BRIEFING: BURKINA FASO'S REVERSAL ON GENETICALLY MODIFIED COTTON AND THE IMPLICATIONS FOR AFRICA Downloaded from

Full text

BRIEFING: BURKINA FASO’S REVERSAL
ON GENETICALLY MODIFIED COTTON
AND THE IMPLICATIONS FOR AFRICA
BRIAN DOWD-URIBE*AND MATTHEW A. SCHNURR
CAN GENETICALLY MODIFIED (GM) CROPS help smallholder farmers in
sub-Saharan Africa? To date, only two GM crops –insect-resistant forms of
cotton and maize –have made it into the hands of African farmers. Of
these, GM cotton has the longest empirical track record, having been the
first GM crop ever introduced in Africa, and the only one that has been
grown in multiple countries –first South Africa, then Burkina Faso.
1
The
performance of this crop has received intense scrutiny, as it offers the best
indication of how the suite of other GM crops slated for commercial ap-
proval may perform across the continent.
This briefing reviews the experiences of South African farmers with GM
cotton, which has emerged as the crucial precedent highlighting the value of
GM crops for poor farmers. It then turns to the case of Burkina Faso, which
became the showcase for how GM crops can benefit smallholder African
farmers. However, as shown here, Burkina Faso has begun a complete phase-
out of GM cotton, citing the inferior lint quality of the GM cultivars as the
reason for abandoning its cultivation. Burkina Faso’s phase-out could stall or
even end negotiations to adopt GM cotton in other Francophone African
countries with similar concerns over cotton quality. More generally, Burkina
Faso’s reversal could undermine public trust in GM crops across the contin-
ent at a time when many African countries are grappling with the politicized
and polarized debate over whether to adopt these new breeding technologies.
We argue that the retreat of Burkina Faso, one of the most prominent and
*Matthew A. Schnurr (matthew.schnurr@dal.ca) is Associate Professor in the Department of
International Development Studies at Dalhousie University. Brian Dowd-Uribe (bdowduribe@
usfca.edu) is Assistant Professor in the International Studies Department at the University of
San Francisco. Fieldwork undertaken in summer 2015 was funded by the Social Sciences and
Humanities Research Council of Canada. We would also like to acknowledge the indispensable
insights and assistance offered by our friend and colleague Hamadou Diallo. Both authors con-
tributed equally to this article.
1. GM cotton has also been grown in Sudan since 2012. See Nagala A. Abdallah, ‘The story
behind Bt cotton: Where does Sudan stand?’,GM Crops and Food 5, 4 (2014), pp. 241–3.
African Affairs,1–12 doi: 10.1093/afraf/adv063
© The Author 201 . Published by Oxford University Press on behalf of Royal African Society. All rights reserved
1
6
African Affairs Advance Access published January 4, 2016
at University of San Francisco on January 13, 2016http://afraf.oxfordjournals.org/Downloaded from

vocal supporters of GM crops on the continent, could have significant impli-
cations for the future of GM crops in Africa.
2
GM cotton in South Africa
The first GM crops grown on the continent were planted in South Africa,
which permitted the commercialization of Monsanto’s insect-resistant (Bt)
cotton in 1997. Bt cotton is genetically modified by inserting a bacterial
gene that secretes a protein fatal to larvae from the genus Lepidoptera,
which are among the most pernicious cotton pests. Larger-scale, commer-
cial growers, who immediately realized the financial benefit of reduced
pesticide applications and increased yields, readily adopted Bt cotton. One
year after its initial release, Monsanto launched a targeted campaign to in-
crease adoption among smallholder cotton farmers in the Makhathini Flats,
a remote rural district just south of the border with Mozambique. Initial
accounts of Bt cotton’s performance in Makhathini were extremely positive,
reporting gains in average yields and profits, as well as a significant reduc-
tion in pesticide applications.
3
Makhathini emerged as the prototype for
how GM crops could improve yields and livelihoods for farmers across
sub-Saharan Africa.
