AGRICULTURAL
ECONOMICS
Agricultural Economics 46 (2015) 1–15

The estimated ex ante economic impact of Bt cowpea in Niger, Benin
and Northern Nigeria
S. D. Gb`egb`el`egb`ea,∗ , J. Lowenberg-DeBoera , R. Adeotia , J. Luskb , O. Coulibalyc
a Department

of Agricultural Economics, 403 W. State St., West Lafayette, IN, 47907-2056, USA
of Agricultural Economics, 411, Agriculture Hall, Stillwater, OK, 74078, USA
c International Institute of Tropical Agriculture (IITA), BP 08-0932, Cotonou, Benin, Africa

b Department

Received 4 April 2013; received in revised form 1 October 2014; accepted 19 October 2014

Abstract
Genetically modified (GM) crops could increase economic growth and enhance living standards in Africa, but political issues have slowed
the use of biotechnology. This is the first study that assesses the potential impact of GM crops in Africa while considering the preferences of
producers and consumers towards GMOs as well as the income and price risks they face. The study uses a choice experiment to estimate the ex ante
economic impact of a novel technology, Bacillus thuringiensis (Bt) cowpea, on producers and consumers in Benin, Niger and northern Nigeria.
The experiment involves the simulation of a market transaction similar to those in open air markets in West Africa. During the market simulation,
respondents are informed about the advantages and disadvantages, including health risks, of Bt cowpea. The results from the study suggest that
cowpea growers and consumers in Benin and northern Nigeria prefer Bt to conventional cowpea for health safety reasons. The results estimate that
social welfare in Benin, Niger and northern Nigeria would increase by at least US$11.82 per capita annually with Bt cowpea, if seed sectors are
operating smoothly. With inefficiencies in seed sectors and the potential for cowpea acreage increase, the estimated social welfare increase in the
region would be about US$1.26 per capita annually.
JEL classifications: C9, C30, D81, L65
Keywords: Biotechnology; Africa; Risk; Choice experiment; Cheap talk; premium/discount; welfare

1. Introduction
Generating economic growth for the millions who live in
poverty in developing countries is one of the key world problems of our time. In many cases, economic growth is driven
by technological improvements. One of the technologies which
could generate economic growth and better living standards in
Africa is biotechnology, but political issues have slowed the use
of this option. Genetically modified (GM) crops in Africa have
become a political issue, in part because little is known about
the potential producer and consumer benefits.
Few studies have analyzed the economic impact of GM food
crops in Africa, including the attitudes of both producers and
consumers towards GM crops. Qaim (2001) used a partial equi∗ Corresponding

author. Tel.: +254 (20) 722 4630. E-mail address: g.sika@
cgiar.org (D.S. Gb`egb`el`egb`e).
Data Appendix Available Online
A data appendix to replicate main results is available in the online version of
this paper.

C

2015 International Association of Agricultural Economists

librium model, which involves semi-subsistence agriculture,
full acceptance of GM crops by farmers and consumers and
a closed-economy to estimate the impact of transgenic sweet
potatoes in Kenya. The study results suggest that the adoption
of transgenic sweet potatoes would lead to substantial benefits
for producers and consumers in Kenya. De Groote et al. (2011)
estimated the potential benefits of GM crops in Kenya; the
authors combined spatial modeling and economic techniques
to estimate the impact of Bacillus thuringiensis (Bt) maize in
Kenya. Their economic model involved a partial equilibrium
(PE) model characterized by an open economy for maize and
full acceptance of GM crops by consumers. Their results suggest that the economic benefits from the technology would depend on the ability of Bt maize to be effective against all major
stem borers. These previous studies have assumed full adoption
of GM crops by and no income risk for, producers and consumers. Kostandini et al. (2009) combine spatial and economic
modeling to estimate the benefits of GM drought-tolerant cereals in eight countries in sub-Saharan Africa and Asia. Their
study considers production risk and assumes differing adoption

DOI: 10.1111/agec.12182

2

S. D. Gb`egb`el`egb`e et al./Agricultural Economics 46 (2015) 1–15

rates of the improved technologies by farmers. However, they
also assume full acceptance of the GM crops by consumers.
Langyintuo and Lowenberg-DeBoer (2006) use a spatial and
temporal price equilibrium model to estimate the potential economic impact of Bt cowpea in Central and West Africa. Their
model considers different adoption rates of Bt cowpea by farmers in the region but it also assumes no risk and full acceptance
of Bt cowpea by consumers. Their study results imply that
the adoption of Bt cowpea would be more beneficial if it was
done at a regional level rather than in selected countries.
This study uses choice experiment to estimate the ex ante
economic impact of Bt cowpea in Benin, Niger and northern
Nigeria; it takes into consideration the preferences of consumers
and producers towards GM crops as well as the income risk they
face. The study estimates the impact of inefficient seed systems
on expected social welfare after the introduction of Bt cowpea.
Bt cowpea would be a transgenic crop. The proposed genetic
modification is insertion of DNA from the organism Bt into the
genome of cowpea. This genetic modification would allow the
cowpea plant to produce Bt toxin within its own cells and to
thereby resist attacks by Maruca vitrata without the application
of pesticides. Recent reports suggest that Bt cowpea could be
available by 2017 (AATF, 2011). The next section of the paper
presents the analytical framework used in estimating the ex ante
economic impact of Bt cowpea. Section 3 presents the survey
design and data while Section 4 presents the estimation procedures. Sections 5 and 6 discuss results. Section 7 summarizes
the paper.

≺∗

nity cost of time. M reflects the distribution of the optimal full
≺∗

incomes of the household, and Y (X∗ , T ) reflects the distribution of the optimal output quantities produced via the family
business. X∗ reflects optimal input quantities and T reflects the
maximum amount of fixed inputs available for production. L∗
reflects the optimal labor hours allocated by the household to the
family business, and E is the amount of time available for work
and leisure. Indirect utility is also assumed to be dependent upon
the minimum food consumption requirements of the household
(F ) and fixed attribute levels (AF ) for goods/services; it is assumed to be dependent on the consumption (zc ) and production
(zp ) characteristics of the household. εu is the error term reflecting this portion of utility specific to the household but unknown
to the researcher (Gbegbelegbe D., 2008).
The optimal solution to the problem of the household can be
used to estimate the economic impact of a new technology on
the welfare of the household given risks, and given the absence
of complete and actuarially fair insurance against risks. Without
Bt cowpea, the expected optimal utility of the household is:
ev noBt =
≺∗

E VnoBt P o , P h , w, M noBt , F ; AF,noBt , zc , zp , ε . (2)
With Bt cowpea the expected indirect utility function of this
household becomes:
≺∗

evBt = E VBt P o , P h , w, M Bt , F ; AF,Bt , zc , zp , ε
2. Conceptual framework
The household model is used to capture the problem of the
typical family in West Africa. This model implies that the income of the typical household in West Africa is modeled as
coming from two potential sources: family business that involves one or more family members and a nonfamily business.
The former generates monthly incomes that vary depending
on market conditions, while the nonfamily business tends to
provide relatively constant income. Therefore, the problem of
the typical household can be assumed to consist of maximizing food security and the expected satisfaction derived from
consumption given an environment characterized by various
constraints, risk (price and income), and the nonavailability of
actuarially fair insurance against these risks.
Assuming nonseparability, the optimal solution to the problem of the household can be written as:
≺∗

≺∗

E V P o , P h , w, M Y (X ∗ (P o , w), T ), L∗ , E , F ; AF , zc , zp , ε

,

(1)

where E(V (.)) is the expected utility of the household defined
over prices and random optimal income. P o is a vector of
prices for the consumption goods/services not produced via
the family business; P h reflects the prices of the consumption
goods/services produced by the household; w is the opportu-

(3)

and the distribution of compensating variations that equalize
expected utility with and without Bt cowpea is found with the
following equality:
≺∗

≺

ev noBt = E VBt P o , P h , w, M Bt − CV , F ; AF,Bt , zc , zp , ε

, (4)

≺

where CV is a vector reflecting the distribution of compensating variation brought by Bt cowpea. The latter distribution of
optimal welfare changes can be used to identify the compensating variation that Bt cowpea would bring in the worst-case
scenario or in the state of nature related to the bad outcome
occurring with a substantial probability:
≺

CV = Vector of random compensating variations due to Bt
cowpea
≺

CVcert = f (CV ) = minimum/certain compensating variations brought by Bt cowpea; corresponds to compensating variation related to worst-case scenario
In much of the current economics literature, the notion of certain welfare change is not considered as the appropriate ex ante
willingness-to-pay (WTP) for a nonmarketed product in the face
of uncertainty and unavailability of complete and actuarially fair
insurance; the notion of option price is. For the Bt cowpea case,
where a household can randomly have a low or high income

D. S. Gb`egb`el`egb`e et al./Agricultural Economics 46 (2015) 1–15

Source: Authors’ computations.
Fig. 1. Illustration of the certain compensating variation.

