I recently came across a paper  by Kelsey Jack which is a white paper for the J-PAL and CEGA Agricultural Technology Adoption Initiative (ATAI). This paper systematically explores the barriers to technology adoption that come from market inefficiencies, what we know about these, and what research is going on (under ATAI) to fill these gaps.
The neat thing about this paper is its systematical-ness. By working through the full range of market inefficiencies, what Kelsey is doing is of immense use for policy makers and researchers. Through a fairly tight economic lens, we can see what the problems are and what we need to know more about in order to address them. It gives us some sense of what is important, and what is unexplored – taking a broad area of agricultural policy as the framing motivation (as opposed to an individual question of theory or a super cute experimental design). My first thought was that we need more of these for other areas as well. This does not look at all easy to do, but this is a serious public good.
So let’s take a look at how things stack up. For each area, there is a description of what the manifestations of inefficiency are and also what has been done in terms of impact evaluations and which ones are going on now. Overall, the score on the evidence for policy side is pretty dismal – there isn’t much out there -- the references to perhaps (or not) generalizable lessons from health interventions for example make this painfully obvious. There is some hope in the ongoing work, but still some gaping holes. By topic, it looks very roughly like this:
- Externalities: little to no existing evidence, some ongoing work on payments for adopting environmentally friendly agricultural practices.
- Input and output market inefficiencies: bits and bobs (some on fertilizer) have been done and some pieces of interesting pieces of ongoing work on contracting, insurance, extension, and subsidy distribution.
- Land market inefficiencies: two done, nothing in the pipeline.
- Labor market inefficiencies: nada, bupkis, zilch.
- Credit market inefficiencies: Some work has been done, some of which is rural/agricultural credit and some not, with interesting results on both commitment and lender technology sides, and a couple of ongoing projects looking at repayment periods and commitment.
- Risk market inefficiencies: some existing work which shows (perhaps) puzzling low rates of take up due to trust and/or liability issues, ongoing work which allows farmers to experiment with small amounts of insurance.
- Informational inefficiencies: some work out there on extension, and a fair number of ongoing evaluations of information technologies and farmer-to-farmer information transfers/social networks.
Kelsey also raises a number of interesting over-arching issues. First, given that multiple market imperfections might bind at once, it is worth thinking about evaluating programs that target multiple constraints, as well as individual ones, preferably in the same setting. This is important – while composite programs might work well, it would be good to know which components are absolutely necessary. Second, she draws in results and theory from behavioral economics which, as we saw in the fertilizer nudge paper  by Duflo and others, can be important. And she points to some areas where it might matter and further work would be a good investment. Finally, she points out in a number of places where these constraints might bind on women differentially and this too would be a fruitful area of additional work (and a blog post to come later).
Kelsey’s picture of ongoing work is incomplete – for example, I know of at least four impact evaluations of land related impact evaluations going on at the Bank/MCC which will go some (small) way to reducing the knowledge gap on land. But this is to be expected, as there is no master global database of ongoing impact evaluations. But it seems to me that this paper provides a powerful example and argument for greater coordination as it identifies the important questions/issues and knowledge gaps and provides a coherent, big picture, policy-relevant impact evaluation agenda. Given that this falls in a hazy nexus between policy and research, who should do this for other areas?