4
But the success of GM cotton in Makhathini did not last long. After only
a few years of operation, the cotton company that operated the local gin and
provided credit for the purchase of the more expensive GM seeds went
bankrupt. Production levels recovered briefly following the arrival of a new
cotton company, buoyed by joint ventures in which the company took over
the management of farmers’lands to maximize economies of scale, as well
as incentives that privileged the adoption of Bt seeds (by excluding appropri-
ately sized packages of non-Bt seed and refusing to accept non-Bt seed for
ginning). But this scheme also folded after only a few years, unable to trans-
form a patchwork of smallholder producers into a more financially viable
model that was centralized, heavily mechanized, and revolved around cotton
monocultures. Within ten years of its introduction, most growers had aban-
doned Bt cotton altogether. The most recent cotton production figures
2. The interviews upon which this article is based were undertaken in person in
Ouagadougou and Bobo in early July 2015. Subsequent interviews were undertaken by phone
in August and September 2015. Recognizing the highly contentious nature of the debate over
GM crops in Burkina Faso, we have decided not to reveal informant names. Instead, we refer
to informants by affiliation and interview date only.
3. Richard Bennett, Stephen Morse, and Yousouf Ismael, ‘The economic impact of genetic-
ally modified cotton on South African smallholders: Yield, profit and health effects’,Journal of
Development Studies 42, 4 (2006), pp. 662–77; Colin Thirtle, Lindie Beyers, Yousouf Ismael,
and Jenifer Plesse, ‘Can GM-technologies help the poor? The impact of Bt cotton in
Makhatini Flats, Kwazulu-Natal’,World Development 31, 4 (2013), pp. 717–32.
4. Matthew A. Schnurr, ‘Inventing Makhathini: Creating a prototype for the dissemination
of genetically modified crops into Africa’,Geoforum 43, 4 (2012), pp. 784–92.
2AFRICAN AFFAIRS
at University of San Francisco on January 13, 2016http://afraf.oxfordjournals.org/Downloaded from

available from Makhathini reveal the extent of this collapse. The total number
of Bt adopters in the 2014/15 growing season was below 5 percent of what it
was in the peak production years that followed the introduction of GM.
5
The South African case has three lessons for other African countries con-
sidering the introduction of GM crops. First, institutional dynamics are
crucial to a new technology’s success. In Makhathini, farmer enthusiasm
for the technology eroded once the enabling institutional environment of
easy access to credit and a guaranteed market disappeared. Second, GM
crop evaluations relying on measures of average profits and yields fail to ap-
preciate fully the comprehensive and longer-term impacts of these tech-
nologies. More specifically, these data often occlude which categories of
farmers benefit from these technologies, as well as how these technologies
disrupt farming systems. Aggregate data thus need to be contextualized and
extended over longer time periods to determine the implications of new
GM technologies for resource-poor and marginalized farmers, as well as
the implications for different commodity-chain actors.
Finally, the South African case underlines the potential gulf between the
representation and reality of GM adoption. GM proponents have used the
Makhathini case to promote and advance the introduction of GM crops in
other parts of sub-Saharan Africa, long after the benefits they offered to
farmers had declined.
6
The South African precedent is a reminder of how
an initial success story can endure and influence the political debate in
other locales, even after the real benefits to end users have faded.
Burkina Faso and Bt cotton
Today, Burkina Faso has surpassed South Africa as the model for how GM
crops can help African farmers. Many factors surrounding the introduction of
Bt cotton in Burkina Faso distinguish it from South Africa. Unlike in South
Africa, smallholders account for the vast majority of total cotton production in
Burkina Faso, and the country was the top cotton producer in Africa in 2015
with over 700,000 MT of seed cotton produced.
7
Also unlike South Africa,
Burkina Faso has a highly organized and regulated cotton industry. This vertical-
ly integrated cotton system, where companies operate regional monopolies with
a shared cotton purchase price, ensures the operation of a reliable credit market.
