Why don’t economists study Haiti more?

Tate Watkins, referring to an old Chris Blattman post, asks me such a question.  One might raise a similar query about Sri Lanka.  In both cases there are some obvious answers, having to do with ease of access and languages and also levels of violence (in the past, if not always today).

But I would like to raise a more general issue and I hope the development economists and social scientists reading this can add their expertise in the comments section.  To what extent is the choice of venue for study due to what I will call “social science infrastructure”?

I don’t mean roads and bridges.  I mean having trained armies of local assistants, data gathering and processing facilities, populations which are used to signing informed consent forms, medical clinics which understand how to work with social scientists and register data, and other less visible assets.  I once visited a Poverty Action Lab evaluation in Hyderabad and it struck me immediately just how much local assistance they needed to get their study of micro-credit off the ground.  As far as I could tell, the local assistance seemed really quite able but of course that cannot be taken for granted in all locales.

It has struck me for a while just how many RCT papers appear to be set in a relatively small number of places in Kenya.  Presumably these parts of Kenya have a very good social science infrastructure.

So what is going on here?  Who can shed light on this?  And to what extent does having a good social science infrastructure correlate with other features which will bias the results of these studies?


In my experience donor priorities are the most important reason for why research is conducted in one region and not another. Funding agencies have explicit and implicit priority regions and getting funding for anywhere else difficult (but not impossible). Social scientists know about these when they write grant applications.

The answer is relatively easy: Haiti is a mess. Best efforts and big money could not change the situation for decades, but have just worsened it, including the involvement of former US- president Clinton. There is a lot of analysis, but no strategy how to go on within the prevailing mental models. The paradigm of international donor community has reached its limits. Thus, we wait for a shift of the US-approach (very low probability) or a powerful-enough global player, which is willing to challenge the dominance of the US in Haiti (even much lower probability). Until this time the optimal solution seems to sustain the status quo. It is a trade-off between the costs of change in the international community as well as the political costs in overcoming institutional restrains and the benefits for Haiti. Bad luck. And there are much more candidates, where problems are encountered in a network of structural problems, no international organisation wants to deal with.

Good points ... but one would think that precisely because "Haiti is a mess" that some young economist would rise to the challenge in order to help sort the mess the out ...

I don't know about Sri Lanka, but studying Haiti could very well lead to clashing with important beliefs held by elites of developed countries. Hispaniola has two populations. One is a disaster and the other is not. The potential of bumping into prevailing taboos is just too great to be worth the effort.

Perhaps it's because of this cautionary tale:

"Afraid he would kill me, I pleaded with him to honor my commitment to Haiti, to him as a brother in the mutual struggle for an end to our common oppression, but to no avail. He didn’t care that I was a public choice scholar. He told me to shut up, and then slapped me in the face."


Perhaps this is an extension of the WEIRD problem. We study people in places that are easy and cheap and popular / lucrative to study, and then extrapolate to human universals without recognizing the systemic selection biases introduced by our choices.

I'd like to echo Stefan Siewert in that there exists 'impact bias' or, 'I want you to show me something that gives me justified hope and optimism about the existence of some low-hanging fruit whereby a reasonable about of smart-funding can produce dramatic progress quickly.'

You definitely see that kind of bias in the philanthropy community about, say, cheap soap and vitamins and vaccines, or the enthusiasm for microcredit. The overlap between the social science, policy, and charitable communities in this is definitely not coincidental - the demand signal is for actionable intelligence. It's like wildcatting for oil - you can drill a hole in the middle of nowhere, or you can hope to strike in the general region where people have hit paydirt before.

It's not just Haiti, economists in general don't study informal economies what some call System D. Most economists dismiss the informal activity as not very relevant, but some estimate, that in some parts of the world, most economic activity is off the books, and not part of what is called the formal economy.

Well you need other sources of information to impute economic activity such as consumption / income / capital related information in various types of surveys. Presumably the work on the ground is mostly done by people from the country.