(Fig. 1), the option price would reflect the change in optimal
expenditures needed to equalize expected utility with and without Bt regardless of the state of nature occurring (Boardman,
et al., 1996; Desvousges, et al., 1987). In Fig. 1, the option
price would be reflected by the vector OP, where the compensating variation under the low income (CVpess) equals the
compensating variation under the high income (CVopt). The
certain compensating variation would correspond to CVcert
and would reflect the change in optimal expenditures needed
to equalize expected utility with and without Bt cowpea under
the low income scenario. The difference between the certain
and option prices relates to the assumptions on which they are
based. The notion of option price is based on the assumption
that the economic agent wants stability at all costs. The certain price encompasses the option price and is based on the
assumption that the economic agent is more concerned with a
bad outcome occurring with a substantial probability, i.e., the
worst-case scenario. When the agent can afford stability, the
option price equals the certain price.
3. Survey design and data
Because Bt cowpea was not yet on the market at the time
of the survey, a choice experiment combined with “cheap talk”
is used to estimate the ex ante economic impact of Bt cowpea in Benin, Niger and northern Nigeria which are part of
the Nigerian cowpea grainshed as described by Langyintuo
et al. (2003). The choice experiment involves the simulation
of a market transaction where respondents are incited to exhibit their WTP for a given product. Cheap talk consists of
explaining hypothetical bias to respondents so as to reduce its

3

occurrence during the market simulation, and hypothetical bias
occurs when the simulated market does not seem familiar and
believable to respondents (Freeman, 1993, Lusk, 2003, Lusk
and Schroeder, 2004).
The cheap talk script and the description of Bt cowpea that
was shared with the respondents are presented in Appendix
A. During the market simulation, the respondent is asked to
imagine that he/she is in front of a seller to buy cowpea in a
market. The seller explains the advantages and disadvantages
of conventional and Bt cowpea prior to offering these products
at given prices to the customer. When buyers are interested in
purchasing Bt or conventional cowpea, they are asked to choose
certain quantities, i.e., quantities of cowpea they are sure to buy
and consume regardless of what their income turns out to be. In
many cases, respondents adjusted their estimated cowpea purchase, depending on the prices of Bt and conventional cowpea.
In a few cases, customers provided a fixed amount of money
that they would expect to spend on cowpea regardless of the
level of income they will earn in the future or regardless of
the cowpea prices they might face in the future. In other cases,
they provided a fixed quantity of cowpea that they would buy,
regardless of what the cowpea market price or their income
turned out to be. In these cases, the fixed amount of money or
cowpea quantity provided by respondents reflected the option
price. In the other cases, where the respondents adjusted their
cowpea purchase to the prices of conventional and Bt cowpea,
the stated answers likely reflected the amount of cowpea purchase that the household plans to do, considering the minimum
income it is expecting to earn in the future.
The experimental design for the choice experiment was composed of two products for consumers (conventional and Bt cowpea) and three for producers (conventional cowpea; Bt cowpea;
and chemical insecticide).
For cowpea consumers, four prices for conventional cowpea
were selected from monthly market prices spread over the three
to five years preceding the survey and obtained from parastatals. Four coefficients were then applied to each of the four
conventional cowpea prices to estimate the price of Bt cowpea:
1/3, 2/3, 1, and 4/3. In total, there were 16 price combinations
for conventional and Bt cowpea, so that the full factorial design
used with consumers was composed of 16 choice sets, with
each choice set composed of a specific price for each of Bt and
conventional cowpea.
The same procedure was also used for producers for conventional and Bt cowpea. Normally, seeds are more expensive than
grains used for food consumption; but this was not considered
in this study. Producers were given the choice to buy chemical
insecticide along with conventional and Bt cowpea seeds, and
there were four price levels for chemical insecticide which were
also based on previous market prices. The four price levels for
each of conventional cowpea, Bt cowpea and chemical insecticide produced 64 (64 = 4*4*4) price combinations for the three
products. Therefore, an orthogonal fractional factorial design
was generated to reduce the number of price combinations. The
resulting factorial design was composed of eight choice sets

4

S. D. Gb`egb`el`egb`e et al./Agricultural Economics 46 (2015) 1–15

for producers, with each choice set reflecting a specific price
for each of Bt cowpea, conventional cowpea and chemical insecticide. The D-efficiency of the orthogonal factorial design is
100% (Louviere et al., 2000).
The price coefficients used for Bt cowpea led to low Bt
cowpea prices. Bt cowpea is being developed with public or
foundation funds. Hence, there is no need to recover the costs
of research and development, as is done with privately developed seed. In addition, Bt cowpea is much cheaper to produce
compared to conventional cowpea since it would require less
insecticide.
Figure B.1 in Appendix B illustrates the number and types
of decisions made during the market simulation for cowpea
consumers. These decisions constitute the profiles in the experimental design for consumers (Louviere, et al., 2000). In
Table B.1, respondents had to first decide whether they were
interested in buying cowpea at the prices offered. If they were
interested, they then had to decide which product to buy: Bt or
conventional cowpea. For the third profile, the respondents had
to decide how much of each product to buy during the period of
the year when the household consumes cowpea grown within
the household. In the last profile, respondents had to decide how
much of each product to buy during the period of the year when
the household usually buys cowpea.
In Table B.1, the experimental design for cowpea growers
is composed of five profiles. The first profile implies that respondents have to decide whether to buy cowpea seeds at the
prices offered. If they are interested, then they have to decide
which type of cowpea seed to buy. The next profile consists of
deciding the amount of cowpea seed and the level of chemical
insecticide needed. The last decision relates to the number of
hectares on which to plant the cowpea seed purchased.
Additional data on the demographic characteristics of the
households were collected. These data were used as independent variables in the econometric analysis to estimate the potential demand functions for Bt and conventional cowpea. For
all respondents, self-insurance relative to cowpea production
or consumption was assessed. All respondents who would purchase the same quantity of cowpea or would spend a fixed
amount of money to buy cowpea, regardless of what their income or what cowpea prices turned out to be were considered
self-insured. For such respondents, the certain price/quantity
equaled the option price/quantity.
Stratified random sampling was used for cowpea growers, rural and urban consumers of cowpea. Samples of cowpea growers
were selected in the major cowpea producing agro-ecological
zones in each country. Rural consumers were randomly selected
in each agro-ecological zone of a country whenever possible.
Urban consumers were selected based on a random selection
of houses located near major open-air city markets in a country
(Agazounon, 2003, Aitch´edji et al., 2004).
Budgetary and time constraints implied that a total of 534
households could be sampled for this study. In Benin, 56
farm households growing cowpea were surveyed and they were
spread across the major agro-ecological zones involved in cow-

pea production. The sample of Beninese urban and rural cowpea
consumers was 69 and 83, respectively. In Niger, the sample of
urban and rural cowpea consumers included 39 and 40 households, respectively; the sample of cowpea growers involved 40
households with 20 in each of the two agro-ecological zones
(arid and semi-arid zones) characterized by high cowpea production. The sample of urban and rural cowpea consumers
involved 60 households each in Nigeria; the sample of cowpea growers involved 88 households spread across the three
Nigerian agro-ecological zones characterized by high cowpea
production.
Each cowpea grower provided purchase decisions for eight
price combinations for Bt cowpea, conventional cowpea and
chemical insecticide. Urban and rural consumers were presented with 16 price combinations for Bt and conventional cowpea. After removing the responses with missing data, the total
number of observations by cowpea growers was 608 in Nigeria,
295 in Niger, and 448 in Benin. The number of observations
was 816, 592, and 1280 for the rural consumers in Nigeria,
Niger, and Benin, respectively; whereas the total number of observations was 799, 556, and 1087 for the urban consumers in
Nigeria, Niger, and Benin, respectively. The questionnaire on
cowpea production was only administered to producers while
the one for consumers was only administered to the urban and
rural consumers.
4. Estimation procedures
The certain quantities obtained via the survey were then regressed on the exogenous socio-economic variables to estimate
certain demand functions for both Bt and conventional cowpea.
Preliminary data analysis suggested that exponential functions
would more appropriately capture the relationship between the
exogenous and endogenous variables; moreover, the exponential functional form is consistent with economic theory since
exponential demand functions respect the law of diminishing
marginal utility. Stated WTP for Bt and conventional cowpea
was recorded from each respondent. Hence, exponential demand functions for Bt and conventional cowpea are estimated
using the Seemingly Unrelated Regression (SUR) model:
QBt = eβXBt +εBt

(i)

Qc = eβXc +εc .

