Subscribe to E-Bulletin Donate to CEGA

Microenterprise Growth and the Flypaper Effect

Development Challenge

There has long been debate around giving unrestricted cash to poor households.[1]  While some see it as a tool to empower the poor, others fear it can lead to waste and misspending.  This fear percolates in the microfinance sector, where many organizations restrict loans to ensure spending on business purposes.[2]  While traditional economic theory suggests freedom of cash flow enables the entrepreneur to disperse capital where it most needed, other literature hypothesizes a “flypaper effect”; money sticks where it hits.  As microenterprise grant programs expand across the developing world, understanding differing efficacy between restrictive in-kind grants or more liberal cash grants could promote business growth.   


The study follows small microenterprises in Accra, Ghana, and the nearby industrial city of Tema.  Both locations have strong industry relative to the rest of the country, and are home to many small firms.  Figures suggest as many as half of those employed in Ghana are self-employed.[3] Ghana has a documented history of female entrepreneurship,[4] and Ghanaian women participate in the labor market at an equal rate as their male counterparts. This gender balance makes for an apt context for studying the effects of grant type by gender.

Evaluation Strategy

To test differences between cash and in-kind grants, researchers randomly selected 70 enumeration areas in Accra and 30 in Tema.  In each area, households were screened for individuals 20 to 55 years of age, who were self-employed and working over 30 hours per week on their business.  To ensure the grants could have a substantial effect, the businesses had to be small; selections were restricted to enterprises with no paid employees and no motor vehicle.  After two pre-treatment baseline survey rounds, 793 firms–479 led by females and 314 by males—were selected to take part in the study.  The baseline defined several strata that would later be used to analyze the data: participant businesses were identified as working in a male-dominated industry (eg. carpentry), a female-dominated industry (eg. beauty) or mixed industries (eg. insurance), and whether the businesses were initially profitable or subsistence (low-profit).  Participants were randomized into one control and two treatment arms as follows:

Group Grant Number of Firms
Control No Compensation 396
Treatment 1 Cash Grant: 150 Ghanian cedis (~120 USD)  198
Treatment 2

In-Kind Grant: Equipment, materials (~120 USD)


The grants were substantial: 150 cedis, at the time of the study, were equivalent to approximately two months of profit for the average enterprise. Data collection consisted of quarterly surveys conducted in May, August and November 2009, and February 2010.  Of the 793 firms initially enrolled, 730 stayed in the program and completed the final survey.  The surveys measured profit margins and spending behaviors.  

Results and Policy Implications

Study results support the theorized flypaper effect among Ghanaian microenterprises.  While cash had a positive effect overall, the in-kind treatment were substantially more effective across all firms.  When disaggregated, however, the data show more nuanced results. 

Women with high initial profit margins were the primary driver of an increase in profits for the in-kind treatment group.  These differences were consistent over time, including in the three-year post-treatment follow up.  Women who owned businesses with higher profit margins before the intervention tended to have businesses of larger scale, and they tended to be better educated and to come from richer households.  Men who received cash grants drove up the overall monthly income increases, but this increase was offset by the negative net gains of women who received cash.

One possible explanation for larger effects of the in-kind grants is that these grants function as commitment devices and prevent impulse spending. Alternatively, the in-kind grant can also guard against “social solidarity taxes”, or external pressures to share money among family members.  Estimating these effects from the data, researchers found statistically insignificant, but suggestive, results favoring the commitment device hypothesis.  More research is needed to confirm this effect. 

The results from this study build upon a body of evidence on the effectiveness of microfinance on small businesses.  Findings suggest that cash treatments have a small effect for some subgroups, while in-kind grants have larger overall effects.  Moreover, women drive effects with successful businesses at onset.  This is an especially important finding for grantmakers who target subsistence-level female firms, where grant effects appear weakened. Overall, the study shows the average microenterprise gains from in-kind grants and suggests that microfinance programs focusing primarily on women may be ignoring a large group of enterprises in need of more capital.  



[1] Show Them the Money. (2014, November 23). Retrieved November 23, 2014, from

[2] Karlan, D. & Zimmerman, J. (2012). List randomization for sensitive behavior: an application for measuring use of loan proceeds. J Dev Econ. 98 (1), 71-75.  

[3] Ghanian Labor and Policy Research Institute. (2009). The Trade Union Congress of Ghana.  Retrieved November 22, 2104, from

[4] Hill, P. (1984).  Indigenous trade and marketplaces in Ghana.  Department of History, University of Jos Jos, Nigeria.

Photo: Arne Hoel/The World Bank. Woman in small shop in Ghana.