Cotton producers receive seeds and inputs such as pesticides and fertilizer on
credit provided by the cotton company and later sell their cotton back to the
same company for a guaranteed price. This centralized system has obvious
5. Cotton South Africa, ‘Small-holder cotton farmer production estimates’, 9 September
2015, <http://cottonsa.org.za/Report/GetReport/8> (10 September 2015).
6. Matthew A. Schnurr, ‘Inventing Makhathini’.
7. Ecobank, ‘Middle Africa briefing note, soft commodities, cotton’(Ecobank Research
Centre, Lomé, 12 June 2015).
BRIEFING 3
at University of San Francisco on January 13, 2016http://afraf.oxfordjournals.org/Downloaded from

benefits in ensuring farmers access to premium seeds and expensive inputs
required to achieve higher yields. It is also one of the crucial factors that made
Burkina Faso such an appealing destination for Bt cotton. In the words of one
Monsanto official, the heavily integrated and centralized administrative structure
‘made it attractive in that it would be easier to capture value for our investment’.
8
In 2003, the Burkinabè government signed a contract with Monsanto to
test their insect-resistant Bt cotton in experimental field trials. The first few
years of testing showed good resistance to Lepidopteron infestations and
yield improvements. But Burkinabè officials expressed reluctance over the
importation of American germplasm, which they considered to be inferior to
their own domestic cultivars. The Burkinabè cotton industry was fiercely
proud of the reputation of the cotton it produced, which, along with the
output of its other Francophone West African neighbours, was considered to
be the best on the continent after Egyptian Pima. Burkinabè officials voca-
lized these concerns to their Monsanto collaborators, insisting that the Bt
trait needed to be inserted into their local cultivars, since these were adapted
to local agro-climatic conditions and produced high cotton-quality character-
istics. Monsanto deferred. In the words of one Monsanto representative:
The government officials were not interested in us bringing our varieties in. So we decided
to work with them and introgress the traits into their local germplasm …they wanted it that
way and it preserves the agronomic qualities that they valued locally.
9
Monsanto scientists proceeded to backcross the Bt trait into the three most
widely grown cotton cultivars across Burkina Faso –FK 290, FK 37, and
STAM 59A –inviting government officials to assist in assessing their agro-
nomic performance.
10
After three generations of backcrossing. the company
announced that the new GM lines were stable and ready for commercial
release.
11
The Burkinabè cotton industry and Monsanto patented the resulting Bt
cultivars, releasing them to farmers in 2008. The adoption of Bt cotton
within the country skyrocketed in the ensuing years. By 2013, almost 70
percent of total cotton hectares were planted with Bt cultivars.
12
This rate
of adoption remained consistent up to 2015.
13
The only published studies
regarding the performance of Bt cultivars in Burkina Faso report an average
8. Interview, Monsanto official #2, by phone, St Louis, MO, 21 July 2015.
9. Ibid.
10. Oula Traoré, Sanfo Denys, Jeffery Vitale, K. Traoré, and Koulibaly Bazoumana,
‘Testing the efficacy and economic potential of Bollgard II under Burkina Faso cropping con-
ditions’,Journal of Cotton Science 12 (2008), pp. 87–98.
11. Interview, cotton agronomy expert, by phone, Lubbock, TX, 11 September 2015.
12. International Service for the Acquisition of Agri-Biotech Application (ISAAA), ‘Biotech
facts and trends: Burkina Faso’, 2015, <https://www.isaaa.org/resources/publications/biotech_
country_facts_and_trends/download/Facts%20and%20Trends%20-%20Burkina%20Faso.pdf>
(4 September 2015).
13. Ecobank, ‘Middle Africa briefing note, soft commodities, cotton’.
4AFRICAN AFFAIRS
at University of San Francisco on January 13, 2016http://afraf.oxfordjournals.org/Downloaded from

yield gain of 22 percent over conventional cultivars, and a profit gain of 51
percent for a Bt cotton-producing household with just over 3 hectares.
14
The Bt cottonseed price of US$60 per hectare remains much more expen-
sive than the conventional cottonseed price of $2.
