Chris Blattman

Aspiring PhD students: Should you become a field research assistant for an RCT?

Share on email
Share on facebook
Share on twitter
Share on linkedin
Share on reddit

Tuesday I advertised a new research position with me, on a field experiment in Liberia. One reader asks a very good question:

I have a vague recollection of one of your posts a few weeks (months?) ago saying that if you were to pick dev research like stocks you would go short on RCTs. It’s not great news if the PI I’m working for thinks what I’m doing is a suboptimal use of my time.

Astute. And worth answering not because of my position, but because this is something aspiring PhDs should think about. The answer is most relevant for aspiring economists and increasingly important for aspiring political scientists.

And the answer, like every answer in social science, is the same: “It depends”.


  1. It’s getting harder to distinguish yourself from the herd. JPAL and IPA and other field experiment outfits have exploded in size. Whereas maybe a dozen field RAs applied for economics PhD programs a few years ago, now it is dozens. Maybe dozens upon dozens. This is less true for political science, where the tide is surging but still low. But not for many more years.
  2. Enthusiasm for RCTs is moderating. Enthusiasm is still high, and will remain high. But we’re beginning to see the warts and the limitations.
  3. You might get less interaction with a professor than you think. This varies from academic to academic, and project to project. You might find yourself in daily conversation or you might find that they barely know your name. The latter is seldom the case, but that is the spectrum. If your aim is to get a good letter of recommendation, one side of that spectrum is better than the other.


  1. You may never have a chance to live in a developing country again. In your PhD, you’ll be confined to a trip of a few months maximum, here and there. As a professor, you’ll reach the absurd level of visiting four African countries in 14 days. This is your only chance. And it will make a difference about the kind and quality of questions you will ask and care about.
  2. You will learn hugely useful skills. I can’t speak for other professors, but my RAs learn to design questionnaires and behavioral games, do better qualitative work, code Stata, run real-world regressions and deal with real-world statistical challenges, hire and mange a team, manage and budget hundreds of thousands of dollars, work with policymakers, write for a wonkie audience and, most of all, manage harried and absent-minded professors. You will put all of these to work in any dissertation in the field.
  3. You will be given room to excel. You will be more independent than you expect, which will be scary, but probably good for you. Your bosses will throw as much at you as you can handle, and then some. Over two or three years you will usually get the responsibility you deserve, which can be a lot, and take you great places.
  4. You might get more chances to impress than you think. Even if you’re not on the phone every day with the professors on your project, you have many opportunities to shine, and we look out for those and they land in our letters of recommendation.
  5. It will lead to research ideas. It is no coincidence that Michael Kremer’s students are mostly celebrated academics themselves now, and all run projects in Busia. Likewise you will see the same for the Duflos and Karlans of the world. And my cohort all have PhD students who are collaborators and c o-authors who are finding great dissertations in our field sites, RCT or not. It’s a good machine.
  6. It’s tough to get other kinds of positions with development professors. They are there. I am sure the Besleys and Rodriks and Robinsons and Acemoglus of the world hire full time RAs, but I am not sure how many or how often or simply how. that might be telling.

How to maximize the pros and minimize the cons?

  • See if, after a couple of years in the field, there are opportunities to work with your professor as a US-based RA. This is never assured, but it’s worth asking. It gets you more data experience and more interaction. And there will be hot showers.
  • Do things to distinguish yourself from the herd. this is kind of obvious. For economists, it might mean seeking to spend an extra year working, perhaps at the Fed or with a theorist, or taking extra math classes, or something like that. For a political scientist, it might mean carving out time for an original research project, however small, since writing samples are taken much more seriously. Ask your professor.
  • If you have the luxury, favor projects where you’ll have a closer relationship with the academics.
  • On your applications, don’t say, “I want to randomize everything”. Here’s my thoughts on the future of quantitative field work. Also this.
  • Remember that everyone else is playing the same game as you, and might even be seeing this advice. Paradoxically, that might make all the above advice now strategically sub-optimal.

Most of all, don’t get too calculating or too strategic or too anxious. Otherwise you will be miserable and a jerk.

Frankly, admission is pretty idiosyncratic and largely out of your hands, and  you’re going to have a happy and fulfilling career either way. Really. So chill out and enjoy yourself and do the work for the right reasons: if you enjoy it and can learn a lot. that’s good enough.

Current and former field RAs of all stripes, your comments please…

14 Responses

  1. Another CON: NO pay for internships. This certainly must exclude Africans. NO ETHICS. Without ethics, science ala Mengele.

  2. Convincing. I applied.(….hopefully my app will make the pile that arrives on your desk….)
    Glad/thankful (and a little proud too he he) that my comment triggered this good post & further comments.

