thus, in 2005, 4.2
million individual farms cultivated 65% of the arable area with an average farm size of 3.3ha per
farm (NIS, 2008). These new individual producers lacked the necessary know-how to cultivate
their land. They had no cash to invest and rarely access to credit or agricultural equipment. Up
and downstream sectors had not been restructured to suit the needs of the small farmers which led
to high transactions costs by using the different input and output markets. Such transaction costs
and the lack of capital reinforced the decline in the use of inputs like fertilizer and certified seed.
By responding to these difficulties producers diversified their production, substituted commercial
by non-commercial crops, technical crops by traditional crops and increased subsistence
production. The latter finally further promoted the stagnation in the development of input and
output markets and led to a vicious circle. The increase in maize cultivation in Romania during
this period is basically linked to these developments in the agricultural sector as it is easy to
cultivate with low input use (Balint & Sauer, 2006). The transition process led to an all time
record of 3.2 million hectares sown in 2004 (Table 18). The following years, the sown area
decreased with plantings of about 2.5 million hectares in 2008. During this time Romania also
evolved to becoming a net importer of maize. Romania has the biggest area sown with maize of
all EU-27 member states but the low yields rank it as only third producer in EU-27. Within
Romania maize is an important crop occupying about 32% of the total arable land in 2006. Table
19 shows the regional characteristics of maize cultivation in Romania. The eastern half of the
country, with inclusion of Muntenia, represents about 60% of the area sown with maize. From the
share of maize in crop rotation we see that the north eastern part of the country has the highest
73
density of maize plantings with 42%. In 2006 about 50% of the plantings where in the form of
continuous maize, especially by the small scale farmers (max 2h) covering almost 50% of the
arable land (Rosca, 2006).
WCR was first reported in Romania in 1996 at Nadlac (Arad county), near the Hungarian
border. In the following years, the pest has spread towards the north-east and the population
levels have increased especially in Caras-Severin, Timis, Arad and Mehedinti counties. WCR has
continued to spread towards the east (Eppo, 2008). Since, WCR has spread at a high rate due to
the high percentage of maize in the crop rotation favoring the spread. The pest is well established
now in Romania. In 2006 the infested area was about half of the country (118.3km²) and
preventing spread to the rest of Romania was difficult with a front line of over 400km and the
most densely planted areas in the east not yet infested. Rosca (2006) argues that up to 2006 the
possibility of informing farmers about the risk, importance and presence of the invasive pest have
been limited. Perhaps even more important is the low possibility for farmers to interrupt
monoculture as no suitable crop is available to be grown in rotation with the same ease and
profitability and no chemical control options subsidized by the State. In 2006 some chemicals to
control adult populations were registered but no suitable options to target larvae. Despite these
settings, official damages have been small over the last years, some km² in south west Romania
(Rosca, 2006). This is unexpected due to the experiences in the other infested countries with an
important area of continuous maize. The reason might be the drought during June preventing
larvae development followed by summer rains that help the regeneration of the injured root
system. The reason might also be that the perceived damage is lower than the real damage as it is
difficult to assess the damage caused by WCR in the low yielding Romanian maize sector and
because of the limited knowledge of the farmer about the pest.
74
We assess the different control options taking into account rational farmers. The data used
can be found in Table 23. As a significant part of the area is infested and potentially damaged by
WCR the national average yield may be assessed as a conservative estimation of the potential
yield in Romania. However, because of the spatial and temporal variation we assume the
potential yield is covered by our PDF. Price data for silage maize are not available but as the area
is small in Romania we opted to use the estimation by Schaafsma (1999) of green maize returns
€9.1/ha in the European Union and introducing a coefficient of variation of 0.5. The other data in
the model are set as before. As most of the chemical options are not available in the Romanian
market, we use the prices found in Hungary.
After assessing the technology valuation for Bt maize of grain maize farmers, the revenue
maximizing technology fee can be calculated. With an average technology valuation of €53.7/ha
for the Bt technology the optimal price level would be €36.7/ha. Again we introduce a triangular
PDF into our model to capture the uncertainty of the available data (PDF~(26.7;36.7;46.7)).
In the case of risk neutral farmers, both seed treatment and Bt maize created an average
positive rent compared to no control (see Table 20). Seed treatment offers a positive control of
WCR in 50% of the cases with an average outcome of €5/ha. Bt maize is only outperformed by
no control in 15% of the cases leading to an average rent of €31/ha for the whole population of
grain maize farmers. Both soil insecticide and crop rotation have a negative rent for risk neutral
farmers of respectively €-18/ha and €-61/ha. Compared to no control option at all, crop rotation is
rationally spoken only favored in 13% of the cases and soil insecticides in 20% of the cases. For
CARA farmers the results differ somewhat due to the high variation in outcome on seed
treatment. This high variance implies a high risk premium decreasing the rent created by the
chemical control. Crop rotation comes out a little bit stronger for CARA farmers compared to the
75
other technologies due to the lowest variance on the outcome. Looking at all modeled cases, we
see that Bt maize is the most profitable solution in 98.4% followed by seed treatment in 1.3% of
the cases and a small 0.2 percent for crop rotation for risk neutral farmers. In the absence of Bt
maize the optimal solution would be seed treatment in 95% of the cases and crop rotation in the
remainder of the cases. This indicates that the soil insecticides, at the Hungarian price, are too
expensive relative to the value of the potential damage in the Romanian setting. For CARA
farmers the 1.3% of seed treatments shifts to crop rotation creating a potential adoption on 1.5%
for crop rotation.
For silage maize, which is considered low value and low yielding in Romania, hence the very
small area under cultivation, the results are clearly different. The ranking of control options
remains the same as in the grain maize production but all options have a negative average
valuation. Bt maize, the most valuable option induces an average loss compared to no control of
€9.1/ha. Considering all modeled cases Bt maize is the best of the four control option 100% of
the time. However, the value of Bt maize is only preferred to no control at all in 18% of the cases.
Therefore control options in the silage maize production will remained limited, even with the
introduction of Bt maize in Romanian agriculture.
The countrywide effect of introducing Bt maize, the optimal control option in 98,4% of the
cases for land constrained farmers, we again use the CIR methodology. With about 50% of the
grain maize area under continuous maize and most of the area with small farmers not having the
flexibility and availability of rotation the potential adoption rate is high. We therefore model
adoption as ranging between 40% and 60% of the grain maize area. For silage maize we assume
farmers will prefer no control option at all as the investment only pays off in 18% of the cases.
Whether the adoption level in grain maize will be reached will depend on a high extent on the
76
availability of credit and policy incentives to increase input use as most Romanian farmers now
rely on very low input use and will not be interested in adopting the technology under the actual
institutions. The potential welfare created by introduction of the Bt technology in Romania
creates €7.7 million for the grain farmers while the seed could generate a annual gross revenue of
€11.0 million with the assumed pricing structure. As expected there is a huge value to be created
by Bt maize in Romanian agriculture if institutions change. For now, no control at all or some
seed treatment might be the optimal choice and this is confirmed by the actual behavior in the
Romanian grain maize sector.
77
Control of WCR in the Ukraine
The Ukraine is considered to have the potential to serve as the bread basket of Europe. The
country possesses very fertile soils with about 40% of the worlds black soils within its territories.
Maize is one of the most important grain and forage crops with about 2 million hectares planted
with grain maize annually (see Table 21 for detailed figures). The planted area has increased
since 2000 despite a number of constraints, such as obsolete and inadequate harvesting
equipment, the high cost of production, especially post-harvest drying costs, and pilferage. The
main growing region is eastern and southern Ukraine, although rainfall in some oblasts in the
extreme south is too low to support maize growth. Maize is typically planted in late April or early
May. Harvest begins in late September and is usually nearing completion by early November.
Only about 50% of total maize area is harvested for grain, the remainder being cut for silage,
usually in August. Maize is used chiefly for poultry and swine feed, and production and
consumption have increased since 2000 concurrent with a rebound in poultry inventories (USDA,
2004).
Most of the Ukrainian farms engage in either crop rotation or mixed farming. The distribution
of the production mix changes according to the typology of the farms. Corporate farms generate
on average 70% of their value from crop production while household farms, more subsistence
based, only generate 49% of their output through crop production (Lerman, 2007). Interesting to
note is the difference in feed sourcing between corporate and household farms. The latter only
produce 43% of their own feed which is complemented with communal sources while corporate
farms produce 93% of the feed needed. Maize, through the easiness of cultivation and the good
feeding value is disproportionally popular with the household farms as they have less flexibility
in the crop rotation (Lerman et al., 2007). Moreover, 26 percent of the private farmers have
78
difficulties in following crop rotations. In the Ivano-Frankivsk oblast the proportion is almost
twice as high, at 46 percent. The observation of a crop rotation seems to be a bigger challenge for
smaller farms in the Western region of Ukraine (FA0, 2005). According to the Institute of Plant
Protection (2008) the country wide percentage of maize under monoculture is 10-12% but with a
high regional variation as presented by FAO (2005).
In August 2001, WCR was caught for the first time in Zakarpatya region, west of the
Carpathians near the Hungarian and Romanian borders (districts of Vynogradiv and Beregove).
Since then, the beetle spread throughout Transylvania, the South Eastern part of the country
bordering the infested countries Romania, Hungary, Slovakia and Poland. The spread after six
years of infestation can be seen in Figure 1. Officially there is no yield loss recorded yet.
However, WCR population is increasing and with about 25% of the maize in the infested area in
monoculture (Institute of Plant Protection, 2008) there is a high risk of damage in the near future.
In 2008, 15 000ha of arable land (30% of it planted with maize) were officially under quarantine.
The quarantine zone is created by a 10km radius around a beetle found in 2007. The quarantine
means no maize can be grown without chemical application in the following year. Most known
chemical treatments are available on the Ukrainian market at similar prices as in the rest of
Europe. After a quarantine zone of one year, the decision on how to cope with the infestation lies
with the farmer himself. Especially for those farmers having some land constraint may still
engage in monoculture despite the finding of the beetle in their vicinity. In 2008 the population in
the valleys dropped significantly due to the floods in spring preventing larvae hatching. The
Institute of Plant Protection (2008) expects WCR to spread to all Ukrainian maize growing areas
as the average percentage of monoculture of around 10% is enough for the population to grow.
79
As an essential part of the assessment of crop rotation as a viable control option, the gross
margin (and its heterogeneity) needs to be incorporated. However, gross margin calculations are
not officially available. Furthermore, the heterogeneity among Ukrainian farmers is very large
due to differences in how farmers attempt to overcome the rigidities of the transition period and
their management skills (Zorya and von Cramon-Taubadel, 2006). Therefore, to reduce the
possible bias introduced by the limited (average) data available we choose to adopt the same
strategy followed in the Serbian case. The average gross margins from the literature are
introduced through a PDF using a deviation of 50% as the bordering value. This and other data
are shown in Table 23. As the calculation of the technology fee for Bt maize depends on a high
extent at the variance of the technology valuation we opt for a uniform pricing strategy which
introduces the technology fee assessed for Hungary in the Ukraine.