(ii)

Where QBt QBt is an n-dimensional vector reflecting certain
quantities of Bt cowpea; ‘n’ is the sample size. Qc is also an ndimensional vector reflecting certain quantities of conventional
cowpea. XBt is an ‘n’ by ‘kBt ’ matrix of exogenous socioeconomic variables in equation (i). Xc is an ‘n’ by ‘kc ’ matrix
of exogenous socio-economic variables in equation (ii). β is
a vector (kBt -dimensional) of unknown parameters in equation
(i). αis a vector (kc -dimensional) of unknown parameters in
equation (ii). εBt is an n-dimensional vector reflecting error

D. S. Gb`egb`el`egb`e et al./Agricultural Economics 46 (2015) 1–15

terms in equation (i) while εc is also an n-dimensional vector
reflecting error terms in Eq. (ii).
The descriptive statistics of the variables used in the analysis
are listed in Tables C.1, C.2, and C.3 (see Appendix C) for
cowpea growers, urban and rural consumers, respectively. The
exogenous variables for cowpea growers are in seven groups:
prices; characteristics of the household (size and insurance) and
the household head (age, gender, and education); characteristics of cowpea production (cowpea color, type, and source of
cowpea seed, type of insecticide used in cowpea production); institutional characteristics of the household (credit access, access
to development agencies, access to market inputs, and access
to extension agencies); income and wealth of the household;
knowledge of GMOs; and the households’ perceptions on Bt
and conventional cowpea (yield and problems with insecticide).
The exogenous variables for urban and rural consumers are
also grouped in seven: prices; the characteristics of the household (size and insurance) and the household head (age; gender;
education); the experience of the household with cowpea production (years of experience growing cowpea and uses of homegrown cowpea); the experience of the household with bought
cowpea (cowpea color, uses of bought cowpea); income and
wealth of the household; knowledge of GMOs; and the perception of households on Bt and conventional cowpea (health
safety and ease of cooking).

5. Estimated certain hicksian demands for conventional
and Bt cowpea
Results from the econometric estimation of potential demand
functions for Bt and conventional cowpea for cowpea growers and consumers across the three countries are presented in
Tables 1, 2 and 3. The R2 values for the econometric results in
the tables suggest that the exponential functional form seems
to reflect well the demand function for Bt cowpea by cowpea
growers and consumers.
The results in Table 1 imply that each of chemical insecticide and conventional cowpea is a substitute to Bt cowpea in the
eyes of cowpea growers in northern Nigeria. The coefficients
on the prices of conventional cowpea and chemical insecticide are positive and statistically significant for the estimated
Bt cowpea demand function for the average cowpea grower in
Nigeria (Table 1). For cowpea growers in Niger and Benin, the
prices of conventional cowpea and chemical insecticide do not
affect Bt cowpea demand. One variable which seems to significantly influence the demand for Bt and conventional cowpea
is the knowledge of Genetically Modified Organisms (GMOs).
In Table 1, the estimated coefficient on the variable “Knows
GMOs” is positive and significant for cowpea growers in Niger
and Benin; this implies that cowpea growers knowledgeable of
GMOs seem to have a higher WTP for Bt cowpea compared
to cowpea growers who have never heard of GMOs. Such attitudes imply that farmers in Benin and Niger who hear about
GMOs but have never used them seem to exhibit a favorable

5

attitude towards these products. Previous studies demonstrate
that farmers in the US (who are almost all aware of GMOs)
seem favorable towards growing GMOs (USDA, 2014).
The estimated coefficients in Table 1 were multiplied with
their mean values in Table C.1 to estimate certain Hicksian demands for Bt and conventional cowpea for the average cowpea grower in each region under study. The estimated Bt
cowpea demand function for the average cowpea grower is
(−1.17∗ PBt +3.52)
, where ‘1.17’ is directly taken from
QBt
Nigeria = e
Table 1 and ‘3.52’ is the aggregated product between the significant estimated coefficients in Table 1 and their mean values in
Table C.1. A similar approach was used to estimate the other demand functions whose estimated coefficients are in Table 1. The
parameters for all estimated demand functions are in Appendix
D. The estimated inverse demand functions were then used to
compute premiums/discounts for Bt compared to conventional
cowpea.
The sowing rate for cowpea ranges between 12 and 25 kg/ha
in Benin, Niger and northern Nigeria (Dugje et al., 2009). Farmers in northern Benin allocated on average 0.46 ha to cowpea
in 2005, based on the survey data (data not shown). With an
average sowing rate of 15 kg/ha, the average amount of cowpea
seed planted by the farmers was 6.84 kg. Moreover, much of the
planted cowpea seeds consisted of local seeds kept by farmers
from one season to the next. In northern Benin, only 30% of
the surveyed respondents used cowpea seeds which they had
bought in the market (data not shown). Hence, for a volume
of 6.84 kg, the estimated WTP for Bt or conventional cowpea
based on the parameters in Appendix D are negative: the average farmer in northern Benin would not be willing to buy
6.84 kg of cowpea. However, purchases of Bt cowpea become
positive around 3 kg. For a purchase of 3 kg of cowpea seed,
the average cowpea grower in Benin would be willing to spend
US$30.22 more for Bt compared to conventional cowpea; this
grower would spend US$10.95 on 3 kg of Bt cowpea; however,
their WTP for 3 kg of conventional cowpea would be negative
at about US$19.28. This result suggests that if both Bt and conventional cowpeas were available, the average cowpea grower
in Benin would not buy conventional cowpea; they would need
to receive some compensation of about US$19.28 to take 3 kg
of conventional cowpea. A similar conclusion applies for the
average cowpea grower in Nigeria, although their discount of
conventional relative to Bt cowpea is smaller. For a purchase
of 3 kg of Bt cowpea, the average cowpea grower in Nigeria would spend US$6.21; however, they would require some
compensation of US$1.23 to take 3 kg of conventional cowpea.
If conventional cowpea was priced at US$391/ton (see Appendix D) which translates into US$0.391/kg, the premium of
Bt over conventional cowpea, for a purchase of 3 kg, would
be 833% and 430% for the average cowpea grower in Benin
and northern Nigeria, respectively. This translates into farmers
paying 9.3 and 5.3 times more for Bt compared to conventional
cowpea, in Benin and Nigeria, respectively. Even for high cowpea prices, the average grower in Nigeria and Benin would still
prefer Bt to conventional cowpea.

6

S. D. Gb`egb`el`egb`e et al./Agricultural Economics 46 (2015) 1–15

Table 1
Estimated demand functions for cowpea seed–cowpea growers
Nigeria

Niger

Bt cowpea
Explanatory variable
Price of Bt cowpea
Price of Cnl. cowpea
Price of insecticide
PriceBt – Utility
Hhld size
Insured
Age
Gender
Western educ.
Koranic educ.
Primary educ.
Secondary educ.
Vocational training
Literate
Experience cowpea prod.
Prod – white
Prod – speckled white
Prod – red
Prod – Local seed market
Prod – Local seed home
Prod – Impr. seed store
Prod – improv. seed home
Prod – Icide nonrecom.
Prod – Icide bot
Prod – Icide unknown
Prod – Icide bf plant.
Prod – Icide infestation
Prod – Credit access
Prod – Access to dev. ag.
Prod – Input access
Prod – Access ext.
Expected nonfarm income
Prob. (lowest income)
Ag. wealth
Land size
Knows GMOs
BBC: know. source
Bt – Health prbl. icide
Cnl – Health prbl. icide
Bt – Yield w/o icide
Cnl – Yield w/o icide
AEZ – semi-arid
AEZ – N.G. sav.
AEZ – arid
AEZ – S.G. sav.
R2
No of obs.

Est
−1.17**
0.41**
0.08**
N/A
0.02**
N/A
0.02**
0.54**
N/A
0.00
0.03
0.52**
N/A
N/A
−0.01**
N/A
N/A
N/A
N/A
0.32**
−0.33
N/A
N/A
0.44**
0.11
0.05
0.24**
−0.07
0.39**
−0.22**
0.00**
0.62**
0.00
0.00**
−0.22
N/A
−0.25**
−0.05
0.07**
−0.24**
−0.99**
−0.31**
N/A
N/A
0.98
608

Cnl cowpea
SE

Est

0.08
0.11
0.01

3.67**

0.01
0.00
0.08
0.11
0.07
0.14

0.01

0.10
0.20

0.11
0.14
0.13
0.04
0.07
0.04
0.04
0.00
0.12
0.00
0.00
0.32
0.09
0.04
0.02
0.04
0.18
0.15

−3.64**
−0.33**
N/A
−0.04**
N/A
0.05**
−0.86**
N/A
6.60**
−0.26
−0.09
N/A
N/A
0.00
N/A
N/A
N/A
N/A
−0.91**
−0.83
N/A
N/A
5.48**
−6.88**
−6.27**
0.02
0.26
−1.32**
0.04
0.01**
−2.93**
0.00
0.00
N/A
N/A
0.12
0.31*
0.17
0.09
−2.59**
4.26**
N/A
N/A
0.91
608

Benin

Bt cowpea
SE

Est

0.30
0.46
0.04

−0.62**

0.02
0.02
0.31
1.32
0.18
0.47

0.02

0.37
0.65

2.03
1.92
1.91
0.15
0.31
0.30
0.14
0.00
1.00
0.00
0.00

0.25
0.16
0.13
0.18
0.69
1.11

Cnl cowpea
SE

−0.12
0.00
N/A
N/A
−0.28**
0.03**
N/A
N/A
0.17**
−0.63**
−0.33**
N/A
−0.94**
N/A
0.00
0.33*
−0.31**
0.11
N/A
N/A
N/A
N/A
0.88**
N/A
N/A
−0.17**
N/A
0.15**
−0.12**
0.01**
N/A
0.00**
0.00**
1.08**
−1.07**
0.30**
−0.11**
0.24**
N/A
N/A
N/A
−0.46**
N/A
0.35
26

