The impact of microcredit has become a highly contested question. Lauded as a panacea for poverty and gender inequality, microcredit has conversely been criticized for deepening both. This paper (unfortunately, not available on open access):

• reports the findings of a meta-analysis to assess the impact of microcredit on women’s control over household decisions
• illustrates the difficulties of conducting meta-analysis when faced with a diverse evidence base

There appears to be a gap between the often optimistic (societal) belief in the capacity of microcredit to mitigate the position of women in decision-making processes within the household on the one hand, and the empirical evidence base on the other hand, a gap which this meta-analysis should not be thought to have bridged.

The paper discusses complexities and limitations of interest to those working on microcredit as well as those using meta-analysis.

Microcredit

The authors identified “control over household spending” as a measure of women’s empowerment in developing countries. Through this review, they aimed to assess the impact of microcredit by judging its effect on this key dimension. However, the studies they identified and analysed covered an extremely wide range of methodologies, domains and measures. Most significantly, definitions and concepts of “empowerment” varied widely.

In the course of the meta-analysis, we came to conclusions that there were significant risks of over-interpretation when studies are highly heterogeneous, and there appeared to be publication bias.

To the limited extent that conclusions about microcredit were possible, the authors note that:

• the effect of microcredit on women’s control over household spending is weak
• women’s control over household resources may constitute an important intermediary dimension in processes of women’s empowerment, and/or a proxy for empowerment, but cannot be considered empowerment itself

Meta-analysis

Since the late 1990s, systematic reviews and meta-analyses have risen in popularity in international development to provide evidence on ‘what works’. However, as this paper demonstrates, meta-analysis relies on a minimal degree of comparability among studies.

Mixed results related to women’s empowerment can partly be explained by the heterogeneity of microcredit interventions, contexts and target groups and in large part due to differences in conceptualisation and measurement of empowerment. This paper extracts quantitative effect size information on variables that relate to women’s control over household spending without critiquing the construction of the metric of control and attempts a meta-analysis of 29 studies. 

Among other limitations, the authors note that:

• the studies examined were highly heterogeneous in terms of treatments, outcomes and research designs
• tools for assessing the quality of studies relied as heavily on ‘expert opinions’ as they did on ‘expert systems’
• the risk of bias assessment indicated that most of the included studies displayed serious methodological weaknesses, an aspect which inevitably compromised the synthesis phase
• studies were more likely to be extracted if they were peer-reviewed

Method

This meta-analysis review:

• adapted standard systematic reviews procedures
• developed a protocol setting out inclusion and exclusion criteria, a search process and a synthesis methodology
• featured studies with participants from poor, lower and upper-middle income countries as defined by the World Bank.
• included only studies which examined microcredit and women’s control over household spending, or relevant proxies, as the main outcome variables
• examined quantitative studies which had statistical controls and/or a comparison group