The results of our farm level analysis are shown in Table 22. The top rows show the results
for grain maize production. It seems all control options induce an average rent below the status
quo of no control. This is the result of two factors. The first one is the introduction of an
exogenous technology fee that is not perfectly suited to the Ukrainian settings. We used the
technology of Hungary which seems too high on average for Ukrainian farmers. The second
reason is related to the first and can be found in the low yields and value of Ukrainian maize.
Compared to the other countries in this analysis both yields and prices are far below the average
value. Therefore the percentage loss from WCR is of a lower value and crop protection is valued
lower. The best two control options given these constraints are Bt maize and crop rotation. In
grain maize the use of Bt maize is favored in 64% of the cases and crop rotation in the remaining
cases. However, Bt maize would only offer a result better than no control in 16% of these. Of the
36% of the cases crop rotation scores better than Bt maize, 33% offer also a better control than no
80
control option. This means in 22% of the modeled cases applying some form of protection would
be rational, in the other cases no control option would be optimal. We know from the regression
model in Table 4 that these cases are situated in high value, high yielding and high damaged
fields. This means there is scope for protection in the high value farms in the Ukraine but not for
the smallholder using maize as a feed source in the subsistence oriented production. For CARA
the situation is similar but even less % of the area would be under some kind of control option
(16%) as all the rents decrease.
As silage maize is of even lower worth the control options are certainly too expensive for
those cultivating the crop. Although Bt maize is the best solution in 68% of the modeled cases
compared to the other options, it only has a positive value in as little as 4%. Crop rotation on the
other hand has a positive outcome in 22% of the cases but is lower on average because of the low
average profit compared with Bt maize that offers very high benefits for some farmers with a
maximum profit of €87/ha for Bt maize and a max of €31/ha for crop rotation. The results show
that the adoption potential of control options in silage maize is small.
This adoption rate is what we need in order to assess the welfare from adopting Bt maize in
the Ukraine. We have stated before that about 26% of the farmers have some kind of land
constraint and about 10-12% of the total maize area is under monoculture. We assume that
because of the low value production in the Ukraine, Bt maize will only be adopted by those
farmers generating a positive value from adopting in monoculture compared to no control. For
grain maize this would lead to an adoption rate of 10% of the area under monoculture, for silage
9%. In contrast to some other countries in the analysis there is limited reduction of this adoption
rate through credit constraints. Lerman et al. (2007) show that the credit market, and therefore
availability of cash, is well developed in the Ukraine. Of the corporate firms about 71% of those
81
in need of credit have access while 42% of the households in need have access. We consider that
adoption is only assumed in the high yielding part of the PDF of potential adoption. It seems
reasonable to assume that the most efficient farmers have access to cash for input use. In order to
allow some flexibility in the adoption process we use a PDF~(0.01;0.02;0.05) in the CIR
methodology for both silage and grain maize. The results show that despite the high price of the
technology and the competition with no control, farmers could gain €632 517 annually with
€320 940 accruing to the grain maize farmers and €311 577 to the silage maize farmers. Due to
the high price for the technology, the seed industry captures almost 80% of the benefits with €2.7
million annually. In case the optimal price would be asked in the Ukrainian market, the share of
the seed industry would decrease while the value for the farmers would increase (Dillen, Demont
and Tollens, 2008).
The results show that the Ukrainian maize sector is not in a great demand of control options.
The actual state of production with low input use for a low yielding and low value crop does not
create an incentive to adopt these technologies. If in the coming years the Ukraine lives up to its
possibilities and increases the maize output in a more open trade model (recently entered the
WTO) the situation could change and control options may become economic viable on a larger
scale.
82
Conclusions
The study assessed the competitiveness of the different control options against WCR damage
in eight countries in central Europe. The study has to cope with two major sources of uncertainty
surrounding the problem statement. First of all, the potential damage caused by WCR in Europe
is difficult to assess. From experience in the USA we know that the damage is influenced by
several exogenous factors such as rainfall, soil,… making the prediction of damages under
different control options based on population statistics difficult. Moreover, only fragmented data
on damage is available in Europe up to now. Secondly, as the beetle has not yet spread in the
whole area under research and control options are not yet commercialized in all countries under
research the ex ante impact study cannot rely on actual data. For these reasons the problem
statement has been addressed through a stochastic modeling approach. This framework, explicitly
incorporating the heterogeneity in farm practices, has several advantages. First of all it allows
modeling corporate pricing strategies for those control options not yet commercially available.
Moreover, it gives insight in the spread of outcomes which is a major advantage compared to
other approaches as focusing on the average farmer often neglects a lot of information. Finally
the modeling approach allows for a full sensitivity analysis. This sensitivity analysis can be used
to develop a multi-criteria decision rule for the farmer on which control option to adopt instead of
relying on the ambiguous population statistics. To supplement this approach, a willingness to pay
assessment is executed to capture the non-pecuniary benefits of the one control option, Bt maize,
as literature has shown these can be important in the adoption process.
If all the presented technologies would be available in the market, Bt maize and crop rotation
seem to be the optimal choice for most of the land constrained farmers complemented with seed
treatment in some cases. However, due to the different constraints in different countries, the
83
outcome is spatially heterogeneous. The choice for Bt maize ranges from 14% in Ukrainian silage
maize to 100% in Poland (Table 24). In absence of Bt maize, soil insecticides also offer the
optimal solution in selected cases. However as they are mainly targeted to high yielding high
value producers, Bt maize outperforms them in this setting. For the country specific results on
rents and constraints we refer to the specific countries. Here we summarize the effects of
introducing Bt maize resistant against WCR damage in European maize production, presented in
Table 24. These results can be interpreted as the value forgone by not allowing the cultivation of
this variety in case WCR spread all over the countries under research.
In Hungary the revenue maximizing technology fee amounts to €33/ha, offering a value of
protection against WCR of €63/ha in grain maize and €40/ha in silage maize for risk neutral
farmers. This has to be weighted to the level of protection of the competing technologies. We see
that for land constrained farmers the extra value accruing to farmers is €18/ha and €22/ha while
in monoculture the value increases to €34/ha and €33/ha in grain and silage maize respectively.
Under this conditions, Bt maize offers the optimal control for 80% of the cases in grain maize
and 93% of the cases in silage maize. The last two columns show that the total welfare created on
a per hectare basis for farmers and input suppliers combined ranges from €51/ha to €67/ha
depending on the production system.
Similar results for the other countries can also be found in Table 24. The countries with the
highest demand for Bt maize besides Hungary are Slovakia and Poland. Austria can created a
high value in the south eastern intensive maize growing area, up to €81/ha but has a lower
demand on a national scale. These are the countries where crop rotation as a control option is less
favorable which can be deducted from the same profits in monoculture as in land constrained
production.
84
We also see some negative values in the table. In Czech Republic they stem from the fact that
crop rotation does not bear a high cost for most farmers, and is therefore favored by almost 50%
of the famers. In the case of monoculture however, also Czech farmers can gain substantially
from the adoption of BT maize. For Romania and especially the Ukraine the reason for the
negative rents are different. As yields and prices are low the value of the damage by WCR is low
compared to the prices for protection for specific farmers. Therefore it makes more sense for
most of the farmers to not apply any control options at all. This is reflected by the low
percentages, 17% and 14%, that Bt maize resistant against WCR would be the optimal choice in
the Ukraine. However, we know from the sensitivity analysis that for those farmers in the
Ukraine having high yields through good input use there is some value (around €20/ha on
average) from adopting Bt maize but for the small scale subsistence farmers other strategies may
be preferred.
The WTP analysis shows that Hungarian farmers (and by extend the farmers in the countries
under research) also have a positive valuation for the non-pecuniary effects of the technology.
This value amounts to €12/ha.
Up scaling the regional farm model to a national impact is not easy as data on land constraints
and monoculture are not available. Therefore these results need to be interpreted carefully. The
created welfare accruing to farmers varies from €700 000 in the Czech Republic to €15million in
Hungary annually. These results show that there is a huge value to be created even if the absolute
amount is only an estimation.
The study demonstrates that there are different optimal control options depending on the
situation of the farmer’s constraints. Therefore there is a need for more data on the farm specific
85
properties to aid the farmers in their decision. This paper gives the first tendencies and can act as
a basis for further research.
86
References
AGbios, AGbios GM crop database, 19-12-2008.
AGES (2008). Personal Communication.
Agripolicy (2006). Monitoring of agricultural policy market and trade developments in Serbia. Available at www.agripolicy.net.
AKII (2008a). Livre de poche statistique de l'économie agricole. Research institute for agricultural economics (AKII).
AKII (2008b). Personal Communication.
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.
Alston, J.M., Norton, G.W., & Pardey, P.G. (1995). Science under scarcity: principles and practice of agricultural research evaluation and priority setting. Ithaca, NY: Cornell 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.
Baca, F. (1993). New member of the harmful entomofauna of Yugoslavia Diabrotica virgifera virgifera LeConte (Coleoptera,Chrisomelidae). XXI skup entomologa Jugoslavije. IWGO Newsletters 13: 21-22.
Baca, F., Videnovic, Z., Erski, P., Stankovic, R., & Dobrikovic, D. (2003). Time of planting and choice of maize hybrids in controlling WCR (Diabrotica virgifera virgifera Le Conte) in Serbia and Montenegro. Communications in agricultural and applied biological sciences, 68(4), 79-87.
Baca, F., Levic,J., Stankovic,S., Stefanovic,L., Simic,M., Gosic-Dondo,S.,&Knezevic,S.(2007). Factors contributin to the population changes of major pests in Serbia. Maydica 52, 343-346.
Balint, B., & Sauer, J. (2006). Distorted prices and producer efficiency: Romanian maize. Paper presented at the 46th annual meeting of German Association of agricultural economists, Giessen.
Bateman, I.A., Carson, R.T., Day, B., Hanemann, M., Hanley, H., Hett, T., Jones-Lee, M., Loomes, G., Mourato, S., Özdemiroglu, E., Pearce, D.W., Sugden, R., & Swanson, J. (2002). Economic valuation with stated preference techniques: A manual. Cheltenham, UK: Edgar Elgar.
Baufeld, P., and Enzian, S. (2005a). Establishment potential of the western corn rootworm (Diabrotica virgifera virgifera) in Europe. Paper presented at the.
Baufeld, P., & Enzian, S. (2005b). Maize growing, maize high risk areas and potential yield losses due to Western corn rootworm (Diabrotica virgifera virgifera) damage in selected Euroopean countries. In S. Vidal, U. Kuhlmann, & C.R. Edwards (Eds.), Western corn rootworm: Ecology and
managment (pp. 285-302). CABI publishing.
Baufeld, P., & Enzian, S. (2005c). Western corn rootworm (Diabrotica virgifera virgifera), its potential spread and economic and ecological consequences in Germany. Symposium Proceedings No. 81:
Introduction and Spread of Invasive Species, 149-154.
