Yale welcomes Christopher Udry back for a lecture on surprising new data about structural transformation

At first, we thought it was a smoking gun of the impact of climate change on African agriculture. And to be sure, agricultural yields and output in sub-Saharan Africa are very strongly associated with weather patterns. But we don't actually see in our data any sustained trend in weather outcomes that can explain this yield trend. We also explored the possibility that the trend was due to expansion of cultivation into more marginal areas, or increases in conflict, or changes in the crop mix. We found that all of these things matter, but they don't explain the overall trend. 

What does matter is labor input. In particular, we found that there has been a sustained decline in both the number of days worked in agriculture and, more importantly, the number of hours per day worked on farms. This fully explains the falling yields that we see.  

What do you think is going on in the results?  

There's a simple story, which we think is true. Essentially, options outside agriculture are improving, so people are doing what you would expect them to do: they are substituting work in agriculture for work in something that's better. Indeed, we find that the yield declines are most rapid in areas that have more active labor markets, that are better connected to growing urban areas, and that are closer to increases in population density. 

Yet all of these factors are things that the traditional literature on structural transformation would say are sources of increased yield in agriculture. That literature says that agricultural productivity drives processes of structural change outside agriculture, or vice versa. But in fact, we're finding the opposite. We see something that looks like an apparent negative spillover from growth outside agriculture leading to declining yields in agriculture.  

But of course, it's misleading to think of this as negative. It's positive – it's a reflection of increasing opportunities outside agriculture.  

How does this deviate from what you had expected going into the research? 

I guess my expectations were twofold. First, I was hoping that processes of technological change – new technologies, new seeds, better availability of markets – would have led farmers to adopt new technologies that increased productivity. But we don't see any evidence of total factor productivity increasing in important ways.   

Second, following arguments made by economists like Ester Boserup and others, I would have expected that growing cities and the growing urban demand for food, combined with transportation costs, would have led to increasing yields in peri-urban areas, as farms near growing cities meet that increased demand, start using factors of production more intensively, and move up the marginal cost curve.  

This actually is what makes me so excited to be a development economist now. The field has grown so tremendously, and you see the melding of really good theory, fieldwork, data, and econometrics. – Christopher Udry

But what we're seeing instead, we believe, is that the increased availability of off-farm employment opportunities – maybe jobs, but also informal self-employment – is taking people at least partially out of agriculture. Essentially, the opportunity cost of labor is rising. 

You’ve worked on agriculture in Africa for decades. How does this study link to your prior research? 

Most of my work has actually focused on the opposite. Rather than looking for commonalities and broad patterns, most of my work has focused on the remarkable heterogeneity that's apparent across sub-Saharan Africa – even in geographically close communities, like cities or farms. I’ve often focused on the many differences that can be found across local economies, and on what we can learn from this variance.  

But this astonishing degree of heterogeneity, we argue, does not necessarily translate into inefficiency or what economists would call a market failure. Merely observing differences doesn't necessarily imply inefficiency – and heterogeneity doesn’t necessarily imply a market failure. A given farmer might be allocating her resources as efficiently as possible and there still might be enormous variation within that same household across different plots on the same farm. 

Can you talk about your efforts to influence policy with your research? 

At the Global Poverty Research Lab, we do a lot of directly policy-relevant work. For example, right now we’re engaged in a large meta-analysis of unconditional cash transfer programs across developing countries. We’re trying to see what we can learn from the aggregated experience of providing cash – mobile money or physical cash – as a means of social protection. We're doing something similar with “graduation programs,” or multi-dimensional anti-poverty interventions. These projects are meant to be directly policy relevant, to try to identify the conditions under which these sorts of programs can have big impacts.  

But closer to my heart, actually, is the work I do that tries to understand why such poverty exists and why it is so resistant to change. Why is it so difficult for individuals and communities to get out of poverty? Those types of questions, I think, require a deeper understanding into people’s economic conditions, their institutional arrangements and social constraints, and the multiple dimensions of people's lives that shape their economic wellbeing. That type of work is fundamentally policy-relevant, because we have to understand the situation in order to make good policy decisions. But the relevance is indirect; it doesn’t say directly to “do A rather than B”; it tells you how to think about making that choice.  

What led you to development economics and your focus on Africa?   

Right after graduating from college, I spent a few years as a secondary school teacher in Ghana in the Peace Corps. Those years really shaped a lot of what I think about and why I do this. When you're 22 years old and you meet these kids – who can speak four languages but are out of school after elementary school, because they can't get to the nearest junior secondary school… You can’t help but think about the waste to the world of the contributions that these people could make if only given the opportunity. Once you meet people like that, it's hard to think of anything else. So that's why my work has focused on Africa, and with a particular focus on the poor and West Africa.

 This actually is what makes me so excited to be a development economist now. The field has grown so tremendously, and you see the melding of really good theory, fieldwork, data, and econometrics. The personal engagement of brilliant researchers, more and more from the region in which they're working, shedding light on why institutions are the way they are, how behavior is shaped, and the constraints that people face. It's really exciting, to see all of that just flowering in the field.