15
However, conventional
cotton production requires more pesticide applications than Bt cotton; Bt
cotton farmers report that pesticide use declined from an average of six sprays
per growing season to two.
16
The profit gains for Bt producers stemmed from
the significant reduction in pesticide applications coupled with increased
yield, which outweighed the difference in seed cost.
17
Thedeclineintheuse
of pesticides also led to a reduced incidence of pesticide poisonings.
18
Burkina Faso’s success with Bt cotton has supplanted the Makhathini
Flats as the showcase for how GM crops can perform in the hands of African
smallholders
19
GM crop advocates have mobilized the triumph of Bt cotton
in Burkina Faso to encourage more permissive approaches to GM crops
across the continent. For example, the case of Burkina Faso figured promin-
ently in recent debates over the passage of the Nigerian biosafety law.
20
Burkina Faso also regularly hosts delegations from African nations to tour Bt
cotton fields and visit with cotton officials and farmers. Since the introduc-
tion of Bt cotton in 2008, Burkina Faso has received delegations from at least
seventeen different African nations, with many of these countries represented
on multiple occasions.
21
The goal for these visits, which were organized by
14. Jeff Vitale and John Greenplate, ‘The role of biotechnology in sustainable agriculture of
the twenty-first century: The commercial introduction of Bollgard II in Burkina Faso’,in
David D. Songstad, Jerry L. Hatfield, and Dwight T. Tomes (eds), Convergence of food security,
energy security and sustainable agriculture (Springer, Heidelberg and New York, NY, 2014),
pp. 239–93.
15. Brian Dowd-Uribe and James Bingen, ‘Debating the merits of biotech crop adoption in
sub-Saharan Africa distributional impacts, climatic variability and pest dynamics’,Progress in
Development Studies 11, 1 (2011), pp. 63–8.
16. The remaining two sprays are generally undertaken late in the growing season to ward off
damage by sucking pests such as aphids or jassids, which are not repelled by the Bt toxin.
17. Brian Dowd-Uribe, ‘Engineering yields and inequality? How institutions and agro-
ecology shape Bt cotton outcomes in Burkina Faso’,Geoforum 53 (2014), pp. 161–71.
18. Jeff Vitale and John Greenplate, ‘The role of biotechnology in sustainable agriculture’.
19. The Burkina Faso ‘success’story relies on studies that demonstrate average yield and
profit gain for farmers, though there remains more empirical work to be done on the potential
differential impacts of Bt cotton. Brian Dowd-Uribe, ‘Engineering yields and inequality?’.
20. Business Day,‘Why Burkina Faso has overtaken Nigeria in cotton production’,28
January 2015, <http://businessdayonline.com/2015/01/why-burkina-faso-has-overtaken-nigeria-
in-cotton-production/> (24 September 2015); Jimoh Babatunde, ‘Boosting cotton production:
Why Nigeria needs biosafety law’,Vanguard, 22 February 2015, <http://www.vanguardngr.com/
2015/02/boosting-cotton-production-why-nigeria-needs-biosafety-law/> (24 September 2015).
21. This number is derived from attendance at some of these events by the authors and
ISAAA news reports dated 16 December 2011, 27 November 2013, 8 January 2014, and 12
November 2014. These reports can be found at <www.isaaa.org> (29 October 2015).
Notably, many of these visits happened well after Burkinabè cotton sector officials knew about
cotton quality issues with Bt cultivars. The list of countries is: Togo, Benin, Ivory Coast,
Ghana, Chad, Niger, Nigeria, Senegal, Ethiopia, Kenya, Malawi, Tanzania, Sudan,
Swaziland, Uganda, Zambia, and Zimbabwe.
BRIEFING 5
at University of San Francisco on January 13, 2016http://afraf.oxfordjournals.org/Downloaded from

the International Service for the Acquisition of Agro-biotech Applications
(ISAAA) and Monsanto, was ‘that the participants will use this experience to
expedite the commercialization process in their respective countries for the
benefitofcottonfarmers’.