87
Boriani, M., Agosti, M., Kiss, J., & Edwards, C.R. (2006). Sustainable management of the western corn rootworm , Diabrotica virgifera virgifera LeConte (Coleoptera: Chrysomelidae), in infested areas: experiences in Italy, Hungary and the USA. EPPO Bulletin, 36(3), 531-537.
Cagan, L. (2008). University of Nitra, Slovakia. Personal Communication.
Cameron, T.A., & Quiggin, J. (1994). Estimation Using Contingent Valuation Data from a "Dichotomous Choice with Follow-Up" Questionnaire. Journal of Environmental Economics and Management, 27(3), 218-234.
Ciosi, M., Miller, N.J., Kim, K.S., Giordano, R., Estoup, A., & Guillemaud, T. (2008). Invasion of Europe by the western corn rootworm, Diabrotica virgifera virgifera: multiple transatlantic introductions with various reductions of genetic diversity. Molecular Ecology, 17(16), 3614-3627.
Coates, S.A., Tollefson, J.J., & Mutchmor, J.A. (1986). Study of migratory flight in the Western corn rootworm. Environmental Entomology, 15 620-625.
Cox, W.J., Shields, E., & Cherney, D.J.R. (2007). Seed-applied insecticides inconsistently affect corn forage in continuous corn. Agronomy Journal, 99 1640-1644.
Demont, M., Cerovska, M., Daems, W., Dillen, K., Fogarasi, J., Mathijs, E., Muška, F., Soukup, J., & Tollens, E. (2008). Ex ante impact assessment under imperfect information: GM technologies in the New Member States. Journal of Agricultural Economics, 59(3), 463-486.
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., Oehmke, J.F., & Tollens, E. (2006). Alston, Norton and Pardey revisited: The impact of Bt maize in Spain (Working Paper No. 89). Leuven: Katholieke Universiteit Leuven.
Demont, M., Tollens, E., & Fogarasi, J. (2005). Potential impact of biotechnology in Eastern Europe: Transgenic maize, sugar beet and oilseed rape in Hungary (Working Paper No. 92). Leuven: Katholieke Universiteit Leuven.
Devos, Y., Demont, M., Dillen, K., Reheul, D., Kaiser, M., & Sanvido, O. (2008). Coexistence of genitacally modified (GM) plants and non-GM plants in the European Union. A review. Agriculture for sustainable development, 29 11-30.
Dillen, K., Demont, M., and Tollens, E. (2008). Modelling heterogeneity to estimate the ex ante value of biotechnology innovations. Paper presented at the 12th Congress of the European Association of Agricultural Economists,Gent.
Dillen, K., Demont, M., and Tollens, E. (2009). Corporate Pricing Strategies with Heterogeneous Adopters: The Case of Herbicide Resistant Sugar Beet. Under review
Draper, S. (2007). Choosing seed dressings and seed rates (Agronomy technical note Maize Grower Association, UK.
88
Elliot, N.C., Gustin, R.D., & Hanson, S.L. (1990). Influenc of adult diet on the reeproductive biology and survival of Western corn rootworm (Diabrotica virgifera virgifera). Entomologia experimentalis et
applicata, 56 15-21.
EPPO, Diabrotica virgifera, http://www.eppo.org, 15-11-2008.
European Commission (2006). Final report of a mission in Poland to evaluate the measures taken against Diabrotica virgifera virgifera. DG Sanco 773/2005. Brussels.
European Commission (2008). Prospect for agricultural markets and income in the European Union 2007-2014 Brussels.
Eurostat (2009). Statistics office European Union, available at http://epp.eurostat.ec.europa.eu.
Fall, E.H., & Wesseler, J. (2007a). Report on market analyses for the potential use of biological control products in high value maize systems in Europe (Diabract No. D02.15).
Fall, E.H., & Wesseler, J. (2007b). Report on practical compatibility and economic competitiveness of each biological control option with chemical control and with cultural control of WCR (Diabract No. D02.16).
FAO (2004). Serbia and Montenegro: Review of the sugar sector. FAO investment centre/ EBRD
cooperation program report series,6.
FAO (2005). Ukraine:REvew of the sunflower oil sector. FAO investment centre/ EBRD cooperation
program report series,10.
FAO (2005). Fertilizer use by crop in Ukraine. Rome, FAO.
Foster, R.E., Tollefson, J.J., Nyrop, J., & Hein, G.L. (1986). Value of adult corn rootworm population estimates in pest managment decision making. Journal of Economic Entomology, 79 303-310.
Furlan, L. (2006). The potential of managment of maize planting date in Diabrotica v. virgifera eradication-containment programmes. Paper presented at the.
Furlan, L., Canzi, S., Di Bernardo, A., & Edwards, C.R. (2006). The ineffectiveness of insecticide seed coatings and planting-time soil insecticides as Diabrotica virgifera virgifera LeConte population suppressors. Journal of Applied Entomology, 130(9/10), 485-490.
Gerber, C.K. "Testing the areawide pest management concept on the western corn rootworm (Diabrotica virgifera virgifera LeConte) in northwestern Indiana and east-central Illinois.", Purdue University, 2003.
Gloyna, K., & Thieme, T. (2007). Cereals and weeds as potential host plants for Diabrotica virgifera. IWGO Newsletter, 28(1), 44.
Gómez-Barbero, M., Berbel, J., & Rodrígues-Cerezo, E. (2007). Adoption and socioeconomic impacts of the first genetically modified crop introduced in EU agriculture: Bt maize in Spain (EUR technical report Zaragoza: European Commission.
Goodwin, B. (2008). Payment limitations and acreage decisions under risk aversion: A simulation
approach. American Journal of Agricultural Economics, 91(1), 19-41.
89
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: Springer.
Gray, M.E., Sappington, T.W., Miller, N.J., Moeser, J., & Bohn, M.O. (2009). Adaptation and Invasiveness of Western Corn Rootworm: Intensifying Research on a Worsening Pest*. Annual
review of entomology, 54(1), 303-321.
Haab, T.C., & McConnel, K.E. (2002). Valuing Environmental and Natural Resources: The econometrics of non-market valuation. Cheltenham UK: Edwar Elgar.
Hanemann, M., Loomis, J., & Kanninen, B. (1991). Statistical Efficiency of Double-Bounded Dichotomous Choice Contingent Valuation. American Journal of Agricultural Economics, 73(4), 1255-1263.
Hatala Zsellér, I., Ripka, G., and Vörös, G. (2006). Some economic aspects of Western Corn Rootworm for the Hungarian maize production. Paper presented at the IWGO conference.
Hennessy, D.A. (1998). The production effects of agricultural income support policies under uncertainty. American Journal of Agricultural Economics, 80 46-57.
Herriges,J.A., and Shogren J.F. (1996). Starting point bias in dichotomous choice valuation with follow-up questioning. Journal of Environmental Economics and Management, 30:112-131.
Horak, A., Ayoub, S., Nowakowski, R., Peczze, R., and Skillman, S. (2008). Cruiser seed treatment, Force granular insecticide and Karate with Zeon technology foliar spray as components of an integrated program against Corn Rootworm. Paper presented at the.
Hyde, J., Martin, M.A., Preckel, P.V., & Edwards, C.R. (1999). The economics of Bt corn: Valuing protection from the European corn borer. Review of Agricultural Economics, 21(2), 442-454.
Institute of Plant Protection (2008). Ukrainian academy of agrarian sciences. Personal communication.
Just, R.E., & Pope, R. (2002). A comprehensive assessment of the role of risk in U.S. agriculture. Boston: Kluwer.
Kim, J.S., & Yum, B.J. (2008). Selection between Weibull and lognormal distributions: A comparative simulation study. Computational Statistics & Data Analysis, 53(2), 477-485.
Kiss, J., Komaromi, J., Bayar, K., Edwards, C.R., & Hatala Zsellér, I. (2005). Western corn rootworm and the crop rotation systems in Europe. In S. Vidal, U. Kuhlmann, & C.R. Edwards (Eds.), Western
corn rootworm ecology and managment (pp. 189-220). Wallingford: CABI Publishing.
Knolhoff, L.M., Onstad, D.W., Spencer, J.L., & Levine, E. (2006). Behavioral Differences Between Rotation-resistant and Wild-type Diabrotica virgifera virgifera (Coleoptera: Chrysomelidae). Environmental Entomology, 35 1049-1057.
Krinsky, I., & Robb, A.L. (1986). On Approximating the Statistical Properties of Elasticities. The Review
of Economics and Statistics, 68(4), 715-719.
Krishna, V.V., & Qaim, M. (2007). Estimating the adoption of Bt eggplant in India: Who Benefits from public-private partnership? Food Policy, 32(5-6), 523-543.
90
Krysan, J., & Miller, T. (1986). Methods for the study of Diabrotica. New York: Springer Verlag.
Lerman, Z. (1999). Status of Land Reform and Farm Restructuring in Central and Eastern Europe: A Regional Overview. World Bank Technical Paper 465.
Lerman, Z., Sedik, D. , Pugachov, N., &Goncharuk, A. (2007). Rethinking agricultural reform in Ukraine. Studies on the Agricultural and Food Sector in Central and Eastern Europe, 38. Halle: IAMO.
Levine, E., & Oloumi-Sadeghi, H. (1996). Western corn rootworm larval injury to corn grown for seed production following soybeans for seed production. Journal of Economic Entomology, 89 1010-1016.
Levine, E., Spencer, J.L., Isard, S.E., Onstad, D.W., & Gray, M.E. (2002). Adaptation of the western corn rootworm to crop rotation: Evolution of a new strain in response to a managment practice. American Entomology, 48 94-107.
Ma, B.L., Meloche, F., & Wei, L. (2009). Agronomic assessment of Bt trait and seed or soil-applied insecticides on the control of corn rootworm and yield. Field Crops Research, 111(3), 189-196.
Macleod, A., Baker, R., Cheek, S., Eyere, D., & Cannon, R. (2007). Pest risk analysis for western conr rootworm York,UK: Central Science Laboratory.
Marra, M., Hubbel, B.J., & Carlson, G.A. (2001). Information quality, technology depreciation and Bt cotton adoption in the Southeast. Journal of Agricultural and Resource Economics, 26(1), 158-175.
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.
Marra, M.C., Piggott, N.E., & Carlson, G.A. (2004). The net benefits, including convenience, of Roundup Ready® soybeans: Results from a national survey (Technical Bulletin No. 2004-3). Raleigh, NC: NSF Center for Integrated Pest Management.
Meinke, L.J., Siegfried, B.D., Wright, R.J., & Chandler, L.D. (1998). Adult susceptibility of western corn rootworm (colepotera: Chrysomelida) populations to selected insecticides. Journal of Economic
Entomology, 91 594-600.
Metcalf, R. (1986). Foreword. In J. Krysan, & T. Miller (Eds.), Methods for the study of pest Diabrotica (pp. vii-xv). New York: Springer Verlag.