  3. I’m currently a field RA for an RCT in Africa. I would encourage anyone thinking of pursuing a position like this to consider the following:
    1. Will I enjoy the experience? Being a field RA is really hard, and I have seen a number of unhappy field RAs. You are likely to end up working long hours in tough conditions. You have to be able to manage stress well. If this doesn’t sound like you, you won’t enjoy the experience, and as poster Eli says, you can burn out, perform poorly, and end up unenthusiastic about development research.
    2. Will the skills I learn complement my previous experience? The pay-off from working as an RA really depends on what experience you already have, and what you want to learn. Make sure the particular position you are looking at will give you the experience you want. For example, if you don’t have a lot of experience working with real life regressions, make sure your project will offer the opportunity to do data analysis. If you have a lot of experience working with data, but haven’t had a chance to pursue your own research questions, make sure your project will allow you to be involved with research design.
    In summary, I don’t think anyone should become a field RA for an RCT just because he or she wants to get into an econ PhD program. You should become a field RA because you think you would enjoy the experience, and you would value the skills you would gain from it.

  4. I’d venture to say that another positive is that conducting field research could help you realize whether or not one wants to be a development economics academic. The previous RAs on my project are doing the following: working for a consulting firm in London, going to Seminary school, getting a joint MBA/MPA, earning a JD and yes going on for PhDs in economics at some of the best institutions in the world.

  5. It’s funny last summer I applied and interviewed for a couple of IPA positions, one of which was a project you lead in Liberia, but didn’t get either job. My plan was to get some field experience and then apply for an economics Phd. Now I’m starting a PhD program focusing on development without the field experience, but not because I didn’t try…

  6. Any ideas for those who made the mistake(?) of being liberal-arty/Mathless in undergrad and are hoping to get RCT RA work?

    1. I figured out that I wanted to do an Econ Ph.D. late in the game. I didn’t take any math as an undergrad and then after I graduated I realized that I needed to do that! I ended up taking six math classes while working at the Fed. The need for math prevented me from getting actual field research experience before grad school. Now as a Ph.D. candidate I can say that it’s possible to work as an RA for a professor or even get funding to do exploratory work (what I’m doing now).

      Do I agree wholeheartedly with Chris’ post about the benefits of working as an RA in the field? Yes. Do I think that I would have gotten into any decent program if I had done that? No, because the admissions committees care about math grades and not just practical and useful skills that actually help you do quality research. Do I think not taking math as an undergrad was a mistake? Not really because I think there is value in a wide breadth of knowledge.

  7. Re “You may never have a chance to live in a developing country again” – Is this a challenge caused by the workings of the tenure system for econ PhDs? Would it not be possible to work as an academic in a developing country later? Or take time out of academia and work in another position in a developing country? Apologies if these are stupid questions, I’m not from a US/econ background (I’m doing a geography PhD at a UK university).

  8. Having gone from Busia to an econ PhD program, I really couldn’t have said it better myself. I would just re-emphasize that being an RA on an RCT can be a really hard job. It will require long hours and you may have very little guidance on the day-to-day details; those are your responsibility. Also, remember that “field” is a relative term. You will have lots of opportunities to go places and explore and see the research process from the interviews up, but for the most part, it’ll probably be a desk job.

    Again, the biggest question should be whether it sounds like an experience and a job you’d like to have. If you’re unhappy there, it’s unlikely you’ll do well. But otherwise, it’s one of the best and coolest options available among RA jobs. So I would definitely say the pros outweigh the cons. Definitely.

    1. Good points.

      One caveat: A good many are desk jobs, but Busia might be more desk than some, being so longstanding and well-organized and staffed. It’s a lot of logistics but people coming in at the early stages of a project see a lot of field. I have some RAs who would like to see a little less jungle and a little more desk.

  9. The best advice of the entire post, if an applicant is concerned that working as an RA for an RCT will pigeonhole them, is this: ‘On your applications, don’t say, “I want to randomize everything”’

    I’m ABD in a Poli Sci program. I understand the concern about being pigeonholed based on your work or who you have worked for. But I don’t think this is a very important concern, ultimately. While in grad school I’ve worked for heavily quantitative professors, heavily qualitative professors and one borderline political theorist. I learned something valuable from all of them, but I don’t think most of that work is a reflection AT ALL on what I want to do. But it shows that a variety of people think I’m capable.

    Working on an RCT, in a developing country, would be a great experience, no matter what type of work you want to do later. If you don’t want to do RCT’s, you can say so in the personal statement, or be more vague and say something like “It was a good experience, and I learned a lot about the shortcomings of RCT’s and how they work in practice. I’d like to explore other options with my own research, blah blah blah.”

    I wish I had had a similar opportunity in the years before I applied to grad school.

  10. Chris – are all aspiring PhD students (and/or readers of this blog) from the west? Some of them may have spent a lifetime in Asia/Africa and probably are used to living without hot showers, no?

    1. Ah. Good point. And I should know that, in part because I try (and have been pretty successful at) hiring these folks. But just as some of the advice doesn’t apply to the political scientists in the audience, so too does it not apply to those less accustomed to hot showers.

Comments are closed.