One might raise a similar query about Sri Lanka.

Sri Lanka is just not special enough. On most any dimension. If you look for abject poverty Pakistan or India or African nations work better. Neither a paragon of massive reforms nor massive misrule. Their infant mortality figures etc. are actually pretty good too. If you are looking for impact, there's larger playing fields. No resource curse, nor a basket case. No flamboyant dictators, nor any egregious ones. No birthplace of novel movements a la micro-credit or cellphone money or islamic hedge funds or some such.

In short, it's just not interesting enough for economists. It's just too ummm normal..... (Sri Lanka would be an interesting case study in anti insurgency though)

I'm not sure Sri Lanka is relatively that understudied by economists, compared to many developing countries of similar population size - but of course it is way understudied relative to say the US or UK. Some examples:
- I have a range of papers on small enterprises with Suresh de Mel (based in Kandy) and Chris Woodruff

- Seema Jayachandran and Adriana Lleras-Muney have a paper in the QJE looking at how drops in maternal mortality risk translate into more investments for girl's education:

- The World Bank Policy Research Working Paper series has 20 papers on Sri Lanka over the past 5 years, including a macro paper on savings, work on productivity and the apparel industry, and impacts of land restrictions:

1. I am glad that David McKenzie commented here.

2. Even top development economists underestimate how much research the World Bank does, about everything, everywhere: http://jacobageller.com/2012/03/is-the-world-bank-disconnected-from-critical-areas-of-science-and-research/

I suspect that--although with different colonial masters--the cultural similarities between India and Sri Lanka (and for that matter Pakistan and Bangladesh) mean that lots of research done in India is easily generalizable to Sri Lanka. Over time, because of the hard borders and animosity between India and Pakistan and Bangladesh, the former will diverge culturally from the latter two and development lessons from India will not be applicable in Pakistan and Bangladesh. But we are not there yet.

Also most development economists are English speakers so they will be primarily drawn to English speaking countries. They will also stay clear from countries that seems too unstable, thus they will chose Ghana over Nigeria. I also suspects it's not a coincidence that Kenya is a major tourist destination while also attracting a large number of studies. Researchers may not be indifferent to a country that seems to represent the epitome of "Africa" with its Safaris and wild animals.

Oops! Just noticed I repeated the obvious about languages and violence/instability.

This reminds me of the joke about the guy who was looking under a streetlight for the money he had lost down the block because the light was better under the streetlight .

Perhaps also because the Haiti depicted in the New York Times bears little relation to the actually existing country.
Peter Hallward's Damming the Flood, about the role of foreign powers and local elites in curbing democratic reform in Haiti, and about the way ideology determines coverage of Haitian politics, is the most important book of the last decade.

"to what extent does having a good social science infrastructure correlate with other features which will bias the results of these studies?"

this reminds me of a nice talk that Angus Deaton gave at NYU that touched on this issue. See http://www.nyudri.org/events/past-events/annual-conference-2012-debates-in-development/. The material on RCTs starts a little before minute 11.

I've just completed a field experiment in Haiti which was VERY difficult to organize and manage (a first draft of the paper should be available within a month). There are just too many reasons not to site a field study in such a difficult country unless you have a clear idea why Haiti is of particular interest. There are lots of reasons (and we think we've found a few), but if you just want to test a hypothesis that could in principle be tested somewhere else, you'd almost always pick somewhere else unless you curiosity about Haiti gets the better of you (as it did for me).

"unless you curiosity about Haiti gets the better of you (as it did for me)", he said, destroying any credibility his study might have had...

Mike, I'm curious why you think curiosity is a bad thing in a researcher...

Just guessing, but I don't think the credibility comment was serious about your credibility but about your use of language - "that math problem got the better of me" would generally mean that you tried to solve it but could not, whereas something like "the call of partial differential equations got the better of me" would generally mean that you resisted specializing in PDEs but eventually your resistance was overcome and you were, one hopes, successful ...