Mitchell, P. (2004). 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.
Mitchell, P.D. (2002). Yield Benefit of Corn Event MON863. Faculty paper series, 02-04.
Moeser, J., & Vidal, S. (2004). Do Alternative Host Plants Enhance the Invasion of the Maize Pest Diabrotica virgifera virgifera (Coleoptera: Chrysomelidae, Galerucinae) in Europe? Environmental Entomology, 33 1169-1177.
91
Moeser, J., & Vidal, S. (2005). Nutritional resources used by the invasive maize pest Diabrotica virgifera virgifera in its new South-east-European distribution range. Entomologia experimentalis et
applicata, 114(1), 55-63.
Monsanto (2007). Personal Communication.
Muska, F., & Hrudová, E. (2003). Bázlivec kukuricný - mozne nebezpeci pro kukurici. [Maize beetle - Possible risk for maize.]. Uroda, 51(6), 18-19.
Muska, F. (2008). WCR in the Czech Republic - possibilities protection maize. Paper presented at the.
Nagy, B., Szentkirályi, F., & Vörös, G. (1999). Changes in the pest status within maize insect assemblages in the Carpathian Bassin. In E.S. Pego (Eds.), Proceedings of the XIXth Conference of the
International Working Group on Ostrinia (pp. 223-235).
NASS (National Agricultural Statistics Service) (2008). Acreage USDA.
NIS (2008). National Institute for Statistics Romania. Information available at http://www.insse.ro
Nolte, S. (2007, February 23). Personal Communication.
OANDA, OANDA: The currency site, www.oanda.com, 2008.
Oehlert, G.W. (1992). A Note on the Delta Method. The American Statistician, 46(1), 27-29.
Oehmke, J.F., & Crawford, E.W. (2002). The sensitivity of returns to research calculations to supply elasticity. American Journal of Agricultural Economics, 84(2), 366-369.
Pigott, R., & Marra, M. (2008). Biotechnology Adoption Over Time In the Presence of Non-Pecuniary Characteristics that Directly Affect Utility: A Derived Demand Approach. AGbioforum, 11(1), 58-70.
Pilcher, C.D. (2001). Efficacy of MON863 against corn rootworm and comparison to insecticide treatment: results of the year 2000 trials (Monsanto Petition No. 00-CR-015E).
Pimentel, D., McNair, S., Janecka, J., Wightman, J., Simmonds, C., O'Connell, C., Wong, E., Russel, L., Zern, J., Aquino, T., & Tsomondo, T. (2001). Economic and environmental threats of alien plant, animal, and microbe invasions. Agriculture, Ecosystems & Environment, 84(1), 1-20.
Piorin (2008). Polish Inspectorate Plant Health and Seed Protection, personal communication.
Polish Statistics (2008). Agricultural yearbook, Warshaw.
Pratt, J.W. (1964). Risk Aversion in the Small and in the Large. Econometrica, 32(1/2), 122-136.
Qaim, M. (2003). Bt cotton in India: Field trial results and economic projections. World Development, 31(12), 2115-2127.
Qaim, M., Pray, C., & Zilberman, D. (2008). Economic and social considerations in the adoption of Bt crops. In J. Romeis, A.M. Shelton, & G.G. Kennedy (Eds.), Integration of insect-resistant
genetically modified crops within IPM programs (pp. 329-354). Dordrecht, Netherlands: Springer.
92
Rice, M.E., & Oleson, J.D. (2005). Three year summary of corn rootworm control products. Integrated
crop managment, 494(26), 215-217.
Ripka, G. (2008, October 23). Personal Communication.
Rosca,I.(2006). Corn and Diabrotica virgifera virgifera Leconte in Romania. Governatori,G. (Ed.). Integrated Pest Management for corn: a necessity or an opportunity? The case of the Western
Corn rootworm. (pp56-58).Rome,FAO.
Schaafsma, A.W., Baufeld, P., & Ellis, C.R. (1999). Influence of cropping practices on corn rootworm in Canada as a basis for assessment of the potential impact of Diabrotica virgifera in Germany. EPPO Bulletin, 29(1-2), 145-154.
Simic, D., Ivezic, M., Brkic, I., Raspudic, E., Brmez, M., Majic, I., Brkic, A., Ledencan, T., Tollefson, J.J., & Hibbard, B.E. (2008). Environmental and Genotypic Effects for Western Corn Rootworm Tolerance Traits in American and European Maize Trials. Maydica, 52(425), 430.
Sivcev,I.L. (1997). Harmfullness and control of the Western corn rootworm in Yugoslavia. Presented at the XIX IWGO conference, Portugal.
Sivcev, I.L.(2008). Insitute for Plant Protection and Environment of Serbia. Personal Communication.
Sivcev, I., Stankovic, S., Kostic, M., Ladic, N., & Popovic, Z. Population density of Diabrotica virgifera virgifera Leconte beetles in Serbian first year and continous maize fields, Journal of Applied
Entomology, in press.
Skevas, T., Wesseler, J., Fevereiro, P. (2009). Coping with ex-ante regulations for planting Bt maize-The Portuguese experience. AgBioForum ,in press.
State Phytosanitary Administration, Use of crop protection in 2007in the Czech Republic, 2009.
State Statistical Committee Ukraine (2008). Ukrainian yearbook 2007. Kiev, Ukraine
Statistical Office of Serbia (2008).Statistical yearbook. Belgrade, Serbia
Szõke, C., Zsubori, Z., Pók, I., Rácz, F., Illés, O., & Szegedi, I. (2002). Significance of the European corn borer (Ostrinia nubilalis Hübn.) in maize production. Acta Agronomica Hungarica, 50(4), 447-461.
Takacs, J., Balogh, P., and Nadasy, M. (2007). Rapid forecasting of the Western Corn Rootworm (Diabrotica virgifera virgifera LeConte), a pest of sweetcorn in Hungary. Paper presented at the.
Toth, V. (2005). Protection against Western Corn Rootworm Adults (diabrotica vigifera vrigifera) in Banranya county (Hungary). Journal of Central European Agriculture, 6(3), 309-316.
Trnka, M., Muška, F., Semeradova, A., Dubrovsky, M., Kocmankova, E., & Zalud, Z. (2007). European Corn Borer life stage model: Regional estimates of pest development and spatial distribution under present and future climate. Ecological modelling, 207 61-84.
Tuska, T., Kiss, J., Edwards, C.R., Szabo, Z., Ondrusz, I., Miskucza, P., and Garai, A. (2001). Effect of silk feeding by western corn rootworm adruts on yield and quality of seed and commercial corn. Paper presented at the.
93
Urias-Lopez, M., & Meinke, L.J. (2001). Influence of Western Corn Rootworm (Coleoptera: Chrysomelidae) Larval Injury on Yield of Different Types of Maize. Journal of Economic
Entomology, 94(1), 106-111.
USDA (2004). Country report Ukraine. Foreign Agricultural Service, United States Department of agriculture. Available at www.fas.usda.org.
VUZE (2008). Agricultural economics institute of the Czech Republic. Personal communication.
Ward, D., DeGooyer, T., Vaughn, T., Head, G., McKee, M., Astwood, J., & Pershing, J. (2005). Genetically enhanced maize as a potential mangament option for corn rootworm: YieldGard rootworm maize case study. In S. Vidal, U. Kuhlmann, & C.R. Edwards (Eds.), Western Corn
Rootworm: Ecology and Management (pp. 239-262). Wallingford: CABI publishing.
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). Wilson, T.A., Rice, M.E., Tollefson, J.J., & Pilcher, C.D. (2005). Transgenic Corn for Control of the
European Corn Borer and Corn Rootworms: a Survey of Midwestern Farmers; Practices and Perceptions. Journal of Economic Entomology, 98(2), 237-247.
Zorya, S.,& von Cramon-Taubadel ,S. (2006). (When) will Ukraine be a global player on World
agricultural markets. Presented athe the Xth EAAE congress, Zaragoza, Spain.
94
Figures and Tables
Figure 1: The Presence of WCR in Europe in 2007
Source: Gray et al. (2009)
95
Figure 2: First detection of WCR in the maize field by the farmers
Figure 3:The estimated yield loss under no control by Hungarian farmers
0.00%
20.00%
40.00%
60.00%
80.00%
100.00%
120.00%
cum
mu
lati
ve
pre
sen
ce
in farmer perception
Officially
0.0
05
.01
.01
5D
en
sity
-50 0 50 100 150Expected yield loss of WCR under no control(%)
Kernel density estimate
96
Figure 4: Modeled yield loss from WCR in European maize fields
Figure 5: Technology valuation for Bt maize
97
Figure 6: The parameterization of technology valuation.
Figure 7: Spread of WCR in Czech Republic (source State Phytosanitary Insitute,2009)
98
Figure 8: Maize concentration (ha) in the Czech Republic in 2001 (Muska & Hrudová, 2003)
99
Figure 9: Spread of WCR in Slovakia in the first years.
Source: Sivicek(2004)
Figure 10: Occurrence of WCR in Slovakia in 2007 at fields where maize was grown in monoculture
Dots represent the number of WCR adults per yellow trap during one week. Red: more than 90;
yellow 30-90; green 15-29;blue up to 15 and black: no adults
Source: Personal communication Prof. Cagan
100
Figure 11: Percentage of maize in Austrian crop rotation
Source:(Baufeld & Enzian, 2005b)
101
Figure 12: Presence of WCR in Austrian maize fields in 2008
Legend: Bleu means no beetles captured, red means between 100-1000 beetles a trap and dark red
mean 1001-10000 beetles a trap.