0.08
0.15
0.00

0.07
0.00

0.08
0.12
0.12
0.11
0.00
0.18
0.08
0.08

Est
N/A
9.50**
N/A
N/A
0.09
6.53**
−0.31**
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A

0.12

0.03
0.03
0.02
0.00
0.00
0.00
0.20
0.31
0.10
0.03
0.03

0.11

N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
0.00**
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
0.20
5

SE
1.71
0.00
0.08
1.38
0.07

0.00

Bt cowpea

Cnl cowpea

Est

SE

Est

SE

−0.28**

0.12
0.20
0.01
0.35
0.01
0.09

−0.08
−0.84*
−0.07**
N/A
0.03
4.25**
N/A
N/A
0.89**
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
1.06**

0.28
0.43
0.03

0.05
−0.01
3.91**
0.03**
0.38**
N/A
N/A
−0.27
N/A
N/A
N/A
0.20*
N/A
N/A
N/A
N/A
N/A
−0.42**
N/A
−1.08**
−0.62**
−0.21*
N/A
0.35**
0.41**
N/A
N/A
N/A
N/A
N/A
N/A
0.00**
0.00**
0.61**
N/A
N/A
N/A
N/A
N/A
N/A
N/A
0.82**
N/A
18.00
448

0.10

0.14

0.09

0.29
0.18
0.17
0.15
0.14

0.00
0.00
0.16

0.11

N/A
N/A
−4.68**
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
0.00**
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
9.00
448

0.03
0.46

0.28

0.34

0.55

0.00

N/A: nonapplicable; * and ** represent statistical significance at the 10% and 5% level, respectively
Source: Authors’ computations.

In Niger, the average cowpea grower would discount conventional cowpea for smaller seed purchases; Bt cowpea seeds
would be discounted for larger purchases. More specifically,
for a purchase of 600 g, the average cowpea grower would be
willing to spend 3.1 times more for Bt cowpea seeds, with a total spending of US$2.80 on the 600 grams of Bt cowpea seeds.

However, for a purchase of 4.5 kg, the cowpea grower would
prefer conventional cowpea and would be willing to spend 16%
more for the conventional cowpea seeds.
Figure 2 illustrates estimated demand functions for Bt and
conventional cowpea, for the average cowpea grower in the
semi-arid and arid zones of Niger. In Fig. 2, the demand for

D. S. Gb`egb`el`egb`e et al./Agricultural Economics 46 (2015) 1–15

7

Table 2
Estimated demand functions for cowpea grain–rural consumers
Nigeria

Niger

Bt cowpea
Explanatory variable
Price of Bt cowpea
Price of Cnl. cowpea
PriceBt
Insured
Hhld size
Age
Gender
Koranic educ.
Primary educ.
Secondary educ.
Literate
Prod – grow cowpea
Experience cowpea prod.
Cons – homegr. – Home
Cons – homegr. – food resale
Cons – buy cowpea
Cons – White
Cons – Red
Cons – Brown
Cons – bought – Home
Cons – bought – resale raw
Cons – bought – resale food
Exp. Income
Lowest income
Prob. (lowest income)
Ag. Wealth
Land size
Knows GMOs
Prints: know. source
Friends + Neighb.: know. source
BBC: know. Source
Extension: know. source
Safety of Bt – Cons.
Safety of Cnl – Cons.
Bt – Easy to cook
Cnl – Easy to cook
AEZ – semi-arid
AEZ – South. G. Savanna
R2
N (# obs.)

Est
−1.58**

−0.07
N/A
N/A
−0.14**
−0.01*
−1.00**
−2.74**
−1.09**
−0.39**
N/A
−0.81**
0.04**
−0.27
0.24
−1.18**
0.46**
N/A
N/A
−0.35**
1.34**
N/A
0.01**
N/A
−1.28**
0.00
0.00
−0.67**
−0.27
−1.28**
−0.47
−2.11**
0.36**
−0.62**
0.89**
1.10**
−0.18
N/A
0.93
816

Cnl cowpea
SE

Est

0.13
0.16

3.25**

0.01
0.01
0.16
0.42
0.10
0.13
0.33
0.01
0.24
0.27
0.31
0.15

0.14
0.18
0.00
0.47
0.00
0.00
0.29
0.34
0.38
0.31
0.33
0.07
0.09
0.09
0.08
0.17

−3.79**
N/A
N/A
−0.08
0.26**
N/A
5.87**
0.02
−0.43
N/A
N/A
−0.28**
N/A
N/A
3.13**
−4.51**
N/A
N/A
−0.38
N/A
N/A
−0.07**
N/A
N/A
0.00
−0.17**
8.29**
N/A
N/A
N/A
N/A
−2.94**
5.76**
−2.38**
−1.28**
N/A
N/A
0.93
816

Benin

Bt cowpea
SE

Est

0.23
0.34

−2.05**

0.06
0.03
2.03
0.35
0.41

0.04

1.07
0.82

0.26

0.01

0.00
0.02
0.87

0.28
0.68
0.35
0.24

0.09
2.77**
−0.20
−0.14**
−0.03**
0.67**
0.69**
0.73**
−0.96**
1.87**
N/A
0.08**
−1.10**
N/A
N/A
N/A
N/A
N/A
−0.17
N/A
N/A
0.00*
N/A
−0.56**
0.00**
−0.01**
1.11**
N/A
N/A
N/A
N/A
0.79**
−0.27**
−0.09**
0.23**
N/A
N/A
0.96
592

Cnl cowpea
SE

Est

0.15
0.16
0.94
0.18
0.02
0.01
0.18
0.11
0.11
0.15
0.30

1.69**

0.01
0.35

0.13

0.00
0.15
0.00
0.00
0.18

0.10
0.04
0.04
0.05

−2.69**
N/A
N/A
−0.09
−0.08**
−0.62*
−2.41**
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
0.03**
N/A
2.31**
0.00
−0.01
1.79**
N/A
N/A
N/A
N/A
−3.83**
N/A
0.26
3.33**
N/A
N/A
0.97
592

Bt cowpea
SE

Est

0.30
0.39

−1.63**

0.07
0.04
0.37
0.91

0.01
0.89
0.00
0.01
0.88

1.50
0.37
1.53

1.52
0.00**
−6.26**
0.01**
N/A
N/A
N/A
0.00
0.26**
N/A
N/A
N/A
N/A
N/A
N/A
0.61*
1.16
1.03**
−1.97
0.00**
−0.77**
0.01**
0.00
0.57
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
−0.40**
1.53**
0.92
1280

Cnl cowpea
SE

Est

SE

0.15
0.19
0.00
0.96
0.01

0.33**

0.18
0.29

0.00
0.15

0.10
0.10
0.10
0.08
0.00
0.10
0.00
0.00
0.15

0.10
0.12

−3.44*
N/A
3.51
0.33
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
1.04**
2.51*
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
3.60
0.93
1280

0.17
0.03

1.00
0.36

0.16

N/A: nonapplicable; * and ** represent statistical significance at the 10% and 5% level, respectively.
Source: authors’ computations.

conventional cowpea is almost flat and perfectly elastic; this
occurs because other negative factors outweigh the positive
price coefficient (Table 1).This suggests that, if both conventional and Bt cowpea seeds were available, the average grower
would expect to buy the conventional seed at the same price
regardless of the quantity purchased. The estimated demands
in Fig. 2 also indicate that the average cowpea grower in the
semi-arid zone has a larger premium for Bt cowpea compared
to the grower in the arid zone. This preference in the arid zone
may be related to the nature of insect infestations. Malick Ba
et al. (2009) explain how some cowpea pest populations do not
remain permanently in drier zones and rather migrate each rainy

season to more humid regions. Hence, it is likely that cowpea
pest infestation is more sporadic in the arid zone of Niger than
in the other agro-ecological zones considered in this study. With
sporadic infestations, intensive use of chemical insecticides is
less important for cowpea production. Hence, for the farmers
in this zone, Bt cowpea would less likely lead to higher yields
and hence higher income.
The results in Table 2 suggest that Bt cowpea is a substitute
for conventional cowpea in the eyes of rural consumers in Benin,
Niger, and northern Nigeria. In Nigeria and Niger, the more
educated rural consumers are, the lower their demand for Bt
cowpea. In Nigeria, the coefficients for the variables ‘Primary

Price of Bt cowpea
Price of Cnl. cowpea
PriceBt – Utility
PriceCnl – Utility
Insured
Hhld size
No of able workers
Age
Gender
Koranic educ.
Primary educ.
Secondary educ.
University
Voc. training
Voc. train. – # degrees
Voc. train. – tailor
Voc. train. – elec
Voc. train. – trainer
Prod – grow cowpea
Cons – White
Cons – Spkled white
Cons – Brown
Cons – Red
Cons – dark Red
Cons – home
Cons – resale raw
Cons – resale food
Exp. income
Lowest income
Prob. (lowest income)
Ag. wealth
Land size
Knows GMOs
TV: know. source
CNN: know. source
Prints: know. source
0.57
0.01
0.01
0.22
0.46
0.35
0.30