22
The importance of quality
One of the most important traits distinguishing West African cotton from
its international competitors is its quality. We use the term quality here to
denote two groups of related phenomena. The first relates to the quality fea-
tures of the fibre. Burkinabè cultivars are the product of decades of careful
breeding that has resulted in premium cotton fibres, which are long, strong,
and uniform. These traits are highly sought after for the production of
high-end textiles and fetch a premium on the global market. The second
reason why Burkinabè cotton fibre is of such high quality stems from it
being hand-picked, which ensures that the fibre is free of other organic
matter.
23
Hand picking influences the other quality trait valued by cotton
companies, known as the ginning ratio, which is the percentage of fibre per
unit weight of cotton delivered to the gin. The ginning ratio of Burkinabè
cotton is high, the result of decades of targeted breeding and careful hand
picking. A high ginning ratio is attractive to Burkinabè cotton companies
since it increases the total amount of fibre that it can sell at a high value
compared to the total harvest weight. Burkinabè cotton has gained a
stellar international reputation and a premium price based on these quality
traits.
The higher quality of Burkinabè conventional cotton is the result of a very
successful breeding programme that has spanned almost seventy years.
24
In
1946, the French government founded the Institute for Research on Cotton
and Tropical Textiles, known by its French acronym ICRT, to lead cotton-
breeding programmes for its African colonies. The cultivation of breeding
techniques begun by ICRT became part of the French agricultural research
organization, CIRAD, in 1984, and was eventually absorbed into national re-
search institutes. The main goal of the ICRT-CIRAD breeding programme
was to create cultivars that were well adapted to the growing conditions in
22. International Service for the Acquisition of Agri-Biotech Applications (ISAAA), ‘African
delegation visits Burkina Faso Bt cotton fields’, 26 November 2010, <http://www.isaaa.org/kc/
cropbiotechupdate/article/default.asp?ID=6987> (4 September 2015).
23. Gérald Estur,‘Quality and marketing of cotton lint in Africa’(Africa Region Working
Paper Series No. 121, World Bank, Washington, DC, 2008).
24. It is worth noting that late nineteenth- and early twentieth-century colonial governments
originally sought to improve cotton quality characteristics in Africa by introducing American
cultivars of cotton. Now, over a century later, American transgenic cultivars are the reason for
a decrease in the quality of African cotton.
6AFRICAN AFFAIRS
at University of San Francisco on January 13, 2016http://afraf.oxfordjournals.org/Downloaded from

West Africa, and exhibited desired quality characteristics such as a high
ginning ratio and long staple length.
25
This breeding programme achieved
considerable success. Between the 1970s and 2006/7, the average ginning
ratio for Burkinabè cotton increased from 36 percent to 42 percent. The
improved ginning ratios in the Francophone West African cotton sector in
general, and the Burkinabè cotton sector in particular, were to a large extent
the distinguishing feature that made them more competitive in the global
market.
26
As Figure 1shows, the ginning ratios in other African countries
were unable to match this progress over the same period of time.
The cotton-breeding programme in Burkina Faso also made consider-
able improvement in staple length over this time period. The standard
benchmark in the cotton industry for the more desirable medium-to-long
cotton fibres is 1
1
/
8
(inches). The percentage of total Burkinabè cotton
classified as longer than this benchmark rose from 20 percent in 1995/6 to
80 percent of total cotton production in 2005/6.
27
Given the potential link between breeding and the beneficial traits of
long fibre length and high ginning ratios, one observer wondered whether
adopting Bt cotton would degrade these quality characteristics, as ‘bio-
technology changes may …modify cotton fibre quality and …Bt cotton
may jeopardize West Africa’s world cotton market advantage’.
28
But
Burkina Faso went ahead with Bt cotton adoption despite these concerns.
Burkinabè officials were initially satisfied with the quality characteristics
present after reviewing early field trials, reporting that ‘the fibre’scharac-
teristics were maintained’.
29
In the words of one Monsanto official familiar
with the trials, ‘All I can say is based upon the assessments that we made
with their help, we were achieving things that were satisfactory to them at
the time.’