Source: Lebensministerium Austria
102
Figure 13: Spread of WCR in Poland in 2007
Source: Piorin (2008)
103
Figure 14: Area with economic damage in the Republic of Serbia (1992-2000)
Source: Sivcev (2008)
Figure 15: Average yield grain maize in Serbia (1991-2007)
Source: Statistical Office of Serbia (2008)
0
10000
20000
30000
40000
50000
60000
199219931994199519961997199819992000
Area with economic WCR damage in Serbia
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
Yield (t/ha)
104
Table 1: Characteristics of the Hungarian maize sector
2003 2004 2005 2006 2007
Grain Maize production 1000 ha 1145 1190 1198 1215 1119 1000 t 4532 8332 9050 8282 4026 kg / ha 3950 7000 7560 6820 3600 Total supply grain maize 7676 10418 14270 16696 14393 Demand for grain maize (1000t) Food & Industrial
Use 564 485 573 582 776
Feed Use 3592 3209 3169 3158 2797 Seed Use 36 39 39 39 36 Export 1311 1320 1940 2447 5356 Waste 112 154 159 154 113 Carry-out Stocks 2062 5213 8391 10316 5314 Silage Maize production 1000 ha 133 108 93 91 145 1000 t 2278 3124 2844 2570 2562 kg / ha 16990 28840 30590 28310 17720
Source: AKII (2008b)
105
Table 2: Data used as input for the Hungarian stochastic bio-economic model
Parameter Value Source
Yield grain (t/ha) Pert (0; 7.70; 10.26) Based on AKII (2008b)
Price Grain (Ft/t) Lognormal(29278;11140) “
Yield silage (t/ha) Pert(0;31.85;42.46) “
Price silage (Ft/t) Lognormal(4874;775) “
Yield loss (%) Betageneral(5.56;10.26;0.62;39.05) Based on Mitchell (2002)
Protection Chem (%) Betageneral(5.27; 11.321; 0.519; 35.852) “
Consistency seed (%) Triangular(0.7;0.8;0.95) Estimates based on Horak et al. (2008)
Consistency GM (%) 1 Ward et al. (2005)
Cost seed (Ft) Uniform(10130;12836) Hatala Zsellér, Ripka, and Vörös (2006), (Monsanto, 2007)
Cost soil (Ft) 17472 “
Price premium (%) Triangular(0;2;3)
Technology fee Triangular(23;33;43)
Refuge 0.2 Agbios (2008)
Gross Margin Maize (Ft) Pert(68627.6;87959.5;94444.6) Based on AKII (2008b)
Gross Margin Barley (Ft) Pert(41350.6;58762.0;73847.9) “
Gross Margin Winter wheat (Ft) Pert(48337.0;67416.8;78270.4) “
Exchange rate (Ft/€) 252 Oanda (2008)
Correlations
Price grain-yield grain -0.2 Goodwin (2008)
Consistency seed-yield loss -0.5 Estimation based on Horak et al. (2008)
106
Table 3: Protection induced rents in the Hungarian maize sector (€/ha)
Risk Neutral CARA
Grain Average Coefficient of variation Average CE % best option
Bt maize 63.0 0.8 54.8 79.9
Seed treatment 29.8 1.4 24.3 1.7
Soil insecticide 6.0 6.8 0.6 0
Crop rotation 47.8 1.3 37.1 18.4
Extra value Bt maize 18.2 0.5 18.5 n.a.
Extra value under monoculture
33.9 0.5 32.3 n.a.
Silage
Bt maize 40.2 0.7 35.1 93
Seed treatment 6.9 3.2 3.7 0.2
Soil insecticide -16.8 1.3 -20.0 0
Crop rotation 14.3 2.6 7.4 6.8
Extra value Bt maize 22.5 0.5 22.7 n.a.
Extra value under monoculture
33.1 0.4 32.2 n.a.
Source: Own calculations
107
Table 4: Regression coefficients of the rents for CARA grain maize producers
Bt maize Excess value Bt Grain Excess value Bt under monoculture
Centered yield 13.44 (0.09)*** -2.06 (0.45)*** 2.08 (0.04)***
Price maize +0.00 (0.00)*** -0.01 (0.00)*** +0.00 (0.00)***
Seed treatment cost -0.07 (0.49) 0.36 (0.03)*** 0.97 (0.23)***
Application cost 0.17 (0.14) -0.10 (0.08) 0.04 (0.07)
Consistency seed treatment 3.88 (3.31) -74.50 (1.74)*** -240.69 (1.56)***
Consistency soil insecticide 0.40 (14.36) -25.12 (7.54)*** -81.44 (6.73)***
Gross margin maize 0.00 (0.00) 0.002 (0.00)*** 0.00 (0.00)
Gross margin barley 0.00 (0.00) -0.0009(0.00)*** 0.00 (0.00)
Gross margin winter wheat 0.00 (0.00) -0.0008(0.00)*** 0.00 (0.00)
Technology fee -0.81 (0.36)*** -0.80 (0.02)*** -0.79 (0.02)***
Damage 6.26 (0.38)*** -1.06 (0.00)*** 0.88 (0.02)***
IP premium -0.11 (0.25) -0.13 (0.13) -0.15 (0.12)
_cons -187.71 (15.06)*** 105.28 (13.27)*** 253.45 (7.08)***
Source: own calculations. *** Significant at the 1% level. Standard deviation between brackets
108
Table 5: Variable definitions and results of the WTP model for WCR resistant maize variety (N=225)
Variable Description Coefficient
Mangament Likert scale 1-5 on the importance in control option decision -0.01
Yield value Likert scale 1-5 on the importance in control option decision 0.17
Insurance Likert scale 1-5 on the importance in control option decision -0.06
Personal health Likert scale 1-5 on the importance in control option decision -0.23
Equipment Likert scale 1-5 on the importance in control option decision 0.25**
Environment Likert scale 1-5 on the importance in control option decision -0.2
IPM Experience Integrated Pest management (1=yes,0=no) 0.63***
Danube 1 if the farmer is located in the Danube valley, 0 otherwise 0.19
Plain 1 if the farmer is located in the Great Plain, 0 otherwise 1.15**
Lease The percentage of the cultivated area leased -0.05
Maize The percentage of the cultivated area planted with maize 0.64
Livestock 1 if the farmer has livestock, 0 otherwise 0.21
Price grain Anticipated farmgate price for 2008 season for grain maize (HUF/ton) 0.0002***
Price silage Anticipated farmgate price for 2008 season for silage maize (HUF/ton) 0.0003***
Grain monculture Percentage of grain maize area under monoculture 0.01***
Yield grain Expected yield in 2008 (ton/ha) for grain maize -0.11*
Yield Silage Expected yield in 2008 (ton/ha) for silage maize -0.0004***
Education 1 if university education, 0 otherwise -0.13
Age Age of respondent 0.01
Infestation Years since first detection in field (2009) 0.04
Future damage 1 if farmer expected damage in the future, 0 otherwise -0.18
Off farm income percentage of income from off farm employment -0.03
Intercept -2.74**
Log-likelihood -180.3
chi squared 96.9**
Estimated mean WTP Average WTP for the proposed variety (€/ha) 70.3
Note: the estimated coefficients can be interpreted as the marginal effect on WTP in HUF/ha *,**,*** Statistically significant at the 0.10, 0.05 and 0.01 level respectively
109
Table 6: Major information source on insect control (%)
Other farmers 8.0
Representatives of manufacturers 9.5
Dealers 3.1
Independent advisors 5.5
Public advisors (chamber, authorities) 4.7
Advertising (direct mailing, brochures) 3.8
Internet 5.8
Agricultural magazines 19.1
Lectures 10.6
Events/trade fairs (public) 9.8
Own experience 9.1
Market prices 0.9
Professional literature 7.3
Television/radio 0.2
Training courses 1.6
Others/don't know/no answer 1.1
Total 100
Table 7:Confidence intervals on the contribution of significant variables to WTP
Mean contribution WTP
2.5% 97.5%
Equipment 5037.0 -3671.5 13745.5
IPM 12669.9 -6182.9 31522.8
Plain 22907.0 -13131.9 58946.0
Price grain 0.5 -0.3 1.4
Price silage 7.1 -3.0 17.2
Yield grain -2160.6 -6191.0 1869.8
Yield silage -8.7 -21.3 3.8
Grain monoculture 224.4 -150.1 598.9
110
Table 8: Czech Republic maize production sector
2003 2004 2005 2006 2007 2008
Grain Maize production 1000 ha 85.4 89.9 98.0 89.8 111.7 107.9 1000 t 476 551 703 606 759 792 t / ha 5.58 6.13 7.17 6.75 6.80 7.35 Total supply grain maize (1000t) 649 661 838 739 729 n.a. Silage Maize production 1000 ha 207 213 192 185 162 180 1000 t 5707 6462 6870 6066 5570 6343 t / ha 27.5 30.3 35.7 32.7 34.4 35.3
Source:Eurostat (2009)
Table 9: Protection induced rents in the Czech maize sector (€/ha)
Risk Neutral CARA
Grain Average Coefficient of variation Average CE % best option
Bt maize 89.2 0.5 80.9 50.1
Seed treatment 41.6 1.0 36.1 0.3
Soil insecticide 17.8 2.3 12.4 0
Crop rotation 107.2 0.6 37.1 49.7
Extra value Bt maize
for adopters
11 0.7 12 n.a.
Extra value under monoculture
48 0.4 46 n.a.
Silage
Bt maize 60.6 0.5 54.7 45.1
Seed treatment 13.2 1.9 9.7 0
Soil insecticide -10.5 2.3 -14.1 0
Crop rotation 65.8 0.6 57.9 54.9
Extra value Bt maize for adopters
14 0.9 15 n.a.
Extra value under monoculture
47 0.3 42 n.a.
Source: Own calculations
111
Table 10: Slovakian maize production sector
2003 2004 2005 2006 2007 2008
Grain Maize production 1000 ha 150 140 152 153 157 153 1000 t 601 862 1074 838 623 1148 t / ha 4.0 5.9 7.1 5.5 4.0 7.5 Total supply grain maize (1000t) 754 601 862 1074 838 624(p) Silage Maize production 1000 ha 98 96 89 85 79 77(p) 1000 t 2101 2293 2324 1996 1839 n.a. t / ha 21.5 24.0 26.2 23.5 23.2 n.a. Maize in monoculture % n.a. 13 20 17 15 n.a.
Source: Eurostat (2009), Cagan (2008) (p-=preliminary estimations)
Table 11: Protection induced rents in the Slovakian maize sector (€/ha)
Risk Neutral CARA
Grain Average Coefficient of variation Average CE % best option
Bt maize 35.8 0.9 29.5 90
Seed treatment 18.9 1.4 13.8 10
Soil insecticide -5.0 4.1 -10.0 0
Crop rotation -75.5 0.6 -84.7 0
Extra value Bt maize 20.0 0.6 19.1 n.a.
Silage
Bt maize 14.1 1.8 9.2 97
Seed treatment -2.2 5.7 -6.2 3
Soil insecticide -26 0.7 -30.0 0
Crop rotation -106.0 0.4 -113 0
Extra value Bt maize 16.3 0.9 17.6 n.a.
Source: Own calculations
112
Table 12: Maize producing sector in Austria
2003 2004 2005 2006 2007 2008
Grain Maize production 1000 ha 173 179 167 159 171 194 1000 t 1452 1654 1725 1472 169 2147 t / ha 8.4 9.2 10.3 9.2 9.9 11.1 Total supply grain maize (1000t) 1956 1708 1945 2021 1746 n.a. Silage Maize production 1000 ha 72 75 77 79 80 81 1000 t 3026 3374 3599 3546 3741 3808
(p) t / ha 41.9 44.6 46.7 45.1 46.6 47.0 (p)
Source: Eurostat(2009)
Table 13:Technology induced rents Austria (€/ha)
Risk Neutral CARA
Grain Average Coefficient of variation Average CE % best option
Bt maize 112.3 0.7 99.7 98
Seed treatment 62.5 1.0 54.2 2
Soil insecticide 45.6 1.3 37.4 0
Crop rotation -20 5 -37.9 0
Extra value Bt maize 46.9 0.5 44.5 n.a.
