Finding Missing Markets (and a disturbing epilogue): Evidence from an Export Crop Adoption and Marketing Intervention in Kenya

Publication Details

American Journal of Agricultural Economics, November 2009, v. 91, iss. 4, pp. 973-90. Available From:

Link to Source
Nava Asharf, Xavier Gine and Dean Karlan
Institutional affiliations
None specified
Grant-holding institution
None specified
Sub-Saharan Africa (includes East and West Africa)
Agriculture and Rural Development, Finance
Agricultural Credit, Agro-Industry & Marketing, Microfinance
Agricultural Credit, Agro-Industry & Marketing, Microfinance
Equity Focus
None specified
Evaluation design
Randomised Control Trials (RCT)
Journal Article


This study evaluates the effects of an intervention aimed at improving access of Kenyan farmers to horticultural export markets through the provision of financial and technical services. Smallholder farmers in Kenya face several constraints inhibiting effective access to export markets. Limited knowledge of pricing, excessive presence of middlemen and a lack of access to credit significantly reduce export opportunities. To overcome these barriers, the Kenyan nongovernmental organisation (NGO) Drum-Net designed in 2004 a micro-credit programme in which the NGO acts as a neutral third party linking small farmers to commercial banks, crop exporters and transportation facilitators. The intervention starts with a 4-week course on good agricultural practices. Drum-Net then organises participating farmers into groups of five members jointly liable on all individual credits. In addition, the NGO negotiates prices with exporters and supervises produce transportation. Exporters directly pay Drum-Net, which then transfers the funds to farmers’ individual bank accounts, deducting any pending loan repayments.
The study assesses the improvement of farmers’ production and welfare using a cluster randomised controlled trial (C-RCT) procedure. Using administrative data from the Ministry of Agriculture, the authors selected all horticulture self-help groups (SHGs) active in the study area (Gichugu). They randomly assigned those 36 SHGs to three experimental groups of equal size: (1) a ‘treatment credit’ group receiving all Drum-Net services, (2) a ‘treatment no credit’ group receiving all services except credit, and (3) a ‘control group’ receiving no services at all. In April 2004, the evaluation team conducted a baseline survey with the 726 farmers constituting the 36 SHGs and a follow-up survey in May 2005 with these same farmers and 391 new members. The researchers’ measure intent-to-treat estimates, controlling for a set of variables related to members, land, production and the use of inputs.

Main findings

Impact estimates suggest that the intervention had a significant impact on the production of export crops (French beans and baby corn), with treated farmers being 19.2 percentage points more likely to dedicate a portion of their land to these crops than control farmers. The programme also appears to have increased the production of baby corn in absolute terms, whereas the production of French beans remained unaffected. Although the intervention did not significantly raise the incomes of treated farmers as a whole, the study mentions that first-time growers of export crops did benefit from the programme, which is reflected by an income gain of 32 per cent. The authors do not report any impact differential between the two treatment groups, suggesting that the credit component of the intervention did not affect any of the outcomes of interest.
In light of the positive impact the intervention had on farmers who were not previously producing for the export market, the researchers recommend that such programmes should mainly focus on expanding outreach to new farmers rather than providing support to farmers already accessing export markets. They also call for further research to investigate the role of credit access in such interventions. The study concludes with an epilogue explaining that in mid-2006, export companies refused to continue buying the crops sold by Drum-Net because none of the affiliated farmers had obtained the EurepGAP certification, mandatory to access European markets. Consequently, participating farmers could not repay their loans, and Drum-Net had to cease its activities, leading most participants to favour crops intended to local rather than export markets.

Scroll to Top