0.00
0.21
0.57
0.56
0.68
0.00

0.70
0.22

0.26
0.03
0.05
0.01
0.32
0.31
0.54
0.31
0.25

0.19
0.21

−1.94**

0.42**
N/A
N/A
0.93**
−0.35**
−0.02
−0.05**
1.63**
−4.05**
−7.20**
−4.82**
−2.56**
N/A
N/A
N/A
N/A
N/A
7.40**
0.25
N/A
N/A
0.00**
0.31
8.67**
11.27**
12.74**
0.00**
N/A
−3.47**
−0.09**
0.07**
−0.09
7.80**
−0.44
1.72**

SE

Est

Explanatory variable
0.29
−1.72**
−39.15**
N/A
−14.45**
−2.26**
1.40**
0.45**
−15.36**
18.07**
4.40**
9.99**
4.76**
N/A
N/A
N/A
N/A
N/A
N/A
−13.17**
N/A
N/A
8.37**
N/A
5.33**
15.20**
−4.94
0.01**
N/A
N/A
N/A
N/A
−6.70**
N/A
N/A
N/A

Est

SE

0.47

2.32
2.29
4.19
0.00

0.79

1.34

3.95
0.18
0.26
0.04
3.99
3.95
0.47
0.96
0.60

0.27
0.37
4.85
0.30
−157.45**
N/A
−1.66**
1.30**
N/A
0.37**
−34.84**
−1.50**
−0.69**
0.00
N/A
N/A
N/A
4.43**
67.34**
63.94**
N/A
−29.50**
−30.05**
−16.56**
N/A
N/A
0.57*
11.38**
0.00
−0.04**
0.00
12.72**
N/A
N/A
9.12**
N/A
N/A
N/A

−1.79**

Est

Niger
Bt cowpea

Bt cowpea

Cnl cowpea

Nigeria

Table 3
Estimated demand functions for cowpea grain–urban consumers

0.75

0.32
1.48
0.00
0.00
0.00
1.66

2.93
2.72
2.36

0.76
6.71
5.43

0.04
3.41
0.28
0.32
0.00

0.60
0.12

0.18
0.26
12.92

SE
−2.37**
N/A
−0.32
−0.12
0.14**
N/A
−0.08**
−2.55**
1.72**
−1.14**
0.00
N/A
N/A
N/A
0.00
N/A
0.00
N/A
−3.74**
N/A
N/A
N/A
N/A
−1.22**
−0.68
0.00
−0.02**
0.00
0.10
N/A
N/A
−1.68**
N/A
N/A
N/A

2.01**

Est

Cnl cowpea

0.85

0.29
0.82
0.00
0.00
0.00
0.34

0.75

0.00

0.00

0.01
1.12
0.44
0.26
0.00

2.88
0.14
0.02

0.22
0.35

SE

Benin

0.08
−3.63**
N/A
−0.59**
−0.12**
N/A
N/A
N/A
−0.40**
0.24**
−0.08
N/A
−1.56**
1.03**
N/A
N/A
N/A
N/A
4.93**
N/A
0.00
−0.15*
N/A
−1.15**
3.00**
0.00
0.00**
N/A
0.62**
N/A
−0.36**
N/A
N/A
N/A
N/A

−2.31**

Est

Bt cowpea

0.13

0.23

0.07
0.11
0.00
0.00

0.00
0.09

0.27

0.65
0.41

0.15
0.07
0.11

0.20
0.01

0.15
0.18
0.38

SE

−1.63**
37.83**
N/A
8.63**
−1.35**
N/A
N/A
N/A
N/A
4.72**
−19.54**
N/A
−13.64**
32.15**
N/A
N/A
N/A
N/A
−20.53**
N/A
N/A
−7.29**
N/A
−12.42**
15.28**
N/A
−0.01**
N/A
4.66**
N/A
17.23**
N/A
N/A
N/A
N/A

0.78**

Est

1.07

0.87

0.00

0.79
0.79

0.48

1.38

1.12
2.30

0.26
1.40

0.60
0.12

0.19
0.31
2.57

SE

(Continued)

Cnl cowpea

8
S. D. Gb`egb`el`egb`e et al./Agricultural Economics 46 (2015) 1–15

0.38
0.35
0.63

−1.03**

Radio: know. source
Friends+Neighb.: know. source
BBC: know. Source
Safety of Bt – Cons.
Safety of Cnl – Cons.
Bt – Easy to cook
Cnl – Easy to cook
R2
N (# obs.)

N/A
N/A
N/A
N/A
N/A
N/A
N/A
0.84
799

Est

SE
N/A
N/A
N/A
28.53**
2.00**
−10.18**
−4.65**
0.95
556

Est

Source: Authors’ computations.

N/A: nonapplicable; * and ** represent statistical significance at the 10% and 5% level, respectively

5.24**
1.32**
N/A
N/A
N/A
N/A
0.89
799

SE

Est

Bt cowpea

Bt cowpea

Cnl cowpea

Niger

Nigeria

Explanatory variable

Table 3
Continued

2.47
0.38
1.14
0.44

SE
N/A
N/A
N/A
0.97**
0.97**
−1.15**
0.93**
0.95
556

Est

Cnl cowpea

0.16
0.15
0.36
0.28

SE

N/A
N/A
N/A
N/A
N/A
N/A
N/A
0.94
1087

Est

Bt cowpea

Benin

SE

N/A
N/A
N/A
N/A
N/A
N/A
N/A
0.94
1087

Est

Cnl cowpea
SE

D. S. Gb`egb`el`egb`e et al./Agricultural Economics 46 (2015) 1–15
9

10

S. D. Gb`egb`el`egb`e et al./Agricultural Economics 46 (2015) 1–15

Source: Authors’ computations.
Fig. 2. Illustration of estimated demands for Bt and conventional cowpea—
Average cowpea grower in Niger.

educ.’ and ‘Secondary educ.’ are negative and significant for
Bt cowpea demand. In Niger, consumers who have attended
primary school tend to demand more Bt cowpea compared to the
ones with no westernized education. However, consumers who
have completed primary school and have attended secondary
school tend to demand less Bt cowpea compared to the ones
with no westernized education. On the other hand, in Benin,
the more educated rural consumers tend to demand more Bt
cowpea compared to the ones with no westernized education.
Knowledge of GMOs influences Bt cowpea demand. In Nigeria, a rural consumer who is aware of GMOs tends to demand
less Bt cowpea compared to the one who has not heard of
GMOs (Table 2). The contrary is observed for rural consumers
in Niger: prior awareness of GMOs increases their demand for
Bt cowpea.
The results in Table 2 imply that female-headed rural households in Nigeria would more likely adopt Bt cowpea compared
to male-headed households: the coefficient on ‘Gender’ is negative and significant for Bt cowpea demand in Table 2. In Niger,
male-headed rural households would more readily adopt Bt
cowpea compared to female-headed households (Table 2).
The results in Table 2 were used to estimate certain Hicksian
demand functions whose parameters are listed in Appendix D.
The demand functions were then used to estimate premiums and
discounts for Bt compared to conventional cowpea. The average
rural consumer in Benin, Niger and Nigeria tends to prefer Bt to
conventional cowpea. For example, on a purchase of 1 kg, the
premiums expressed for Bt cowpea were US$3.22, US$2.53,
and US$2.19 in Nigeria, Niger and Benin, respectively. The
average rural household would buy 1 kg of Bt cowpea at about
US$1.46, US$1.20, and US$1.43 in Nigeria, Niger, and Benin,
respectively. However, they would require some compensation
of US$1.76, US$1.32, and US$0.75 to take conventional cowpea home in Nigeria, Niger, and Benin, respectively. If conven-