Silage
Bt maize 99.3 0.6 90 99
Seed treatment 47.2 0.9 41.4 1
Soil insecticide 30.3 1.4 24.5 0
Crop rotation -43.2 1.7 -56.6 0
Extra value Bt maize 51.1 0.4 49.3 n.a.
Source: Own calculations
113
Table 14:Maize producing sector in Poland
2003 2004 2005 2006 2007 2008
Grain Maize production
1000 ha 356 412 339 303 262 317 (p) 1000 t 1884 2344 1945 1261 1722 1844 (p) t / ha 5.3 5.7 5.7 4.2 6.6 5.8 (p) Total imports (1000t) 208 369 232 290 1085 663 Silage Maize production
1000 ha 239 289 326 356 368 410 (p) 1000 t 9581 12099 12741 11539 17491 16656 (p) t / ha 40.1 41.7 39.1 32.4 47.6 40.6 (p)
Source: Eurostat (2009)
Table 15: Technology induced rents for the Polish maize sector (€/ha)
Risk Neutral CARA
Grain Average Coefficient of variation Average CE % best option
Bt maize 34.0 0.8 28.9 100
Seed treatment 8.5 2.5 5.3 0
Soil insecticide -15.2 1.4 0.6 0
Crop rotation -106.8 0.4 -116.2 0
Extra value Bt maize 25.5 0.5 24.4 n.a.
Silage
Bt maize 36 0.7 31.0 100
Seed treatment 5.2 3.2 2.1 0
Soil insecticide -18.5 1.3 -21.6 0
Crop rotation -111.7 2.6 -110.9 0
Extra value Bt maize 30.8 0.4 29.6 n.a.
Source: Own calculations
Table 16:Maize producing sector in the Republic of Serbia
2003 2004 2005 2006 2007 2008
Grain Maize production
1000 ha 1200 1200 1220 1170 1202 1274 1000 t 3817 6569 7086 6017 3905 6158 t / ha 3.2 5.5 5.8 5.1 3.2 4.8
Yield in Vojvodina t/ha 3.4 5.9 6.5 5.9 n.a. n.a. Silage Maize production
1000 ha 24 23 23 23 26 25 1000 t 356 489 486 502 460 459 t / ha 14.8 20.9 21.1 21.6 17.5 18.1
Source: Statistical office of Serbia (2008)
114
Table 17:Technology induced rents in the Serbian maize sector (€/ha)
Risk Neutral CARA
Grain Average Coefficient of variation Average CE % best option
Bt maize 26.9 1.0 22.2 66.5
Seed treatment -0.7 32 -3.7 0.2
Soil insecticide -24.5 0.9 -27.5 0
Crop rotation 11.3 5.2 5.2 33.3
Extra value Bt maize for adopters
21.6 0.5 22.3 n.a.
Extra value under monoculture
26 0.4 27 n.a.
Source: Own calculations
Table 18: The maize producing sector in Romania
2003 2004 2005 2006 2007 2008
Grain Maize production
1000 ha 3159 3239 2592 2484 2525 2460(p) 1000 t 9577 14541 10389 8985 3854 7923(p) t / ha 3.0 4.5 4.0 3.6 1.5 3.2 (p)
Net trade balance 1000t 24 232 -418 -274 30.0 395 Silage Maize production
1000 ha 39 34 24 27 46 27 1000 t 546 654 521 533 650 n.a. t / ha 14.2 19.1 21.4 19.5 14.1 14.2
Source: Eurostat (2008)
Table 19: Regional heterogenity in the Romanian maize sector
Total arable
land Maize
% maize in
rotation
% of maize
area sown
Romania 7883954 2520098 32% 100% North East 1172638 491195 42% 19% South East 1555410 509590 33% 20% South-Muntenia 1784510 473985 27% 19% South- West Oltenia 1055581 348359 33% 14% West 851934 282335 33% 11% North-West 798329 251281 31% 10% Center 584846 142018 24% 6% Bucharest 80706 21335 26% 1%
Source: NIS (2008)
115
Table 20: Technology induced rents in the Romanian agriculture (€/ha)
Risk Neutral CARA
Grain Average Coefficient of variation Average CE % best option
Bt maize 30.7 1.1 25.2 98.4
Seed treatment 5.19 5.1 1.7 1.3
Soil insecticide -18.5 1.4 -22.1 0
Crop rotation -60.8 0.9 -71.7 0.2
Extra value Bt maize for adopters
24.1 0.5 23.6 n.a.
Extra value under monoculture
25.5 0.4 24.2 n.a.
Silage
Bt maize -9.1 1.3 -10.9 100
Seed treatment -30 0.3 -31.8 0
Soil insecticide -54.4 0.2 -55.5 0
Crop rotation -114 0.4 -121.8 0
Source: Own calculations
Table 21: Maize producing sector in the Ukraine
2002 2003 2004 2005 2006
Private firms
Grain Maize production
1000 ha 77.3 154.8 231.8 122.6 146.2 1000 t 214.9 414.5 690.5 409.5 388.3 t / ha 2.8 2.7 3.0 3.3 2.7 Share of total production % 5.1 6.0 7.8 5.7 6.0 Silage Maize production
1000 ha 39.6 44.1 33.8 28.6 29.0 1000 t 463.0 552.2 432.1 412.6 388.8 t / ha 11.7 12.5 12.8 14.4 13.4 Share of total production % 2.7 2.9 3.0 3.3 3.8
Other structures
Grain Maize production
1000 ha 1110 1833 2068 1537 1573 1000 t 4213 6908 8852 7184 6471 t / ha 3.5 3.5 3.9 4.3 3.7 Silage Maize production
1000 ha 1466 1520 1126 1041 763 1000 t 17148 19041 14403 12503 10231 t / ha 10.0 10.0 10.7 12.0 n.a.
Source: State statistics committee Ukraine (2008), FAOstat (2008).
116
Table 22: Technology induced rents in Ukrainian agriculture
Risk Neutral CARA
Grain Average Coefficient of variation Average CE % best option
Bt maize -5.0 2.0 -6.8 63.7%
Seed treatment -27.7 0.3 -28.9 0
Soil insecticide -51.2 0.1 -52.6 0
Crop rotation -11.9 2.6 -18.3 36.3%
Extra value Bt maize for adopters
17.2 0.5 17.4 n.a.
Monoculture 23.0 0.4 22.1 n.a.
Silage
Bt maize -13.4 0.5 -14.8 68.2%
Seed treatment -36.1 0.1 -36.9 0
Soil insecticide -59.8 0.1 -60.5 0
Crop rotation -24.4 1.3 -30.3 31.2%
Extra value Bt maize for adopters
18.2 0.6 17.9 n.a.
Monoculture 23.0 0.4 22.0 n.a.
Source: own calculations
117
Table 23: Data used for the simulation model
Czech Republic Slovakia Austria
Grain yield Pert(0; 6.8; 9.1) Eurostat Pert(0;8.3;11.1) Eurostat Pert(0; 10.7566; 14.342) Eurostat
Silage yield Pert(0; 38.3; 51.1) Eurostat Pert(0;27.6;36.8) Eurostat Pert(0; 46.2; 65.3) Eurostat
Price grain Lognorm(150.7;43.1) VUZE Lognorm(89;4.8) Statistical office Slovakia + Eurostat
Lognorm(127.2;49.76) Landwirtschaftskammer
Price silage Lognorm(18.0;2.9) Fall, Wesseler
Lognorm(28.4;4.8) Fall Wesseler Lognorm(25.1;5) Fall Wesseler
Gross margin maize Uniform(15;108) Brookes Pert(260.25;347;433.75) Brookes Pert(-83;69;216) Brookes
Gross margin wheat Uniform (10;48) Brookes Pert(81;108;135) Brookes Pert(-171;-165;-147) Brookes
Gross margin barley Uniform(16;51) Brookes Pert(60.75;81;101.25) Brookes Pert(-271;-269;-260) Brookes
Gross margin oilseed rape
n.a. n.a. n.a.
Techfee Triang(7;17;27) Triang(39.5;49.5;59.5) Triang(19.8;29.8;39.8)
Triang(23;33;43)
Exchange rate (€1) 28.32
38.9
n.a.
Poland Serbia Romania Ukraine
Pert(0; 6.205; 8.27) Eurostat Pert(0; 5.9; 8.85) Statitics Office of Serbia
Pert(0; 3.9; 5.85 Eurostat Pert(0; 3.7; 5.55) FAOstat
Pert(0; 47.91; 63.88) Eurostat n.a. Pert(0; 19.2; 28.8) Eurostat Pert(0; 12; 18) FAOstat
Lognorm(102;7.7) Eurostat Lognorm(87.76;30.99 Statitics Office of Serbia
Lognorm(149.7;49.2) Eurostat Lognorm(55;8.08) FAOstat
Lognorm(12.5;2.862) Fall, Wesseler
n.a. Lognorm(9.1;4.5) Schaafsma Lognorm(9.1;3) Schaafsma
Pert(363;489;626) Brookes Pert(170;340;510) FAO(2004) Pert(156;303;450) Brookes Pert(104;208;312) FAO(2005)
Pert(186;229;240) Brookes Soya= Pert(110.5;221;331.5) FAO(2004) Pert(58;116;174) Brookes Pert(87.5;175;262.5) FAO(2005)
Pert(118;131;172) Brookes Sunflower=Pert(148;296;444) FAO(2004) Pert(32;64;96) Brookes Pert(63.25;126;189.25) FAO(2005)
Pert(268;277;316) Brookes n.a. n.a. n.a.
Triang(27.6;37.6;47.6) Triang(23;33;43) Triang(26.7;36.7;46.7) Triang(23;33;43)
82
3.38
7.79
118
Table 24: Average effects of introducing Bt maize resistant to WCR in Central European countries (€/ha)
Technology fee
Protection Bt Maize
Extra value Bt
maize
Extra value Bt in monoculture
Bt maize optimal
control (%)
Total welfare created land
constrained
Total welfare created
monoculture
Hungary
Grain 33 63 18 34 80 51 67
Silage 33 40 22 33 93 55 66
Czech Republic
Grain 17 89 -18 48 50 -1 65
Silage 17 61 -5 47 45 12 64
Slovakia
Grain 50 36 20 20 90 70 70
Silage 50 14 16 16 97 66 66
Austria
Grain 30 112 47 47 98 77 77
Silage 30 99 51 51 99 81 81
Poland
Grain 38 34 26 26 100 64 64
Silage 38 36 31 31 100 69 69
Serbia
Grain 33 27 22 26 67 55 59
Romania
Grain 37 31 24 25 83 61 62
Silage 37 -9 4 8 18 41 45
Ukraine
Grain 33 -5 17 23 17 50 56
Silage 33 -13 18 23 14 51 56
Source: Own calculations
119
List of Available Working Papers
1. BEERLANDT, H. en L. DRIESEN, Criteria ter evaluatie van 'duurzame landbouw', Afdeling Landbouweconomie, Katholieke Universiteit Leuven, januari 1994, 35 p.