30
25. Tom Bassett, A peasant cotton revolution (Oxford University Press, Oxford, 2001); Jim
Bingen, ‘Cotton in West Africa, a question of quality’, in Jim Bingen and Lawrence Busch
(eds), Agricultural standards: The shape of the global food and fibre system (Springer, Dordrecht,
2006), pp. 7–8; Alfred Schwartz, ‘L’évolution de l’agriculture en zone cotonnière dans l’Ouest
du Burkina Faso’. In Jean-Claude Devèze (ed), Défis Agricoles Africains (Karthala, Paris,
2008), pp. 153–172.
26. The difference of just a few percentage points in the ginning ratio can lead to severe fluc-
tuations in revenues. Consider Burkina Faso’s cotton output in 2008/9. In that year Burkina
Faso produced 452,000 tons of seed cotton. At an average ginning ratio of 42 percent this
translates to 189,840 tons of cotton lint at a price of roughly $1,000 per ton. A 2 percent drop
in ginning ratio to 40 percent would translate into 180,800 tons, or a difference of about 9,000
tons of fibre, valued at more than US$9 million.
27. Gérald Estur,‘Quality and marketing of cotton lint in Africa’.
28. Jim Bingen, ‘Genetically-engineered cotton: Politics, science and power in West Africa’,
in William. G. Moseley and Leslie C. Gray (eds), Hanging by a thread: Cotton, globalization,
and poverty in Africa (Ohio University Press, Athens, OH, 2008), pp. 227–57.
29. Interview, cotton company official, Ouagadougou, Burkina Faso, 2 July 2015.
30. Interview, Monsanto official #2.
BRIEFING 7
at University of San Francisco on January 13, 2016http://afraf.oxfordjournals.org/Downloaded from

Problems with quality
Burkinabè officials noticed declines in both staple length and ginning ratios
during the first years of commercial release.
31
Monsanto officials were scep-
tical, suggesting that these initial declines in staple length and ginning
ratios were due to exceptional water stress and other climatological varia-
tions.
32
But this deterioration in ginning ratios and staple length persisted
over time. Reports from Burkinabè officials, which were corroborated by
Figure 1. Ginning ratios in three African countries, 1980–2007
Source: David L. Tschirley, Colin Poulton, and Patrick Labaste, ‘Organization and
performance of cotton sectors in Africa: Learning from reform experience’(World Bank,
Washington, DC, 2009), pp. 180–1.
31. Brian Dowd-Uribe, Engineered outcomes: The state and agricultural reform in Burkina Faso
(University of California, PhD thesis, 2011); Interview, cotton company official.
32. The Burkinabè cultivars in use were known to exhibit variance in ginning ratios due to
environmental considerations. Dominique Dessauw and Bernard Hau, ‘Cotton breeding in
French-speaking Africa: Milestones and prospects’, paper presented at the World Cotton
Research Conference 4 (Omnipress, Lubbock, TX, 2008).
8AFRICAN AFFAIRS
at University of San Francisco on January 13, 2016http://afraf.oxfordjournals.org/Downloaded from

Monsanto, confirm that Bt cultivars produced fibres that were
1
/
32
of an
inch shorter than conventional varieties.
33
In the 2013/14 season, over two-
thirds of the nation’s total crop was classified as lower-quality medium
(with a staple length between 1
3
/
32
and 1
1
/
16
), with only a third retaining
its previous classification as medium to high staple length. This represented
a decline of over 40 percent since 2005/6.
34
The precise decline in ginning
ratios is more difficult to measure, though Burkinabè officials confirm that
it remains well below the 42 percent achieved by conventional cultivars.
35
This decline in staple length has undermined the reputation of
Burkinabè cotton and cut into its value on the international market. When
coupled with the decline in overall lint due to the lower ginning ratio, the
inferior quality characteristics of the Bt cultivars have compromised the
economic position of Burkinabè cotton companies. The lower-quality
fibre was valued less highly by spinners who could only use the poorer
grade for the production of lower-quality textiles, such as bedding. It also
complicated trading arrangements among other West African producers
such as Côte d’Ivoire and Mali. All Francophone West African producers
aim for a homogeneous product that can be interchanged to facilitate
timely delivery to clients, but Burkina Faso’s poor staple length undercut
this flexible sourcing mechanism.