tional cowpea is priced at US$391/ton (see Appendix D), then
for a purchase of 1 kg, the premium of Bt over conventional
cowpea would be 273%, 207%, and 267% in Nigeria, Niger,
and Benin, respectively.
Based on the results in Table 3, the average urban consumer
in Nigeria, Niger, and Benin would consider conventional and
Bt cowpea as substitute products. As in the case of rural consumers, education seems to influence urban cowpea demand.
Education negatively influences the urban demand for Bt cowpea in Nigeria and Niger; it seems to positively influence the
urban demand for Bt cowpea in Benin. In Nigeria, educated
urban consumers tend to demand less Bt cowpea compared to
consumers with no westernized education. The coefficients on
‘Primary educ.’, ‘Secondary educ.’, and ‘University’ are negative and significant for Bt cowpea demand (Table 3). In Niger,
respondents who have attended primary school tend to demand
less Bt cowpea compared to the ones with no westernized education. However, in Benin, respondents who have attended
primary school demand more Bt cowpea compared to the ones
with no westernized education.
Prior knowledge of GMOs also influences urban demand
for Bt cowpea. In Niger, respondents with prior knowledge of
GMOs would demand more Bt cowpea compared to the ones
without any prior knowledge (Table 3). In Nigeria, respondents
demand more Bt cowpea when the source of knowledge on
GMOs consists of print media; the BBC radio or friends and
neighbors. Respondents who have heard about GMOs from
other radio sources apart from ‘radio BBC’ had a tendency to
demand less Bt cowpea compared to the ones who had not heard
of GMOs (Table 3).
Interestingly, expected income (“Exp. income”) is negatively
related to the quantity of Bt and conventional cowpea in Niger.
In Benin, expected income is negatively related to the quantity
of conventional cowpea. This suggests that Bt and conventional cowpea would be considered inferior goods for urban
consumers in Niger. In Benin, conventional cowpea would be
considered an inferior good.
The results in Table 3 were used to estimate certain demand
functions (see Appendix D) which were used to estimate premiums and discounts for Bt cowpea. The results suggest that
the average urban consumer in Benin and Nigeria seem to prefer Bt to conventional cowpea, but the average urban consumer
in Niger would discount Bt cowpea relative to conventional
cowpea.
For example, for a purchase of 1 kg, the premiums of Bt over
conventional cowpea are US$5.01 and US$9.63 for the average urban household in Nigeria and Benin, respectively. The
average urban consumer in northern Nigeria is willing to pay
US$0.51 for 1 kg of Bt cowpea; however, their WTP turns out
to be negative at US$ 4.49 for 1 kg of conventional cowpea. A
similar result applies to the average urban consumer in Benin
who is willing to pay about US$0.84 for 1 kg of Bt cowpea and
has a negative WTP for conventional cowpea. For a purchase
of 500 g, the average urban consumer in Nigeria and Benin
exhibits a negative WTP for conventional cowpea; in this case,

D. S. Gb`egb`el`egb`e et al./Agricultural Economics 46 (2015) 1–15

the premium for Bt over conventional cowpea is US$2.48 and
US$4.75, respectively. These results imply that if both Bt and
conventional cowpea were offered to urban consumers in Benin
and northern Nigeria, they would no longer buy conventional
cowpea; they would actually need to be compensated to take
conventional cowpea home. However, the average urban consumer in Niger would exhibit a negative WTP for Bt cowpea
and would spend about US$0.08 for 500 g of conventional cowpea. Their premium for conventional over Bt cowpea is about
US$2.91 for a purchase of 500 g of cowpea.
With a price of US$391/ton for conventional cowpea (see
Table 4) in the analysis above, the premium of Bt over conventional cowpea for a purchase of 1 kg becomes 31% and
124% for the average urban consumers in Nigeria and Benin,
respectively.
The preferences of the growers and consumers are supported
by their perceptions of Bt and conventional cowpea (Appendix
C.1, C.2, and C.3). Respondents’ perceptions were collected in
Nigeria and Niger, only; hence, comparable data is not available for Benin. The average cowpea grower in Nigeria thinks
that the health risks related to misusing chemical insecticide
would be lower with Bt compared to conventional cowpea
(Table C.1). In addition, the average grower thinks that cowpea
yield without the use of chemical insecticide would be higher
with Bt compared to conventional cowpea (Table C.1). Such
perceptions can explain the preference of cowpea growers in
Nigeria for Bt over conventional cowpea. The perceptions of
the average cowpea grower in Nigeria are also shared by the
average grower in Niger.
Multiple reasons can explain the preferences of rural consumers in Nigeria and Niger for Bt over conventional cowpea.
In both countries, the average consumer believes that Bt cowpea
is easier to cook compared to conventional cowpea (Table C.2).
Higher producer income, from the sale of Bt cowpea, might
also explain their preferences.
The average urban consumer in northern Nigeria believes
conventional cowpea to be slightly safer for human consumption but more difficult to cook compared to Bt cowpea
(Table C.3). Conventional cowpea were formerly cooked by
boiling once, but now in some parts of Nigeria, families often boil conventional cowpea three times and discard cooking
water before using the boiled cowpea it as a cooking ingredient. They do this because they believe that this cooking technique removes pesticide residues and makes the cowpea safer
to eat.
Based on Table C.3, the average urban consumer in Niger
believes that Bt cowpea is slightly healthier and slightly easier
to cook compared to conventional cowpea. These reasons do
not explain why this average urban consumer unit seems to
prefer conventional to Bt cowpea, as shown via the estimated
demand functions. However, information collected during the
survey suggests that urban consumers in Niger tend to believe
that their current cooking procedure of conventional cowpea,
which involves the use of natron (a naturally-occurring type of
sodium carbonate), eliminates any type of residues, including

11

pesticide residues. However, studies have actually shown that
‘natron’ reduces cooking time for cowpea but too much of it
can create health problems (Balla and Barag´e, 2006).
Beninese respondents were asked to state the reasons behind their preferences, whenever possible. The recurrent reason
among all respondents in Benin relates to health. Most cowpea
in Benin is currently produced with chemical insecticides not
labeled for cowpea; insecticides which farmers tend to misuse
to the detriment of their health and the health of their family
members. Moreover, cowpea sellers in the open-air markets of
Benin tend to conserve cowpea with insecticides not approved
for use on food, and hence, they put consumers’ health at risk.
There are various reports of on-farm casualties caused by the
misuse of chemical insecticide in relation to cowpea between
1999 and 2008 (PAN UK, 2001, PAN and IPEN, 2009). In Beninese cities, in 2005 and 2006, when the surveys were conducted,
there were many rumors of casualties caused by contaminated
cowpea. Such reports are consistent with survey reports that imply that cowpea heavily contaminated with nonrecommended
insecticides is sold in some major markets in Benin, including
Ouando and Pob`e (Adigoun, 2002).

6. Estimated ex ante economic impact of Bt cowpea
Figure 3(a) illustrates aggregate demand and supply for conventional cowpea in Benin, Niger, and northern Nigeria, prior
to the introduction of Bt cowpea. The numbers used to derive
Fig. 3 are estimates from the 1990s from Langyintuo (2003)
who used a methodology where observed market data is combined with main market pattern determinants such as seasonal
price changes and storage losses to estimate cowpea supply
and demand functions for countries in West Africa. Given the
available data and methodological construct, linear demand and
cowpea functions turned out to best reflect cowpea market equilibrium characteristics in the region (Langyintuo, 2003). The
slope and intercepts of cowpea supply and demand functions
derived from Langyintuo (2003) are presented in Appendix D.
Since our study considers northern Nigeria and not all of Nigeria for production and consumption, the intercept and slope of
the estimated cowpea demand function for Nigeria is multiplied
by a factor of about 40% to reflect cowpea demand for northern Nigeria only. We assume that the average market-clearing
price of about US$391/ton is observed across the region and
it is used to adjust the parameters for cowpea supply from all
of Nigeria, Niger and Benin such that regional cowpea demand
equals regional cowpea supply at the market clearing price. The
market-clearing price of US$391 is the average price for Benin,
Niger, and Nigeria around 1999; the national prices were from
Langyintuo (2003) and are presented in Appendix D. With a
market clearing price of about US$391/ton, cowpea regional
demand stands at about 1,260,000 tons of cowpea and hence,
cowpea supply at the regional level has to be reduced by a
factor of about 44% to equalize regional cowpea demand and
supply. Hence, the parameters for the supply function estimated

12

S. D. Gb`egb`el`egb`e et al./Agricultural Economics 46 (2015) 1–15

Source: authors’ computations.
Fig. 3. Illustration of economic impact of Bt cowpea in block formed by Benin,
Niger, and Northern Nigeria.

by Langyintuo (2003) were multiplied by a factor of about 44%
to obtain the adjusted parameters used in this study. Our approach implies that Benin, Niger, and northern Nigeria form
a regional trade block involving no barriers to cowpea trade
among efficient markets within the block and no cowpea trade
with other regions in the world. Evidence suggests that cowpea
trade for Benin, Niger, and Nigeria is mostly with each other
(Langyintuo, 2003).
In Fig. 3(a), the market equilibrium prior to the introduction of Bt cowpea involves a cowpea market price of
about US$391/ton and annual cowpea quantity sold of about
1.26 million tons. Figure 3(b) illustrates aggregate demand