2. BEERLANDT, H. en L. DRIESEN, Evaluatie van herbicide-resistente planten aan criteria voor duurzame
landbouw, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, januari 1994, 39 p.
3. BEERLANDT, H. en L. DRIESEN, Evaluatie van bovine somatotropine aan criteria voor duurzame landbouw, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, januari 1994, 63 p.
4. BEERLANDT, H. en L. DRIESEN, Evaluatie van gemanipuleerde planten met biopesticide eigenschappen
afkomstig van Bacillus thuringiensis aan criteria voor duurzame landbouw, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, januari 1994, 32 p.
5. BEERLANDT, H. en L. DRIESEN, Evaluatie van haploide planten aan criteria voor duurzame landbouw,
Afdeling Landbouweconomie, Katholieke Universiteit Leuven, januari 1994, 17 p.
6. BEERLANDT, H. en L. DRIESEN, Evaluatie van genetische technieken voor diagnosebepaling,
immunologische technieken ter verbetering van de landbouwproduktie en transgene dieren en planten als
bioreactor aan criteria voor duurzame landbouw, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, januari 1994, 28 p.
7. BEERLANDT, H. en L. DRIESEN, Evaluatie van verbetering van de stikstoffixatie bij planten aan criteria voor
duurzame landbouw, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, januari 1994, 17 p.
8. BEERLANDT, H. en L. DRIESEN, Evaluatie van porcine somatotropine aan criteria voor duurzamelandbouw,
Afdeling Landbouweconomie, Katholieke Universiteit Leuven, januari 1994, 29 p.
9. BEERLANDT, H. en L. DRIESEN, Evaluatie van tomaten met een langere houdbaarheid aan criteria voor
duurzame landbouw, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, februari 1994, 30 p.
10. CHRISTIAENSEN, L., Voedselzekerheid: van concept tot actie: een status questionis, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, april 1994, 106 p.
11. CHRISTIAENSEN, L. and J. SWINNEN, Economic, Institutional and Political Determinants of Agricultural
Production Structures in Western Europe, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, May 1994, 40 p.
12. GOOSSENS, F., Efficiency and Performance of an Informal Food Marketing System, The case of Kinshasa,
Zaire, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, July 1995, 41 p.
13. GOOSSENS, F., Failing Innovation in the Zairian Cassava Production System, A comparative historical
analysis, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, July 1995, 18 p.
14. TOLLENS, E., Cadre conceptuel concernant l'analyse de la performance économique des marchés, Projet-FAO "Approvisionnement et Distribution Alimentaires des Villes de l'Afrique Francophone", Afdeling Landbouweconomie, Katholieke Universiteit Leuven, août 1995, 35 p. (Deuxième version, avril 1996, 77 p.)
15. TOLLENS, E., Les marchés de gros dans les grandes villes Africaines, diagnostic, avantages et éléments
d'étude et de développement, Projet-FAO "ApprovisioMement et Distribution Alimentaires des Villes de l'Afrique Francophone", Afdeling Landbouweconomie, Katholieke Universiteit Leuven, août 1995, 23 p. (Deuxieme version, septembre 1996, 32 p.)
16. ENGELEN, G., Inleiding tot de landbouwvoorlichting (heruitgave), Afdeling Landbouweconomie, Katholieke Universiteit Leuven, augustus 1995, 17 p.
17. TOLLENS, E., Agricultural Research and Development towards Sustainable Production Systems: I.
Information Sources, Surveys; II. Conceptualisation of the Change Process, NATURA-NECTAR course: "Agricultural Economics and Rural Development", module 1, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, August 1995
18. TOLLENS, E., Planning and Appraising Agricultural Development programmes and Projects: I. Farm
Planning; II. Aggregation, Sensitivity Analyses and Farm Investment Analysis; III. Guidelines on Informal
120
Surveys and Data Collection, NATURA-NECTAR course: "Agricultural Economics and Rural Development", module 2, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, September 1995
19. TOLLENS, E., Structural Adjustment and Agricultural Policies: I. Market Theory: the State and the Private
Sector; II. Output Markets and Marketing Institutions; III. Input Markets; IV. Case Study: Cameroon,
NATURA-NECTAR course: "Agricultural Economics and Policy Reforms", module 1, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, September 1995
20. TOLLENS, E., Theory and Macro-Economic Measures of Structural Adjustment – Methods of Evaluation and
Linkages to the Agricultural Sector: I. Development Models and the Role of Agriculture, NATURA-NECTAR course: "Agricultural Economics and Policy Reforms", module 2, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, September 1995
21. TOLLENS, E., Theory and Macro-Economic Measures of Structural Adjustment – Methods of Evaluation and
Linkages to the Agricultural Sector: II. Implementation of Policy Reforms: Case Study of Market Liberalisation
in Cameroon for Cocoa and Coffee, NATURA-NECTAR course: "Agricultural Economics and Policy Reforms", module 2, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, September 1995
22. TOLLENS, E., Supply Response within the Farming Systems Context: I. Input Supply and Product Markets; II.
Agricultural Supply Response Assessment, NATURA-NECTAR course: "Agricultural Economics and Policy Reforms", module 3, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, September 1995
23. GOOSSENS, F., Agricultural Marketing and Marketing Analysis: I. Agricultural Marketing Research
Frameworks. II. Agricultural Market Performance Criteria and The Role of Government Intervention,
NATURA-NECTAR course: "Agricultural Economics and Rural Development", module 3, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, September 1995
24. GOOSSENS, F., Agricultural Marketing and Marketing Analysis: Demand Analysis, NATURA-NECTAR course: "Agricultural Economics and Rural Development", module 3, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, September 1995
25. CHRISTIAENSEN, L. en H. BEERLANDT, Belgische voedselhulp geanalyseerd met betrekking tot
voedselzekerheid, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, november 1994, 15 p.
26. CHRISTIAENSEN, L. en H. BEERLANDT, De Belgische ontwikkelingssamenwerking met Rwanda
geanalyseerd met betrekking tot voedselzekerheid, Afdeling Landbouweconomie, KU.Leuven, november 1995, 36 p.
27. BEERLANDT, H., Identificatie van de meest kwetsbaren in Monduli distrikt, Arusha regio, Tanzania, A.C.T.- Afdeling Landbouweconomie, Katholieke Universiteit Leuven, april 1995, 40 p.
28. BEERLANDT, H., TOLLENS, E. and DERCON, S., Methodology for Addressing Food Security in
Development Projects, Identification of the Food Insecure and the Causes of Food Insecurity based on
Experiences from the Region of Kigoma, Tanzania, Department of Agricultural Economics and Centre for Economic Research, Katholieke Universiteit Leuven, Leuven, December 1995, 19 p.
29. BEERLANDT, H., Koppelen van noodhulp en strukturele ontwikkelingssamenwerking: opties voor een Belgisch
beleid, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, december 1995, 23 p.
30. TOLLENS, E., La crise agraire au Zaïre: pour quelle politique de développement dans la phase de transition?,
Une contribution au colloque “Le Zaïre en Chantier: Quels Projets de Société”, Anvers, 18 février 1993, December 1995, 14 p.
31. GOOSSENS, F., Rôle des systèmes d'alimentation dans la sécurité alimentaire de Kinshasa, Une contribution au projet GCP/RAF/309, AGSM, FA0, mai 1996, 78 p.
32. BEERLANDT, H., DERCON, S., and SERNEELS, I., (Project co-ordinator: E. TOLLENS), Tanzania, a Food
Insecure Country?, Department of Agricultural Economics, Center for Economic Research, Katholieke Universiteit Leuven, September 1996, 68 p.
33. TOLLENS, E., Food security and nutrition 2. Case study from Tanzania, Nectar Programme, Agricultural Economics and Policy Reforms, module 4, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, Septembre 1996, 47 p.
121
34. BEERLANDT, H., en SERNEELS, J., Voedselzekerheid in de regio Kigoma, Tanzania, Afdeling Landbouweconomie en Centrum voor Economische Studiën, Katholieke Universiteit Leuven, september 1996, 45 p.
35. BEERLANDT, H., Identificatie van verifieerbare indicatoren ter toetsing van de voedselzekerheidssituatie in de
regio Arusha, Tanzania, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, november 1996, 60 p.
36. GOOSSENS, F., Commercialisation des vivres locaux en Afrique Subsaharienne, le secteur informel dans un
perspectif dynamique, Une contribution au projet GCP/RAF/309, AGSM, FAO, novembre 1996, 58 p.
37. GOOSSENS, F., The Economics of Livestock Systems: I. Marketing Problems and Channels of Livestock in
Subsahara Africa, NATURA-NECTAR course: "Agricultural Economics and Rural Development", module 4, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, November 1996.
38. GOOSSENS, F., The Economics of Livestock Systems: II. Price Stabilization in the Livestock Sector,
NATURA-NECTAR course: "Agricultural Economics and Rural Development", module 4, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, November 1996.
39. GOOSSENS, F., The Economics of Livestock Systems: III. Consumer Demand for Livestock Products,
NATURA-NECTAR course: "Agricultural Economics and Rural Development, module 4, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, November 1996.
40. JASPERS, N., I. La Seguridad Alimenticia en el departamento de Quiché: Identificación e Impacto del
Programa de Créditos, II. Informe Sobre Estudio Seguridad Alimenticia, ACT - Afdeling Landbouweconomie, Katholieke Universiteit Leuven, November 1996, 39 p.
41. TOLLENS, E., Social indicators with an illustration from Thailand, NATURA-NECTAR course: "Agricultural Economics and Policy Reforms", module 4, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, January 1997, 38 p.
42. BEERLANDT, H., en SERNEELS, J., Handleiding voor een voedselzekerheidsdiagnose, Afdeling Landbouweconomie en Centrum voor Economische Studiën, Katholieke Universiteit Leuven, februari 1997, 131 p.
43. BEERLANDT, H., and SERNEELS, J., Manual for a Food Security Diagnosis, Department of Agricultural Economics and Center for Economic Research, Katholieke Universiteit Leuven, March 1997, 125 p.
44. GOOSSENS, F., Aangepaste vormen van samenwerking als hefboom voor de sociaal-economische promotie van
boeren in het zuiden - algemene conclusies, Seminarie georganizeerd door Ieder Voor Allen, Brussel, 17-18 maart 1997, 8 p.
45. GOOSSENS, F., Commercialisation des vivres locaux en Afrique Subsaharienne - neuf études de cas, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, Mai 1997, 50 p.
46. BEERLANDT, H., en SERNEELS, J., Food Security in the Kigoma Region of Tanzania, Department of Agricultural Economics and Center for Economic Research, Katholieke Universiteit Leuven, May 1997, 42 p.
47. BEERLANDT, H., and SERNEELS, J., Manuel Pour un Diagnostic de Securité Alimentaire, Département d’Economie Agricole et le Centre d’Etudes Economiques, Katholieke Universiteit Leuven, Juillet 1997, 134 p.