36
In the most recent growing season,
Burkina Faso produced over 700,000 MT of cotton, while its western
neighbour Mali produced only 500,000 MT, yet within a few months
Mali’s entire product had been sold on the international market, while
most of Burkina Faso’s languished awaiting export. As one high-ranking
official lamented, ‘What is the point in being the top producer if you can’t
even sell your cotton?’
37
Breeders are struggling to account for these declines in ginning ratio and
staple length. In theory, inserting the Bt gene into the Burkinabè germ-
plasm should have left the resultant progeny identical to its parent in every
way except for the inserted trait conferring insect resistance. But, in reality,
the process of introgressing the Bt trait into the local variety appears to have
interfered with some of its most important characteristics. Monsanto scien-
tists are at a loss to explain the precise mechanism that has created these
problems.
38
The company is attempting to identify and correct this fault.
In the short term, Monsanto has proposed forming a technical committee
of local and international experts to investigate this issue of declining
33. Interview, Monsanto official #1, Bobo, Burkina Faso, 1 July 2015; Interview, cotton
company official, Ouagadougou, Burkina Faso, 2 July 2015.
34. Ecobank, ‘Middle Africa briefing note, soft commodities, cotton’,p.2.
35. Interview, cotton company official.
36. Ecobank, ‘Middle Africa briefing note, soft commodities, cotton’,p.3.
37. Interview, cotton company official.
38. Interview, Monsanto official #1.
BRIEFING 9
at University of San Francisco on January 13, 2016http://afraf.oxfordjournals.org/Downloaded from

quality and propose recommendations for moving forward. In the medium-
to-long term, Monsanto has embarked upon a new process of backcrossing
the Bt trait into a new local cultivar, known as FK64. The company pro-
mises to use ‘new tools and processes’to ensure that the resulting back-
crosses do not suffer similar deteriorations in quality.
39
Burkina Faso’s cotton companies have grown impatient and decided to
take matters into their own hands. Frustrated with Monsanto’s inability to
identify and correct these declines in quality, the companies set a timeline
for abandoning Bt cotton and returning to conventional Burkinabè culti-
vars. Their centralized control over the country’s seed supply allowed them
to reduce the availability of Bt cottonseed from the peak rate of adoption of
73 percent in 2014/15 to 53 percent in 2015/16. They plan on reducing this
amount to 30 percent in the 2016/17 growing season, with the goal of a
complete return to conventional cotton in time for the 2017/18 season. The
cotton companies also made a formal request to Monsanto for losses in-
curred due to these declines in quality. They are demanding more than
FCFA 30 billion (approximately US $280 million) as compensation for
losses sustained since 2010.
Implications for GM crop adoption across Africa
The story of Bt cotton in Burkina Faso raises important questions about the
commercial production and dissemination of GM crops in Africa. The first
concerns the narrow scope of the GM insect-resistance breeding pro-
gramme. Bt cotton was originally bred in the United States with the sole
aim of conferring the Bt trait into a cultivar that would express the toxin
consistently. This exclusive focus on pest mitigation contrasts sharply with
the Francophone West African breeding programmes, which spent decades
integrating a broad spectrum of adaptability to growing conditions along-
side multiple characteristics of fibre quality. The Burkinabè cotton industry
astutely tried to remedy the undesirable characteristics of the American cul-
tivar by backcrossing it into its own cultivars. But quality suffered. This
failed breeding programme calls into question the potential for combining
GM technology and local cotton cultivars to produce new technologies that
offer desired performance across multiple criteria, as well as focusing on the
GM trait rather than the suite of characteristics of the germplasm into
which it is conferred.
