As part of our interview series, we ask renowned experts in the field about the future of research in development economics, and for their advice for young researchers. For this interview, we got the opportunity to talk to Prof. de Roux.
Nicolás de Roux is an Associate Professor at the Economics Department of Universidad de Los Andes Colombia and Visiting Associate Professor at the School of International and Public Affairs of Columbia University.
ETRM: Your research covers different topics in development economics like credit, labor markets, education or trade. What drives your interest in these areas?
Nicolás: If I had to put a label on what I do, I would say I study production in developing countries and the barriers producers in these countries face. You can think of a big manufacturing plant, an agricultural producer, or a small enterprise owner looking for a loan. Or you can think from the perspective of a bank that’s unwilling to lend. The barriers they face can be related, for example, to problems in labor markets, access to credit, lack of access to foreign markets, or to what I presented today, small farms having to sell their land to smooth consumption after a weather shock.
ETRM: When we interviewed Eric Verhoogen, who is one of your co-authors, he said that the topic of firms in developing countries is not well studied yet.
Nicolás: I think the way to really improve people’s living conditions is to provide jobs. Those who offer stable employment are mainly firms, especially larger ones, so it’s important to think about how to help those firms grow. That’s why I believe firms in developing countries are so important.
ETRM: On the topic of firms and production in developing countries, what do you think is the most exciting frontier of development right now?
Nicolás: It depends on whether you want to talk about the topics or the methods. I’m going to focus more on the methods. There was a time when regression estimates weren’t identified because there was no credible source of variation. Then came the Credibility Revolution and RCTs. Nowadays, good identification is still important, but we’re using these identified parameters to feed models that allow us to answer bigger questions. What I presented today is a good example: we have reduced-form effects of weather shocks on the farm size distribution, and we use those facts to estimate an empirical model. The intersection of modeling with well-identified parameters from credible exogenous variation is becoming increasingly important in development, and I find that very exciting.
The other thing I want to mention in terms of methods is that we’re making a big push toward working with institutions, first, to get them to run experiments. For example, two weeks ago at PacDev (the Pacific Conference for Development Economics), Karthik Muralidharan talked about working with the government to increase state capacity. I feel that development is increasingly focused on finding ways to engage with the institutions that are actually implementing policies and thinking about how to help them do it better.
When working with institutions, you start using a lot of administrative data, which allows you to answer questions that couldn’t be answered in the past. A lot of development research is beginning to rely on data from administrative sources, where the information comes directly from the operations of the institutions implementing the policies. I think that’s a promising avenue for research in development.
ETRM: How do you stay updated with all the various topics of research that you do, and how do you identify the questions that are worth pursuing?
Nicolás: To stay up to date, I try to keep reading papers in econometric methods or in other topics that interest me. Additionally, something I highly recommend for people working in development is to sign up for the VDEV/CEPR/BREAD Seminars, which I’m now co-organizing. Let me tell you the story. I did my PhD at Columbia University and then became an Assistant Professor at Universidad de los Andes in Colombia. The one thing I missed the most when I returned to Bogotá was the level of the presentations. When I organized the seminar at Los Andes, it was very hard to get top-quality speakers to come, because Bogotá is far away.
When the pandemic started, some professors in Europe launched a virtual seminar, now called VDEV/CEPR/BREAD, that brings in a superstar in development every two weeks to give a talk. This year, for example, we saw Clément Imbert present the other week, and next week we have Ach Adhvaryu, followed by Joe Kaboski, Eric Verhoogen, and so on. So even if you’re not in the US or at a top university in Europe, you’re still exposed to very high-quality talks every two weeks. It’s a one-hour seminar, and the speaker usually stays for an extra half hour in a Zoom room to chat with attendees.
Now, on to your question about generating ideas. I usually get ideas while I’m reading. Whenever something comes to mind, I write it down on a Post-it, and eventually I type it up. Once in a while, I go back and review those notes. Then I’ll talk to a colleague and share an idea, if they’re interested, we start working together. Sometimes, colleagues also invite me to join projects they’re working on.
I would say, though, that we have this scientific ideal: you come up with an idea at your desk, then think of a context where you can test it. You find funding, run the test, and write a paper. However, many very good papers that get published don’t happen that way. Often, it’s just about an opportunity; you meet a partner, an NGO, or gain access to a dataset. So, one piece of advice I give to students is: find an opportunity to work in a specific context. Get to know that context, talk to people, look for datasets, and think about potential ideas within that smaller setting. It’s within the context that it becomes easier to come up with ideas that can turn into a paper. Because it’s only much later in your career that you can say, “I have this idea, and I’ll find the resources to test it in a particular setting.”
ETRM: Thanks for this advice. To finish this interview, we want to know how you see your research influencing policy in Colombia?
Nicolás: I collaborate with multiple institutions and share my research findings with them. For example, I have a paper on loan repayment where we tested different behavioral motives in text messages to encourage repayment. We found that messages based on social norms led to higher repayment than messages with other motives. We presented the results to the bank, and they agreed it made sense to have a behavioral team focused on these issues. That team now exists, in part, because of the research we did. This kind of impact can also come from starting conversations with the key actors in the sectors your work speaks to.
Also, sometimes policymakers reach out to ask about your research. As you progress in your career, they begin to ask you to present your papers and think about policy recommendations. For example, I currently have a draft paper that studies the effect of the collapse of trade with Venezuela on Colombian manufacturers. Venezuela was Colombia’s second-largest export partner in 2009. In 2013, Venezuela’s GDP per capita was twice that of Colombia. But between 2013 and 2018, Venezuela’s GDP per capita fell by a factor of three. As a result, Colombia’s exports to Venezuela dropped from around 13–14% in 2009 to almost zero in 2018. We study the impact on firms that were relatively more exposed to that market. We use data on customs transactions linked at the firm level to a manufacturing census, which allows us to construct nice measures of firm performance. What we’re analyzing is the long-term effect of losing a major trade partner.
We wrote this draft, and then I got a call inviting me to present the paper at Colombia’s Ministry of Trade, Industry, and Tourism. As you move forward in your career, policymakers can become interested in the work you’re doing. This goes back to a broader question in our profession: do we write papers to understand the economic mechanisms that explain how the world works, or because we care about policy relevance? In practice, it’s often both. What we write is motivated by those two goals, and ideally, our work should also be policy relevant.
ETRM: Thank you so much for having the interview with us today!