and supply curves for both Bt and conventional cowpea in
all of Benin, Niger and northern Nigeria, given the introduction of Bt cowpea. The estimated demand functions for Bt
and conventional cowpea, given the introduction of Bt cowpea, are derived directly using the demand functions whose
parameters are in Table 8. For example, the aggregated urban demand functions for Bt cowpea in northern Nigeria
(−1.94)∗ PBt +1
) where the value of
is QBt
Nigeria−aggr = (1.15E8)(e
1.15E8 is the product of 12 months and the size of the urban
population in northern Nigeria in 2005 (Appendix D).
We use an adjusted version of the method developed by
Masters et al. (1996) to estimate cowpea supply given the introduction of Bt cowpea. We incorporate the plausible cowpea
acreage changes that could occur with the introduction of Bt
cowpea; the method proposed by Masters et al. (1996) assumes
no changes in crop acreage with the introduction of a new
technology. The approach used in estimating the supply of Bt
cowpea before the adjustment in cowpea acreage is explained
in detail in Appendix E. We incorporate cowpea acreage adjustments through supply elasticity changes which are based
on changes in the price elasticity of cowpea demand with and
without Bt cowpea.
At the market equilibrium, before the introduction of Bt cowpea, the price elasticity of demand for conventional cowpea is
about –0.19,1 based on the results from Langyintuo et al. (2003)
which are illustrated in Fig. 3(a). With the hypothetical introduction of Bt cowpea, the price elasticity of demand for Bt
cowpea is 3.5 times higher than the price elasticity of conventional cowpea demand before the introduction of Bt cowpea,
at the price of US$391/ton (computations not shown). The estimation of the price elasticity of the Bt cowpea demand is
derived from one aggregated function consisting of the sum of
the national demand functions whose parameters are listed in
Appendix D.
Based on the changes in cowpea demand before and after
the hypothetical introduction of Bt cowpea, we assume that the
price elasticity of supply observed at the original market equilibrium would increase by 1.76 times (half of the increase in
price elasticity of demand) from about 0.20,2 after the introduction of Bt cowpea. The higher supply elasticity of about 0.35
implies that the slope of the aggregate cowpea supply function
(639.07 = 552.04 + 73.53 + 13.50) must also be multiplied
by 1.76. The percentage changes, as estimated by the approach
proposed by Masters et al. (1996), are then applied to the more
elastic supply function to obtain the Bt cowpea supply function.
Based on Fig. 3(b), with the hypothetical introduction of
Bt cowpea in all three regions, Bt cowpea would become the
only cowpea product sold in the market and the new market
equilibrium would involve about 2.2 million tons of Bt cowpea
traded and an optimal price of US$748.02/ton for Bt cowpea.
–0.19 = (aggregate demand slope) × ((equilibrium price)/(equilibrium
quantity)) = sum(–535.15, –19.33, –49.52)×(391) / 1,257,889.
2 0.20 = (aggregate supply slope) × (equilibrium price)/(equilibrium quantity) = sum(552.04, 73.53, 13.50) × (391)/(1,257,889).
1

D. S. Gb`egb`el`egb`e et al./Agricultural Economics 46 (2015) 1–15

13

Table 4
Economic impact of Bt cowpea in block formed by Benin, Niger, and Northern Nigeria

Variable

Change caused by Bt
cowpea – no inefficiencies in
seed sector and acreage
increase for Bt cowpea
supply function

Change caused by Bt
cowpea – no inefficiencies in
seed sector and no acreage
change for Bt cowpea
supply function

Change caused by Bt
cowpea – inefficiencies in
seed sector and acreage
increase for Bt cowpea
supply function

Consumer surplus
Producer surplus
Net social welfare
Net social welfare per capita

−$172,544,267
$1,020,644,350
$848,100,083
$ 11.82

−$394,790,781
$1,080,458,295
$685,667,513
$9.56

$425,109,355
−$334,698,806
$90,410,548
$ 1.26

Without Bt cowpea
Market equilibrium
Consumer surplus
Producer surplus
Net social welfare

(Price: US$ 391; qty: 1,257,889)
$654,343,723
$442,647,917
$1,096,991,641

Source: authors’ computations.

The changes that Bt cowpea would cause in the cowpea demand and supply in Benin, Niger and northern Nigeria can be
used to estimate its net impact on societal welfare in the region. In Fig. 3(b), with the introduction of Bt cowpea, market
equilibrium in the region would move from ‘e1 ’ to ‘e2 ’.
Table 4 presents results from applying mathematical integration techniques on the supply and demand functions illustrated
in Fig. 3(b) to estimate the net economic impact of Bt cowpea in the regional block formed by Benin, Niger and northern
Nigeria. Based on Table 4, given no inefficiencies in cowpea
markets across the block formed by Benin, Niger and northern Nigeria, consumers would experience a net welfare loss of
about US$172million/year, while producers would experience
a welfare improvement valued at about US$1,020 million/year,
if Bt cowpea were introduced in the region. As a result, net social welfare would increase by about US$848 million/year due
to Bt cowpea. The net welfare increase brought by Bt cowpea
translates into some net welfare increase of about US$11.82 per
person in the study region. It is important to note that this welfare change is the minimum welfare change that Bt cowpea can
bring given the yield, income and price risks faced by economic
agents in the region.

7. Sensitivity analysis on welfare effects of Bt cowpea
If we assume no change in the acreage allocated to cowpea
production, the slope of the aggregated Bt supply function is the
same as that of the aggregated supply of conventional cowpea.
In such a scenario, consumer loss would rise to US$395 million;
producer gain would reach US$1,080 million and social welfare
would increase by US$9.56 per person in the region (Table 4).
In both scenarios, where the increase in Bt cowpea acreage is
null and where it is positive, consumers stand to lose with the
introduction of Bt cowpea. However, producer gains would be
large enough to compensate the consumer losses.

The ex ante adoption rate of Bt cowpea in our study is high,
since most surveyed farmers would buy and plant Bt cowpea
if the latter is sold at the average market price. Other studies
have reported high ex post adoption rates of an improved cowpea variety in the semi-arid (Inaizumi, et al., 1999) and dry
savannas of Nigeria (Kristjanson, et al., 2005). A key factor
that could hinder the adoption of improved cowpea technologies, including Bt cowpea, is a weak seed industry. The cowpea
seed business is not well developed in West and Central Africa
where most cowpea is grown. Hence, it becomes difficult for
farmers to access improved cowpea.
Table 4 presents the net economic impact of Bt cowpea given
inefficiencies in the seed industry in that farmers in Benin,
Niger and northern Nigeria can access Bt cowpea seeds 50%
of the time. In such a scenario, it can be assumed that half of
the cowpea supplied every year would be Bt cowpea and the
other half would be conventional cowpea. If consumers’ fear
of unsafe conventional cowpea is still present and they cannot
distinguish between Bt and conventional cowpea, the demand
for cowpea would remain the same as what was observed before
the hypothetical introduction of Bt cowpea. Fig. 3(c) illustrates
cowpea market equilibrium before and after Bt cowpea given
inefficiencies in the seed sector. In Fig. 3(c), market equilibrium without Bt cowpea is represented by e1 . With the hypothetical introduction of Bt cowpea and given the inefficiencies
in the seed sector, market equilibrium moves from e1 to e3 in
Fig. 3(c). The values in Table 4 imply that producers are big
losers in such a scenario: they would experience net welfare
deterioration with the hypothetical introduction of Bt cowpea,
assuming that the technology makes supply more elastic. Consumers would gain more than in the scenario where the seed
sectors are working smoothly, as cowpea prices fall from about
US$391 to US$100/ton. All in all, society as a whole would
still benefit from Bt cowpea when seed sectors are inefficient:
net social welfare would increase by about US$89 million with
the hypothetical introduction of Bt cowpea, even if there are
inefficiencies in the seed sector; such increase corresponds to

14

S. D. Gb`egb`el`egb`e et al./Agricultural Economics 46 (2015) 1–15

about 10% of the net welfare increase brought by Bt cowpea
given no inefficiencies.

8. Conclusion
Because Bt cowpea is not yet on the market, this study uses a
choice experiment combined with cheap talk to estimate certain
demands for Bt cowpea and quantify the ex ante economic impact of Bt cowpea for cowpea growers and consumers in Niger,
Benin, and northern Nigeria. Our study reports a premium for
GM cowpea for urban consumers in Benin and Nigeria; however, urban consumers in Niger tend to prefer conventional to
Bt cowpea. Rural consumers in the three regions tend to prefer
Bt over conventional cowpea. Our results are similar to those
of Kikulwe et al. (2011) who found that rural consumers in
Uganda were willing to buy GM bananas if the latter shared
the same price as non-GM bananas but had additional benefits
including reduced pesticide use.
Relative to farmers’ WTP for GM products, our results on
the premium of Bt over conventional cowpea seeds for cowpea growers in Nigeria and Benin compare well with those of
Krishna and Qaim (2008) who found that eggplant farmers in
India would be willing to pay about five times more for Bt
eggplant seeds compared to conventional seeds. Just like the
farmers in West Africa, those in India have also been exposed
to the health hazards caused by insecticide misuse (Krishna and
Qaim, 2008).
In line with other studies (Lusk, 2003, Gonz´alez et al., 2009),
our results suggest that prior knowledge of GMOs affects WTP.
Other variables which also influence WTP are the gender and
education of the household head.
Since there is no clear evidence that cheap talk eliminates hypothetical bias, the results on the demand functions for Bt and
conventional cowpea could have been validated with experimental auctions using conventional cowpea only. The results
from such auctions would be similar to the results illustrated in
Fig. 3(b) for these consumers who would prefer Bt to conventional cowpea once the former is available. On another note, the
results from the experimental auctions might not be as accurate
as the ones from the choice experiment: Corrigan et al. (2009)
have shown that WTP elicited through an open ended choice
experiment, like the one used in this study, might provide more
reliable results compared to experimental auctions.
Under the assumption of an efficient seed system and with no
barriers to trade, the adoption of Bt would increase social welfare by at least US$848 million/year with most of the benefits
accruing to farmers. Other empirical studies have found that the
potential benefits of GM crops to reduce pesticide application
primarily benefitted farmers (Qaim, 2001; Dillen et al., 2009).
In this study, producers’ gains are larger than consumers’ in
part because the Bt cowpea supply is more price inelastic than
the Bt cowpea demand.
Using population proportions, the annual welfare gain
brought by Bt cowpea is around US$625, US$ 154, and US$