48. GOOSSENS, F., Rural Services and Infrastructure - Marketing Institutions, NATURA-NECTAR course: "Agricultural Economics and Policy Reforms", module 4, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, June 1997, 20 p.
49. TOLLENS, E., International Trade and Trade Policy in Livestock and Livestock Products, NATURA-NECTAR COURSE: "Agricultural Economics and Rural Development", module 4, Afdeling Landbouweconomie, Katholieke Universiteit Leuven, October 1997,43 p.
50. DESMET, A., Working towards autonomous development of local farmer organisations: which role for
development agencies?, Department of Agricultural Economics and Center for Economic Research, March 1998, 49 p.
51. TOLLENS, E., Catalogue de titres dans la bibliotheque ALEO sur le Zaïre - Congo, Département d'Economie Agricole, Katholieke Universiteit Leuven, Mars 1998, 96 p.
122
52. DEMONT, M., JOUVE, P., STESSENS, J., et TOLLENS, E., Evolution des systèmes agraires dans le Nord de
la Côte d’Ivoire: les débats « Boserup versus Malthus » et « compétition versus complémentarité » révisités, Département d’Economie Agricole et de l’Environnement, Katholieke Universiteit Leuven, Avril 1999, 43 p.
53. DEMONT, M., and TOLLENS, E., The Economics of Agricultural Biotechnology: Historical and Analytical
Framework, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, October 1999, 47 p.
54. DEMONT, M., en TOLLENS, E., Biologische, biotechnologische en gangbare landbouw : een vergelijkende
economische studie, Afdeling Landbouw- en Milieueconomie, Katholieke Universiteit Leuven, Maart 2000, 53 p.
55. DEMONT, M., JOUVE, P., STESSENS, J., and TOLLENS, E., The Evolution of Farming Systems in Northern
Côte d’Ivoire: Boserup versus Malthus and Competition versus Complementarity, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, August 2000, 25 p.
56. DEMONT, M., and TOLLENS, E., Economic Impact of Agricultural Biotechnology in the EU: The EUWAB-
project, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, January 2001, 16 p.
57. DEMONT, M., and TOLLENS, E., Reshaping the Conventional Welfare Economics Framework for Estimating
the Economic Impact of Agricultural Biotechnology in the European Union, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, March 2001, 32 p.
58. DEMONT, M., and TOLLENS, E., Uncertainties of Estimating the Welfare Effects of Agricultural
Biotechnology in the European Union, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, April 2001, 81 p.
59. DEMONT, M., and TOLLENS, E., Welfare Effects of Transgenic Sugarbeets in the European Union: A
Theoretical Ex-Ante Framework, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, May 2001, 39 p.
60. DE VENTER, K., DEMONT, M., and TOLLENS, E., Bedrijfseconomische impact van biotechnologie in de
Belgische suikerbietenteelt, Afdeling Landbouw- en Milieueconomie, Katholieke Universiteit Leuven, Juni 2002, 65 p.
61. DEMONT, M., and TOLLENS, E., Impact of Agricultural Biotechnology in the European Union’s Sugar
Industry, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, June 2002, 55 p.
62. DEMONT, M., and TOLLENS, E., The EUWAB-Project: Discussion, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, August 2002, 20 p.
63. DEMONT, M., DELOOF, F. en TOLLENS, E., Impact van biotechnologie in Europa: de eerste vier jaar Bt
maïs adoptie in Spanje, Afdeling Landbouw- en Milieueconomie, Katholieke Universiteit Leuven, Augustus 2002, 41 p.
64. TOLLENS, E., Food Security: Incidence and Causes of Food Insecurity among Vulnerable Groups and Coping
Strategies, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, September 2002, 30 p.
65. TOLLENS, E., La sécurité alimentaire: Incidence et causes de l’insécurité alimentaire parmi les groupes
vulnérables et les strategies de lutte, Département d’Economie Agricole et de l’Environnement, Katholieke Universiteit Leuven, Septembre 2002, 33 p.
66. TOLLENS, E., Food Security in Kinshasa, Coping with Adversity, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, September 2002, 35 p.
67. TOLLENS, E., The Challenges of Poverty Reduction with Particular Reference to Rural Poverty and
Agriculture in sub-Saharan Africa, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, September 2002, 31 p.
68. TOLLENS, E., Het voedselvraagstuk, Afdeling Landbouw- en Milieueconomie, Katholieke Universiteit Leuven, December 2002, 59 p.
123
69. DEMONT, M., WESSELER, J., and TOLLENS, E., Biodiversity versus Transgenic Sugar Beet: The One Euro
Question, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, November 2002, 33 p.
70. TOLLENS, E., and DEMONT, M., Biotech in Developing Countries: From a Gene Revolution to a Doubly
Green Revolution?, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, November 2002, 8 p.
71. TOLLENS, E., Market Information Systems in Liberalized African Export Markets: The Case of Cocoa in Côte
d’Ivoire, Nigeria and Cameroon, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, November 2002, 19 p.
72. TOLLENS, E., Estimation of Production of Cassava in Bandundu (1987-1988) and Bas Congo (1988-1989)
Regions, as Compared to Official R.D. Congo statistics, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, December 2002, 29 p.
73. TOLLENS, E., Biotechnology in the South: Absolute Necessity or Illusion?, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, December 2002, 29 p.
74. DEMONT, M., BONNY, S., and TOLLENS, E., Prospects for GMO’s in Europe, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, January 2003.
75. FRANCHOIS, L., and MATHIJS, E., Economic and Energetic Valuation of Farming Systems: A Review, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, February 2003, 36 p.
76. VANDERMERSCH, M. en MATHIJS, E., Performantie en bedrijfsprofiel in de melkveehouderij, Afdeling Landbouw- en Milieueconomie, Katholieke Universiteit Leuven, Februari 2003, 33 p.
77. TOLLENS, E., L’état actuel de la sécurité alimentaire en R.D. Congo : Diagnostic et perspectives, Département d'Economie Agricole et de l'Environnement, Katholieke Universiteit Leuven, Février 2003, 40p.
78. VANDERMERSCH, M., MESKENS, L. en MATHIJS, E., Structuur van de Belgische melkveehouderij, Afdeling Landbouw- en Milieueconomie, Katholieke Universiteit Leuven, Februari 2003, 60 p.
79. DEMONT, M., HOUEDJOKLOUNON, A., HOUNHOUIGAN, J., MAHYAO, A., ORKWOR, G., STESSENS, J., TOLLENS, E. et TOURÉ, M., Etude comparative des systèmes de commercialisation d’igname en Côte-
d’Ivoire, au Bénin et au Nigeria, Département d'Economie Agricole et de l'Environnement, Katholieke Universiteit Leuven, Juin 2003, 30 p.
80. TOLLENS, E., Current Situation of Food Security in the D.R. Congo: Diagnostic and Perspectives, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, August 2003, 37 p.
81. TOLLENS, E., Poverty and Livelihood Entitlement, How It Relates to Agriculture, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, August 2003, 30 p.
82. TOLLENS, E., Sécurité alimentaire à Kinshasa: un face à face quotidien avec l’adversité, Département d'Economie Agricole et de l'Environnement, Katholieke Universiteit Leuven, Septembre 2003, 33 p.
83. DEMONT, M. and TOLLENS, E., Impact of Biotechnology in Europe: The First Four Years of Bt Maize
Adoption in Spain, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, October 2003, 28 p.
84. TOLLENS, E., Fair Trade: An Illusion?, Department of Agricultural and Environmental Economics, Katholieke Universiteit Leuven, October 2003, 17 p.
85. TOLLENS, E., DEMONT, M. and SWENNEN, R., Agrobiotechnology in Developing Countries: North-South
Partnerships are a Key, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, December 2003, 29 p.
86. TOLLENS, E., Les défis : Sécurité alimentaire et cultures de rente pour l’exportation – Principales orientations
et avantages comparatifs de l’agriculture en R.D. Congo, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, Mars 2004, 67 p.
124
87. DEMONT, M., JOUVE, P., STESSENS, J. et TOLLENS, E., Boserup versus Malthus revisités: Evolution des
exploitations agricoles dans le Nord de la Côte d’Ivoire, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, Avril 2004, 20 p.
88. DEMONT, M., JOUVE, P., STESSENS, J. and TOLLENS, E., Boserup versus Malthus Revisited: Evolution of
Farms in Northern Côte d’Ivoire, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, April 2004, 17 p.
89. DEMONT, M., OEHMKE, J. F. and TOLLENS, E., Alston, Norton and Pardey Revisited: The Impact of Bt
Maize in Spain, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, April 2006, 31 p.
90. VANDENBOSCH, T., NANOK, T. and TOLLENS, E., The Role of Relevant Basic Education in Achieving
Food Security and Sustainable Rural Development, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, August 2004, 23 p.
91. VANDERMERSCH, M. and MATHIJS, E., Consumer willingness to pay for domestic milk, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, October 2004, 34 p.
92. DEMONT, M., TOLLENS, E. and FOGARASI, J., Potential Impact of Biotechnology in Eastern Europe:
Transgenic Maize, Sugar Beet and Oilseed Rape in Hungary, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, January 2005, 63 p.
93. DAEMS, W., DEMONT, M., MUŠKA, F., SOUKUP, J. and TOLLENS, E., Potential impact of biotechnology
in Eastern Europe: Genetically modified maize, sugar beet and oilseed rape in the Czech Republic, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, February 2007, 68 p.
94. TOLLENS, E., Manual on Cocoa Market Information Systems Based on Experiences in Nigeria, Cameroon and
Côte d’Ivoire, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, October 2006, 29 p.
95. TOLLENS, E., Markets and Institutions for Promoting Rice as a Tool for Food Security and Poverty Reduction
in Sub-Sahara Africa, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, October 2006, 27 p.
96. DAEMS, W., DEMONT, M., DILLEN, K., MATHIJS, E., SAUSSE, C. and TOLLENS, E., Economics of
Spatial Coexistence of Genetically Modified and Conventional Crops: Oilseed rape in Central France, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, March 2007, 53 p.
97. DILLEN, K., DRIES, L. and TOLLENS, E., The Impact of the EU sugar reform on sugar and sugar substitutes
industries, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, October 2006, 25 p.
98. DEMONT, M., CEROVSKA, M., DAEMS, W., DILLEN, K., FOGARASI, J., MATHIJS, E., MUŠKA, F., SOUKUP, J. and TOLLENS, E., Genetically modified crops in the New Member States: How much value and
for whom?, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, March 2007, 52 p.
99. DILLEN, K., DEMONT, M. and TOLLENS, E., Modelling heterogeneity to estimate the ex ante value of
biotechnology innovations, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, April 2007, 31 p.
100. DEMONT, M., DILLEN, K. and TOLLENS, E., Global welfare effects of genetically modified sugar beet under
changing EU sugar trade policies, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, April 2007.
101. STOEFS, E. and MATHIJS, E., Framing as strategy for effective communication about Fair Trade products in
Flanders, Centre for Agricultural and Food Economics, Katholieke Universiteit Leuven, November 2007.