Business Training and Mentoring: Experimental Evidence from Women-Owned Microenterprises in Ethiopia

Very excited to share results from our recent publication in Economic Development and Cultural Change (University of Chicago Press), titled “Business Training and Mentoring: Experimental Evidence from Women-Owned Microenterprises in Ethiopia.” Adding the abstract below:

“Recent research shows that microenterprises in developing countries are constrained by their managerial capacity, especially in the areas of marketing, record keeping, financial planning, and stock control. In a stratified randomized controlled trial, experienced businesswomen in Ethiopia were given a formal business training that addressed these constraints. A second-stage mentoring component in which a random selection of female mentees within the social and business network of the trainees from the first-stage business training received customized mentoring from these “trained mentors.” Pooled results using three rounds of post-training surveys carried out over three years show that business training causes profit and sales to improve by 0.21 standard deviation, while business practices improve by 0.13 standard deviation. The overall impact of mentoring is muted—strong impacts are observed on the adoption of business practices among mentees, but there is no statistically significant impact on profits.”

Working on this paper was a hugely rewarding experience. I received close guidance from the great Markus Goldstein (Director of the Africa Gender Innovation Lab at the World Bank), who and his group are doing cutting-edge empirical research on women’s empowerment in Africa.

About the study – it highlights the potential positive effects of a well-targeted and well-implemented business training program. One novel aspect of the study is the evaluation of a mentoring intervention, where trained mentors were encouraged to teach their mentees what they had learned in the training program. While we don’t see a statistically significant impact of the mentoring component, we show some neat (I think!) evidence on how the improvement in the business performance of the mentees is in line with what their higher adoption of business practices would predict – suggesting that business practices (such as record-keeping, marketing, financial planning, etc.) indeed matter for profits. I believe that some of our results indicate promising possibilities of a well-designed (and incentivized!) mentoring programs.

The ungated/free manuscript is available in my personal website.

Latest Food Security Situation in Bangladesh (data collection ongoing)

Since the beginning of the COVID pandemic, IFPRI has been collecting data to assess the effect of the pandemic on Bangladeshi households, and to chart the recovery process. We have been interviewing households (over the phone) from all districts of Bangladesh by using two national samples of Bangladesh – one for rural households and another for urban households. The pre-pandemic in-person collected data of these households was collected in 2019, while after the pandemic started, we have been carrying out phone surveys in June 2019, January of 2021 and finally, now (September-October) in 2021. The data collection of the current survey is still going on – we have finished about 75% of the sample. But since the incoming data pipeline is automated (thanks to the prodigious coding skills of the IFPRI-Bangladesh research team!), we can start visualizing some of the data as it comes in.

Food insecurity is an extremely important measure of household welfare. We have been asking eight questions on household food insecurity which were designed to elicit information on food behaviors and actions taken by households when the resources needed to access food are constrained. For example, these questions asked whether the household worried that it would not have enough food to eat; skipped meals because the household lacked money or other resources to access food; been hungry but gone without eating; or gone without food for an entire day. We asked these questions using a 30-day recall period. Responses to these questions allowed us to construct the four categories based on the Food and Agriculture Organization’s (FAO) Food Insecurity Experience Scale (FIES): (a) food secure—answer yes to none of the eight questions; (b) mild food insecurity—answered yes to 1, 2, or 3 questions, indicating some element of food insecurity; (c) moderate food insecurity—answered yes to 4, 5, or 6 questions; and (d) severe food insecurity—answered yes to 7 or 8 questions.

Anyway, here is the exciting news – despite the observed increase in food insecurity in the early months of the pandemic (last year), the proportion of households reporting moderate or severe food insecurity has NOT increased in September 2021, despite the recent COVID-related surges and the national lockdown. It’s possible that things were worse a few months ago (during the strict lockdown period), but in that case, a rebound seems to be happening. The recovery of urban households seems to be a bit slower than rural households. These are early indications of resilience among Bangladeshi households.

Although the incidence of moderate and severe food insecurity has gone back to pre-pandemic levels, it is important to note that overall food insecurity of Bangladesh has increased compared to pre-pandemic times. This is because mild food insecurity has increased substantially compared to 2019. In other words, a large number of households which used to be food secure are now experiencing mild food insecurity. This shows up clearly when you tabulate the incidence of any food insecurity reported by households (i.e., mild, moderate or severe food insecurity).

For your kind information, this is based on incoming data from an incomplete sample; so, the estimates will likely change when the final sample is available. We are hoping to have the report ready by the end of October or in early November. But so far, the interim results are quite heartening to see.

If you are looking for our report based on the January 2021 (and earlier) data, please click here.

Near real-time data from field-level ag extension officers to aid high-level government decision-making

I am quite excited about some upcoming ‘action research’ projects at IFPRI-Bangladesh. One such pilot project (with support from CIMMYT and 1-CG COVID hub funding) involves developing a tool which will allow field-level officials of the Department of Agriculture Extension (DAE of the Ministry of Agriculture) to report the most pressing problems farmers are facing all around Bangladesh. These could be, for example, higher agricultural input or lower output prices, or the lack of availability of certain types of input (fertilizer, pesticide, etc.), as well as other environmental and market-related problems (say, lack of buyers during the COVID lockdown) faced by farmers. The idea is to produce a near real-time, nicely visualized data for top-level policymakers, who don’t otherwise have this kind of collated information available to them. Geographical hotspots, where potential adverse events are taking place, can also be identified faster.

This tool could potentially be very useful for the government . It’s also likely that this type of data will be trusted more by top-level decision-makers, since these would essentially be their own data (reported by field-level government officials). Geographical hotspots, where adverse events are taking place, can also be identified faster. For example, locations where fertilizer supply has taken a dip for some reason or some other type of input prices has suddenly increased can be identified sooner, which could lead to a proper investigation and a faster resolution of such issues.

If the piloting goes well, we plan to implement this work nationally and afterwards, (if there is enough interest) with some other government agencies. For example, the Ministry of Women’s and Children’s’ Affairs could have data from sub-district level health complexes about the height/weight of new-borns. If there is a location, where suddenly the Weight-for-Height-Z score (which is a measure of thinness/wasting of children) has decreased due to a possible aggregate shock in the local economy, it can be identified sooner.

I think these tools could help with faster and more efficient decision-making from the top level of the government, and also identify geographical hotspots where possible adverse events are happening. This pilot could make a difference and I am quite excited about working in it.