68 million for northern Nigeria, Niger, and Benin, respectively.
These results are consistent with those from Krishna and Qaim
(2008) and higher than those from Kostandini et al. (2009).
Our estimates are gross benefits and should be higher than the
net effects of Bt cowpea. The results also highlight the importance of an efficient seed system in ensuring that producers and
consumers can fully benefit from improved technologies.
One policy implication consists of strengthening the seed
multiplication and dissemination systems in Africa. Bottlenecks in seed value chains in Africa have been documented
and African countries are trying to address them through their
trade agreements. Another policy recommendation is for countries in West Africa to streamline their bio-safety regulations.
As shown in this study, some African farmers would prefer
to grow GM crops. Given the porous borders in West Africa
along with the fact that some countries, such as Burkina Faso,
are currently growing GM crops, it would not be surprising
to see farmers grow GM crops in other countries where governments are opposed to GMOs. In Burkina Faso, bio-safety
regulations are already implemented and Nigeria is currently
testing Bt cowpea. Without streamlined bio-safety regulations
across West Africa, it would also be a difficult for seed companies to make a commercially viable endeavor from Bt cowpea.
Acknowledgments
The research was funded by the African Agricultural Technology Foundation. However, the ideas expressed here are those
of the authors and do not necessarily reflect those of AATF. The
authors would like to thank the following people for collecting
the data used in the study: Edmond Kpoffon, Martial Zannou,
Pierre Kpoffon for Benin; Andr´e Boubakar, Belko Moussa,
Illyassou Abdou, Kolo Katiela, Malam Ma¨ı Adji, Souleymane Abdou, and Tanko Adamou for Niger; Ibrahim Maina,
Kakka K. Aluwong, Mohammad Hadi Haruna, Aminu Sanusi,
Kabiru Yusuf, Margaret Adegbulu, Nura Gambo, Hashim
Ibrahim Hashim, Yohanna Ezekiel, Yunana Ciroma, and Yusuf
Abubakar for Nigeria; and Casimir Aitchedji for the data collected in Benin, Niger, and Nigeria.
References
AATF, 2011. Maruca Resistant Cowpea Project Progress Report. African Agricultural Technology Foundation.
Adigoun, F.A., 2002. Impact des traitements phytosanitaires du ni´eb´e sur
l’environnement et la sant´e des populations: cas de Klou´ekanm´e et de la
basse vall´ee de l’Ou´em´e (B´enin), Faculty of Letters, Arts and Human Sciences (FLASH). University of Abomey Calavi, Abomey Calavi.
Agazounon, R.C.B.E., 2003. Cadre th´eorique et m´ethodologique de l’´etude in
Etude de rentabilit´e financi`ere et e´ conomique de la transformation du ni´eb´e
au B´enin. Universit´e d’Abomey-Calavi, Abomey-Calavi.
Aitch´edji, C., Coulibaly, O., Lowenberg-DeBoer, J., 2004. Adoption of cowpea storage technologies in the main cowpea growing area of Benin (west
Africa): preliminary report. Purdue University and IITA, Cotonou.
Balla, A., Barag´e, M., 2006. Influence de la vari´et´e, du temps de stockage et du
taux de natron sur la cuisson des graines de ni´eb´e. Tropicultura. 24, 39–44.

D. S. Gb`egb`el`egb`e et al./Agricultural Economics 46 (2015) 1–15
Boardman, A.E., Greenberg, D.H., Vining, A.R., Weimer, D.L., 1996. Option
price, option value, and quasi-option value. In: Jewell, L. (Ed.), Cost-Benefit
Analysis: Concepts and Practice. Prentice-Hall, Inc., Upper Saddle River.
Corrigan, J.R., Depositario, D.P.T., Nayga Jr, R.M., Wu, X., Laude, T.P., 2009.
Comparing open-ended choice experiments and experimental auctions: An
application to golden rice. Am. J. Agric. Econ. 91, 837–853.
De Groote, H., Overholt, W.A., Ouma, J.O., Wanyama, J., 2011. Assessing the
potential economic impact of Bacillus thuringiensis (Bt) maize in Kenya.
Afr. J. Biotechnol. 10, 4741–4751.
Desvousges, W.H., Smith, V.K., Fisher, A., 1987. Option price estimates for
water quality improvements: A contingent valuation study for the monongahela river. J. Environ. Econ. Manage. 14, 248–267.
Dillen, K., Demont, M., Tollens, E., 2009. Potential economic impact of GM
sugar beet in the global sugar sector. Int. Sugar J. 111, 638–643.
Dugje, I.Y., et al., 2009. Production du ni´eb´e en Afrique de l’Ouest: Guide du
paysan. International Institute for Tropical Agriculture (IITA), Ibadan.
Freeman, A.M., 1993. The Measurement of Environmental and Resource Values: Theory and Methods. Resources for the Future, Washington, DC.
Gbegbelegbe D.S., 2008. Ex ante economic impact of Bt Cowpea in Nigeria,
Niger and Benin. Department of Agricultural Economics. Purdue University,
West Lafayette.
Gonz´alez, C., Johnson, N., Qaim, M., 2009. Consumer acceptance of secondgeneration GM foods: The case of biofortified cassava in the North-east of
Brazil. J. Agric. Econ. 60, 604–624.
Inaizumi, H., Singh, B.B., Sanginga, P.C., Manyong, V.M., Adesina, A.A.,
Tarawali, S., 1999. Adoption and Impact of Dry-Season Dual-Purpose Cowpea in the Semiarid Zone of Nigeria. IITA.
Kikulwe, E.M., Wesseler, J., Falck-Zepeda, J., 2011. Attitudes, perceptions,
and trust. Insights from a consumer survey regarding genetically modified
banana in Uganda. Appetite. 57, 401–413.
Kostandini, G., Mills, B.F., Omamo, S.W., Wood, S., 2009. Ex ante analysis
of the benefits of transgenic drought tolerance research on cereal crops in
low-income countries. Agric. Econ. 40, 477–492.
Krishna, V.V., Qaim, M., 2008. Potential impacts of Bt eggplant on economic
surplus and farmers’ health in India. Agric. Econ. 38, 167–180.
Kristjanson, P., Okike, I., Tarawali, S., Singh, B.B., Manyong, V.M., 2005.
Farmers’ perceptions of benefits and factors affecting the adoption of

15

improved dual-purpose cowpea in the dry savannas of Nigeria. Agric. Econ.
32, 195–210.
Langyintuo, A.S., 2003. Cowpea trade in West and Central Africa: A spatial and temporal analysis. Department of Agricultural Economics. Purdue
University, West Lafayette.
Langyintuo, A.S., Lowenberg-DeBoer, J., 2006. Potential regional trade implications of adopting Bt cowpea in west and central Africa. AgBioForum. 9,
111–120.
Langyintuo, A.S., Lowenberg-DeBoer, J., Faye, M., Lambert, D., Ibro, G.,
Moussa, B., Kergna, A., Kushwaha, S., Musa, S., Ntoukam, G., 2003. Cowpea supply and demand in West and Central Africa. Field Crop. Res. 82,
215–231.
Louviere, J.J., Hensher, D.A., Swait, J.D., 2000. Experimental design. Stated
Choice methods: Analysis and Application. Press Syndicate of University
of Cambridge, Cambridge.
Lusk, J.L., 2003. Effects of cheap talk on consumer willingness-to-pay for
golden rice. Am. J. Agric. Econ. 85, 840–856.
Lusk, J.L., Schroeder, T.C., 2004. Are choice experiments incentive compatible? A test with quality differentiated beef steaks. Am. J. Agric. Econ. 86,
467–482.
Malick Ba, N., Margam, V.M., Binso-Dabire, C.L., Sanon, A., McNeil, J.N., Murdock, L.L., Pittendrigh, B.R., 2009. Seasonal and
regional distribution of the cowpea pod borer Maruca vitrata (Lepidoptera: Crambidae) in Burkina Faso. Int. J. Trop. Insect Sci. 29, 109–
113.
Masters, W., et al., 1996. The economic impact of agricultural research: A
practical guide. Purdue University, West Lafayette.
PAN and IPEN, 2009. ‘Endosulfan in West Africa: Adverse effects, its banning, and alternatives’. http://ipen.org/sites/default/files/documents/ipen_
endosulfan_westafrica-en.pdf. 10–11.
PAN UK, 2001. Rapport d’enquˆete sur l’effet de l’utilisation des pesticides
chimiques sur l’homme en Afrique de l’Ouest, Pesticide Action Network
UK, London.
Qaim, M., 2001. A prospective evaluation of biotechnology in semi-subsistence
agriculture. Agric. Econ. 25, 165–175.
USDA, 2014. Biotechnology. http://www.ers.usda.gov/topics/farm-practicesmanagement/biotechnology.aspx.