David McKenzie and Michaela Weber, September 2009
Around the world a growing minority of small and medium enterprises are owned by women. Understanding which policies work for getting female-owned enterprises to grow is critical for enabling women to have a bigger impact on private sector development.
Over the past five years the World Bank and IFC have placed more emphasis on programs increasing access to finance and other policies needed for female-owned businesses to grow. To date, a small portfolio of operations is beginning to form, which can provide first indications on the relevance and performance of interventions in this area.
Training and Credit in Uganda
Women in Uganda receive only 9 percent of commercial credit, despite owning 39 percent of registered businesses. IFC put together a comprehensive program for DFCU, a local bank; the program was comprised of a $6 million SME credit line (including $2 million targeted to women entrepreneurs), and of a tailor-made advisory program that combined training for the Bank and for its potential female SME borrowers.
The courses were dedicated to the bank’s potential loan clients, spanned 12 days, and covered banking requirements and processes, financial literacy, separation of business and personal accounts, and provision of opportunities for business networking and peer mentoring. 51 women were trained in a first group in 2007. At the same time, DFCU drew down the women entrepreneur specific credit line rapidly (within 3 months instead of 2 years) and reported providing loans to 297 women entrepreneurs, with an outstanding default rate of less than 0.5%.
To understand the role that the training provision played within this context, evaluations can play a helpful role. Two evaluation strategies were pursued. In the first, undertaken with financing from the Gender Action Plan, a randomized experiment was designed for a 2009 training group; 75 women were selected from a list of 200 DFCU clients to whom training would be offered. However, only half of those offered the trained attended, and other firms from the experimental control group ended up being invited to make up the target number to be trained. Given the small sample size and other project limitations, this led to us having to abandon the randomized experiment.
However, the experiment is still useful for understanding the take-up decision among clients offered the training. Contrary to what one might expect, the main determinant of the take-up decision was the firm manager’s existing level of formal education: 77 percent of those attending the training had some university education.
The second evaluation methodology implemented was a difference-in-differences methodology, comparing the change in firm outcomes over the 2007 to mid-2009 period for firms in the first training group in 2007 (treatment group) to that in firms whose owners started training in mid-2009 (control group).
The 2007 training was implemented without an impact evaluation design or baseline survey. We therefore had to design an ex-post evaluation which attempted to find and contact the firm owners from the 2007 training group, and ask both them and the 2009 group about the conditions of their firm today and in 2007, and about their training experience.
There were some differences between the two years’ training groups. Those trained in 2007 had existed longer, had several more employees in 2007, and were somewhat less in traditionally female sectors of the economy. Survey tracking rates were lower for the 2007 training group, perhaps due to less successful businesses closing or being less willing to participate in the survey.
Self-assessed results are high
- 72% of the 2007 training group said the training has significantly improved the performance of the business and reports from those who agreed to participate in a focus group following the survey were positive.
Is training the key to success?
- Training did lead firms to be more likely to hold separate business and personal accounts, to specify a salary they draw each month, and to have budgets they monitor.
- There were no changes in other business practices such as willingness to delegate business tasks, payment of taxes, and how they employ relatives.
- Training is not associated with greater use of professional accountants, or more access to finance.
- Those trained in 2007 did not experience significantly higher sales or employment growth over 2007-mid 2009 than those trained in 2009.
The main goal of the training was to create a pool of women entrepreneurs who would then qualify for commercial financial services to achieve their maximum growth potential; the training was intended to have most benefit when coupled with access to commercial financial services. In our sample only half the firms trained in 2007 had bank loans in 2009. It may be that the training has beneficial effects on firm growth for this subgroup, but we are unable to detect this with the sample size which we have.
- Targeting. Best uptake seems to come from women with some university training; featuring this into marketing strategies may save cost and effort. However, reaching out to less educated women with high potential businesses may require a different approach than through the commercial banks.
- Continuously refine content based on entrepreneur segment and beneficiaries feedback. The material on business planning and accounting and financial management practices, including separating business and personal accounts was practical and useful, other content was thought too conceptual. The material is specifically targeted at a narrow range of relatively educated women with perceived ability to grow their businesses, and achieving the right level for content to meet their needs and time constraints appears to require further refinement and testing of content and delivery mechanisms.
- Further impact studies are needed, to test whether training matters for those receiving credit. Such a design should be implemented ex ante. Budget constraints and the pilot nature of the project limited on-the-ground-presence, and as a result, sample size and evaluation methodology. Randomization is always difficult to explain to development partners, in particular to a commercially operating partner. More face-to face contact with all partners can help making it happen right.