Agreed. Haiti is a case of such persistent resistance to development that studying it ought to provide some real insight into the causes of persistent poverty and underdevelopment. Why IS Haiti so resistant to development? Of course it would take a really objective observer to do that, and my guess is that most development economists show up up with one axe or another to grind already in mind.

Eminent domain, maybe?

I think the “social science infrastructure” is very important in facilitating successful research, but also think that it is not exogenously allocated. In terms of RCTs, I think it is more important the willingness to conduct them by governments and NGOs. If you find a place where, for example, a government is willing to conduct rigorous evaluations, over time, the “infrastructure” will be developed. I did an RCT in Peru around 6 years ago and had to do quite a lot of investment in training myself. Now there is a “friendly” government there and many RCTs are being conducted.

An IPA office tends to be of great help. In Kenya, where I am conducting one right now, the office its great.

Finally, yes, the World Bank does research everywhere. In terms of observational studies, the key resource is the availability of high quality data. There, again, in my opinion, the World Bank did a superb job –one that should be even increased- by developing and collecting the LSMS.

This post would have meant more to me had the acronyms been actual words.

Sorry. RCT: Randomized Control Trial.
IPA: Innovation for Poverty Action - http://www.poverty-action.org/
LSMS: Living Standard Measurement Study - http://econ.worldbank.org/WBSITE/EXTERNAL/EXTDEC/EXTRESEARCH/EXTLSMS/0,,contentMDK:21610833~pagePK:64168427~piPK:64168435~theSitePK:3358997,00.html

You need an economy to have economists study it, otherwise it's anthropology.

Actually I think that Pakistan is very very understudied by economists. Perhaps economists think of it as a substitute to India, but certain problems (particularly religion-related terrorism, gender attitudes) are much more pronounced in Pakistan.

Could I dispute the premise? Haiti has just 10 million people, about the size of Ohio. Suppose research is distributed roughly by population. We would expect 100 times more papers on India, 20 times more on Indonesia and 20 times more on Brazil, and even 4 times more on Kenya. Is there any evidence that if we predicted research on the basis of population and average education and perhaps gdp per capita (things that could naturally account for the magnitude of research, perhaps causally via "social science infrastructure" and perhaps via other channels) that Haiti is under researched?

Would the number of economics Ph.D. students (or professors for that matter) in major U.S. universities from the country matter? This should be more relevant if they do undergraduate studies at home so the information is indicative of the "social science infrastructure" of the given country.

I once had a professor who said sri lanka is a country of contradictions. We have high literacy but poor productivity etc. Good people but lots of bad things happening. We fight during the morning, discuss peace during lunch, and dance and drink at the beach in the evening. Same routine next day. Mind you we've had colonial economy, closed economy, open economy, mixed economy, post war economy so on, with the latest trials, on hyper infrastructure and investment. However the common man and woman still is resistant to anything outside the status quo and really has little imagination for creating, transforming or transacting between the complex. Just a simple trader mentality. Lots of socio-cultural reorienting will be needed if we are to get it right (this time) and truly go up the economic ladder. Otherwise the nuts will not fit the bolt (again), and we have a history of that. And too many nuts also with no bolts, and no machines to put them on also. Haiti may be a basket in different ways.

Apart from the infrastructure built on concrete, it is crucial for an economy to build up on the social infrastructure as well. Having said that, even some of the developed countries are facing problems with nurturing their human resource. Case in point: Quality education and affordable healthcare in America and U.K.

Differences in social science infrastructure are indeed the proximal reason why so few economists study Haiti. More fundamentally, though, we might ask why social science infrastructure is so drastically different across developing countries. The reason is that development research is incredibly path-dependent. In several important senses, most social scientists working in Malawi are still riding on the coattails of a project that started there over 15 years ago.


Comments for this post are closed