The second question concerns the role private ownership played in Bt
cotton’s decline in Burkina Faso. Three generations of backcrossing were
undertaken, which is standard practice in the United States where quality
issues are much less pronounced given the heavy reliance on mechanized
39. Ibid.
10 AFRICAN AFFAIRS
at University of San Francisco on January 13, 2016http://afraf.oxfordjournals.org/Downloaded from

pickers. But in Burkina Faso, where quality concerns are paramount, some
breeders advise a minimum of five generations of backcrossing to ensure
the carry-over of the desired beneficial traits.
40
As a result, the desire for sta-
bility and quality clashed with the desire to get to market, as each back-
crossed generation takes a year of careful breeding and selection. Monsanto
officials acknowledged that they ‘wanted to go faster’, and were confident
that three generations of backcrossing were sufficient to maintain these
quality characteristics.
41
The process of introgression is complex and time-
consuming, and potential conflicts can emerge when the priorities of
private patent holders clash with those of other actors.
A third set of questions revolves around the nature of Burkina Faso’s
pull-back from Bt cotton. In Burkina Faso, the decision to phase out Bt
cotton was made by the cotton companies, not cotton farmers. Burkinabè
cotton companies were frustrated by the declining profits associated with
the poorer lint quality of Bt cultivars. The position of the cotton companies
contrasts with most of the farmers we have spoken to over the past few
years, who tended to be enthusiastic adopters of Bt. Both farmers and
cotton companies benefit from a vibrant and profitable cotton sector; the
cotton price paid to farmers is ultimately a function of the price at which
the cotton company sells it on the world market. In this particular case,
though, the interests of the companies and the farmers diverged: the higher
yield of Bt cotton meant more income for farmers while the lower ginning
ratio and shorter staple length meant less fibre, and of a lower quality, for
cotton companies to sell. The case of Bt cotton in Burkina Faso exposes the
conflicting interests within the cotton value chain, underlining how GM
crops can produce different outcomes for different stakeholders.
It remains to be seen how the news of the phase-out in Burkina Faso will
influence the positions taken by other African countries in their delibera-
tions over whether to adopt Bt cotton. Unless these quality characteristics
are fixed, other Francophone African countries such as Mali, Côte d’Ivoire,
Togo, and Benin are unlikely to adopt Bt cotton given the implications
for their own highly valued reputations. By contrast, some Anglophone
African countries, whose ginning ratios and staple lengths do not achieve
the heights of their Francophone African counterparts, might be less
concerned, and move forward with adoption regardless of this latest
controversy.
Perhaps a more enduring legacy of the Burkina Faso case will be its effect
on the polarized debate over GM crops across Africa. As occurred with the
example of Makhathini, the representation of Burkina Faso’s experiences
with Bt cotton may end up straying far from the reality. Actors on both
40. Interview, cotton agronomy expert.
41. Interview, Monsanto official #1.
BRIEFING 11
at University of San Francisco on January 13, 2016http://afraf.oxfordjournals.org/Downloaded from

sides of this debate will work hard to shape the narrative that emerges. GM
crop opponents are likely to use this case to raise questions about public
trust: will GM crops perform as intended or will they have unknown
impacts and risks? Can the institutions charged with the creation and regu-
lation of GM crops be trusted to ensure the proper development and regu-
lation of these crops? On the other side, supporters are likely to stress the
yield and profit gains achieved in Burkina Faso, asserting that concerns
with germplasm expression are isolated to this particular case and do not
signal a broader issue with GM crops in general.
A number of African countries –including Uganda, Kenya, and Ghana –
are poised to make decisions about whether to adopt Bt cotton in the next
few years. The version of Burkina Faso’s experience with Bt cotton that
filters down to key decision makers will play an important role in deciding
whether these countries move forward with this technology. A key deter-
minant in these debates will be the extent to which African governments
and citizens are reassured that the transplantation of GM traits into their
own cultivars will leave their most valued characteristics unchanged.
12 AFRICAN AFFAIRS
at University of San Francisco on January 13, 2016http://afraf.oxfordjournals.org/Downloaded from