Impact of Microcredit in Rural Areas of Morocco: Evidence From a Randomized Evaluation

Publication Details

Massachusetts Institute of Technology (MIT), March 2011, Working Paper 6659. Available from:

Link to Source
Bruno Crepon, Florencia Devoto, Esther Duflo, William Pariente
Institutional affiliations
None specified
Grant-holding institution
None specified
Middle East and North Africa
Finance, Private Sector Development
Microfinance, Small Scale Enterprise
Microfinance, Small Scale Enterprise
Equity Focus
Evaluation design
Propensity Score Matching (PSM), Randomised Control Trials (RCT)
Published Working Paper


This paper examines the effects of a microcredit programme implemented in rural areas of Morocco. With approximately 130 million beneficiaries worldwide, microcredit programmes have received much attention from the academic and policy arenas. However, little is known about the impact of such programmes on general outcomes such as poverty, employment or gender inequality. The present paper extends the existing literature on the impact of microfinance initiatives by focusing on the case of Al Amana, a leading microcredit institution operating with an overall portfolio of 2,736 million Moroccan dirhams (MAD) and serving over 400,000 clients (as of December 2009). In 2006 Al Amana expanded its activities to rural and peri-urban areas with the opening of 60 new branches offering both group-liability and individual loans ranging from 1,000 MAD to 48,000 MAD.
The authors assess the effects of Al Amana's expansion using a matched-pair randomised controlled design. Before the opening of the new branches, the researchers collected data in at least six villages located within the intervention area of each branch. Villages from similar intervention areas were then matched (80 pairs in total) on the basis of characteristics such as the number of households, infrastructure, type of agriculture and income-generating activities. Within each pair, one village was randomly assigned to be covered by the programme and the other served as the control. Approximately 5,000 households were surveyed at baseline and again two years after opening the branches. Surveys collected information on household assets, socioeconomic characteristics, credits, consumption, agricultural production and livestock activities. The analysis measures intention-to-treat estimates and uses a specification that includes a dummy variable indicating whether a household is in a treatment area and variables indicating outcome levels.

Main findings

Two years after opening the new branches, programme uptake remained low, with only 16 percent of the 5,000 households surveyed by the research team having received a credit from Al Amana. Overall, the intention-to-treat estimates show that programme availability in treatment areas had a significant and positive impact on sales (by 3,305 MAD), expenses (by 14 percent), in-kind savings (by 11 percent) and self-consumption (by 12 percent). However, these impacts were driven mainly by agricultural businesses, and no effects were found for non-agricultural businesses. The researchers also observed that households living in treatment areas diversified their livestock production to include a broader range of animals and products.
Furthermore, households in treatment villages tended to work less outside the village (wage earnings dropping by 6 percent), suggesting that household members dedicated more time to their own income-generating activities. In addition, treatment businesses hired employees for a longer period than control businesses (that is, 19 more days per year, corresponding to a 52-percent increase).
The study does not report any significant effect on general outcomes such as poverty or household consumption. Moreover, subgroup analyses reveal that only part of the population benefited from the introduction of the programme. Indeed, only households with pre-existing businesses reported positive effects—a 15-percent increase in sales, a 12-percent increase in business expenses and a 7-percent increase in savings—whereas no significant effect was observed among households starting a new business.

Scroll to Top