BERF Scoping Study: Gender Differences in Enterprise Surveys, April 2017

The aim of this scoping study is to support efforts to enhance overall business performance in low income and emerging countries through contributing to the efficacy of business environment reforms. In the vast majority of countries, women’s businesses are fewer in number, smaller in size and concentrated in less capital and asset intensive sectors, compared to men’s businesses. In some regions, they are also weaker in certain dimensions of performance.

This study explores the extent to which the relative weakness of women’s businesses is attributable to the Business Environment, and aims to identify the potential for business environment reforms to assist the growth of women’s businesses. It lays the ground for a more rigorous assessment of these two issues, using the largest available database on this topic: the global World Bank Enterprise Survey.

The study assesses the suitability of the Enterprise Survey datasets for sex disaggregated analysis, sets out the key gender-related findings that have emerged from previous research, and explores possible hypotheses and methodologies for understanding the different impacts that the Business Environment may have on the prevalence and performance of women’s and men’s businesses. Newly available repeat (panel)  datasets that track the situation of individual enterprises over time appear to have great promise; but they have limitations, particularly in respect of data on women’s businesses.

Taking a broad view of what constitutes the Business Environment, including aspects of the general legal framework of particular concern to women, the study puts forward four hypotheses to be tested by further detailed analysis using survey data, complemented by other relevant data, using appropriate statistical methods. The hypotheses are a package of interrelated ideas, covering equity as well as efficiency aspects.

Our first hypothesis is that certain elements of the Business Environment are more constraining to women’s than to men’s enterprises, although the effects will differ across countries.

Our second hypothesis is that individual Business Environment Reforms have had different effects on the performance of men’s and women’s businesses, such that, as before, Business Environment Reform affects enterprises by gender in different ways in different countries.

Our third hypothesis is that certain Business Environment and related reforms have induced larger improvements in the performance of newly- (or soon to be-) formalised enterprises relative to more established businesses.

Finally, given that female employment rates seem to be significantly higher among women’s than men’s businesses, the fourth proposed hypothesis that we wish to test is that the gender gap in employment (captured by the difference in share of female employees in the enterprise workforce) is related to the gender of the enterprise owner or manager.

Three levels of analysis can be considered to examine these hypotheses: cross-country analyses using aggregate country-level data; enterprise level analysis both within and across countries; and cohort-level analysis, grouping companies by categories (sector, gender in size and other possible categories).  Our assessment is that cohort-level analysis is likely to be most fruitful.

We anticipate that analysis to test one or more of these hypotheses will produce findings that can increase the gender sensitivity of DFID and partners’ policy and programming on Business Environment Reform in ways that enhance both the prevalence and performance of women’s businesses.