The yawning evidence gap in agriculture

|

This page in:

co-authored with Michael O'Sullivan

If you think that agriculture is less than interesting, then perhaps issues of gender in agriculture make you even more likely to yawn. Indeed, our friends over at the Campaign for Boring Development did a nice job of summarizing a new report that a team of us at the World Bank did with the ONE Campaign on gender and agriculture in Africa. But hopefully this report doesn’t put you to sleep.

We're not going to summarize the report, but there are two points that bear thinking about from the perspective over here at Development Impact.   
 
First, the report combines both inferential, old-school research with the existing impact evaluation research. There has been another spate of "impact evaluation isn't the answer to all development problems" discussion, including a thought-provoking recent post by Jeff Hammer. One of Jeff's points is, rightfully, that we can't answer some questions with impact evaluation. In the case of this agriculture report, that particular set of questions centers around the relative productivity of male and female managed farms. First, how much less do women produce per hectare (answer: a lot). Second, why do they produce less? Using work coming out of an LSMS-ISA project, the report provides a summary of decompositions of the productivity differences in six African countries (although Nigeria really should count twice). This work shows that levels of factors of production matter -- for example, women in Niger use less male household labor on their farms than men do. But it also shows that returns matter -- for example, those Nigerien women also get lower returns to applying a kilogram of fertilizer than male farm managers do. And by doing this in a decomposition framework, we can get some sense of the main factors behind the gap. So this helps isolate areas for policy. But it still comes up a bit short. Take the case of lower returns to fertilizer: why do female farm managers have lower returns to fertilizer than males? Is it an information problem (e.g., when to apply, how much to use)?  Is it an aggregate liquidity problem? Is it an inter-temporal liquidity problem? Is it that women are more likely to use adulterated fertilizer (they have smaller fields and thus are more likely to use a repackaged bag of fertilizer than the off-the-boat 50 kg bags)? All of these have significantly different policy responses, and now we are in the land where a judicious mix of additional research (did someone say qualitative work?) and policy experimentation is in order.
 
The second thing that struck us as we worked on this report is how very little impact evaluation work there is in the agricultural policy space. If you look at policies for enterprise development, heck, if you just focus on business training evaluations, there is quite a bit. But on agriculture, without any attention to even gender disaggregated effects, there is precious little. And, of course, there are even fewer papers that have anything to say about gender.  The good news is that a lot of the impact evaluation papers we covered in the report were recent -- indeed, a bunch haven't even been published yet, so maybe the tide is turning. (In addition, one interesting side note is that interventions from other development areas might also provide some insight -- for example, might what worked for getting legitimate malaria drugs out there also help get better quality fertilizer out there -- that is, if quality is the problem...).
 
Taken together, these two pieces -- more insights into what is likely behind these gaps in productivity, and some (limited) insights into what interventions might help – opens up a research agenda to tackle the big issues facing women in agriculture. 
 

Authors

Markus Goldstein

Lead Economist, Africa Gender Innovation Lab and Chief Economists Office

Agnes Quisumbing
April 09, 2014

I agree with Markus that rigorous impact evaluations addressing issues of gender in agriculture are long overdue, and that there are gains to using both quantitative and qualitative methods in impact evaluation, particularly because gender norms are so context-specific. IFPRI and ILRI have just released a series of notes summarizing results from the initial (hopefully) phase of the Gender, Agriculture, and Assets Project (GAAP). Full papers will be on the IFPRI website, but the summaries are available here: http://www.ifpri.org/publication/learning-eight-agricultural-developmen…