Monday, January 4, 2010

Rant on Development Economics

Here's my take on Development Economics, wrong or right.

Development Economics is not a "main-stream" Economics field. Most Development Economists are Development Economists because they really do care about the worlds' poor, they tend to be more liberal and less dogmatic than your average economist, and many of them view other, more mainstream fields such as Macroeconomics or Macro Growth as witchcraft. They often point out that in Macro, you've always got these really difficult identification issues, and it's often impossible to say w/ any certainty that you've identified a causal relationship as opposed to just correlation. With Micro-Development Randomized Trials, it is possible to know "things" with much more certainty.

While I can understand the views of Development Economists, I cannot really remember being impressed by any Development Economists, or have ever walked away from a Development Econ presentation/class feeling like I've really learned a new, important insight into the fundamental question of Economics: Why are some countries poor? And what can "we" do about it? And this was my feeling today when I sat in on a first graduate development lecture of the term. First I should say that the professor was an eminently likable guy who's heart is in the right place, and that there are many readings on the syllabus which I was not familiar with which look like they could be very interesting papers: Bliss & Stern's (1978) "Productivity, Wages and Nutrition", and Hoddinott et. al. "Effect of a Nutrition Intervention During Early Childhood on Eocnomic Productivity in Guatemalan Adults". At the same time, I can only sit thru so much blather on the Household Production Model and blather about rural credit markets (maybe this means I have ADD).

The real problem with the course and syllabus is what it did not have: There was not a single geography paper. Not a single Economic History paper. Not a single Macro-Growth paper. Not a single book of any kind.

Is geography really altogether unimportant? Does it really make no difference for manufacturers whether they produce high up in the Bolivian mountains or next to Iron Ore Deposits on a navigable waterway next to Chicago? Is the whole of History irrelevant because it's impossible to go back and do a randomized trial? (For that matter, I suspect that only a fraction of the papers on the syllabus are randomized trials -- it looks like many are simply rural Micro-Theory papers -- and of those which are, I wonder how many tell us something we didn't already know and carry important policy implications.) Perhaps the whole of Macro Growth is irrelevant (I don't actually think so, I think there are some diamonds in the rough), but is the whole field of Economic Geography irrelevant? Urban Economics too? Is there nothing to be learned by looking at detailed case studies at countries which have managed to develop?

I'm of the theory that a syllabus for a graduate course is a very intimate, personal thing which says everything about you as a scholar. Everything you've ever learned about your field which was not already obvious to someone outside of your field should on that syllabus. Since Development is a wide-ranging topic, the syllabus should be wide-ranging as well. It goes w/out saying that 2/3rds of the stuff listed can be "background" readings, and that the class itself should focus on a more specific, focused slice of the literature.

The problem is that I didn't get the impression that this professor, or other development economists, had read, say, Jeffrey Sachs "Tropical Underdevelopment" paper, or Alfred Crosby's "Ecological Imperialism" and found it wanting. My sense is that they haven't even read them.

What they have read is what other Development Economists who they see at conferences each year write, who all think the same way, who all edit and publish in the same set of journals.

The question is, then, what is the difference between a "Macro/Micro Development Economist" an "Urban Economist", a "Growth Economist", an "Economic Geographer", or an "Economic Historian?" They all try to get at the same two questions -- why are some countries rich and other countries poor, and what can we do about it. But they do not talk to each other. They cannot be bothered to read each others papers, or attend each others conferences. The lines between these fields are delineated, basically, in a more-or-less completely arbitrary (and heavily path-dependent) manner based on who your friends are, what kind of papers they write, and what their and their friends' priors are about the fundamental question of economics.

The syllabus I received will be an extremely helpful resource if I try to publish any papers in a development journal -- as it will allow me to cite the key development papers which will impress (and likely have been written by) referees. Reading all of these papers will also show me what kind of papers, data, methods, and writing style is considered acceptable for the field. Yet, it's much less clear how many of these papers will teach me something I do not already know and have policy implications -- after all, I already suspected that having poor nutrition growing up would be correlated with time spent out of the labor market later in life, however nice it is to see a confirmation of this.

Few, if any, of the papers on the syllabus are "big picture" papers. This follows from the anthropology of Development Micro people -- randomized trials are King, it's impossible to do a randomized trial of industrial policy for different countries, so Development Economists would prefer to be silent. Fair enough.

The only problem is that if I'm the President of Madagascar, I've still got to answer these "big picture" high-development questions and decide on a "grand strategy" and choose my actual policies. (Just because Esther Duflo can't think of how these questions can be answered doesn't mean that they aren't actually answered by people making policy choices.) Wouldn't it be nice if, instead of relying on my priors, there were an academic discipline designed to help provide insight into answering these questions, as difficult as they are? Perhaps we could call this discipline "Development Economics"!

In any case, I'd like to hear why I'm wrong.

UPDATE: I just surfed over to Chris Blattman's blog, and read clicked on his syllabus for his: African Poverty and Western Aid" Course, and most of the criticisms I have above are not valid. His syllabus actually looks quite good...


  1. Geeze, you really ARE Thorstein Veblen!

    I have been long been convinced that Veblen's interest in the economics of development is because he spent his youth watching people doing nation-building at the very edge of human settlements. It why his insights into the process are SO much more interesting than say, Marx's--who got his information sitting in a library.

    Anyway, I gave your blog a shout-out today because I really think what you are doing in important. The reason is simple--if we are ever going to build a sustainable society, we are going to need profound insight into the process of economic development.

    As for why people interested in development economics are so fragmented, just remember, their type have been marginalized in the economics profession for at least 35 years. Perhaps you could introduce them to each other and perform some crude taxonomy on the various currents of thought.

  2. Hmmm, this seems to be actually a "Rant on Fill-the-Gap-with-Your-Favorite-Flavor Economics". Academia has been too taken away of late with Adam Smith's old tales about the wonderland of specialization...

  3. I think when you say "Development economists" what you mean is someone focused solely on randomized controlled trials--and there are a few people like that, but just a few. Even Banerjee doesn't fall into that camp. Duflo might be the only one--and she's the best economist of her generation!

    All of these people have read or are familiar with those geography papers. The 1st point everyone makes is the Robinson, Acemoglu, Johnson paper from 2001 and then point to its sequels. Lots of deveopment readers meant for undergraduates belabor these macro points at the expense of the micro issues (which I prefer honestly), which is why Blattman's syllabus does, it's for an undergrad course for non-econ (or non-math) people.

    Also, there are lots of development macroeconomists too. It might not look that way on your syllabus because there is a huge schism between macro and micro development economists (Dani Rodrik has a good paper on this). And speaking of Rodrik, he has a good book that does exact what you ask for: detailed case studies of countries that managed to develop.

    Anyway, since your beef is (apparently) only with development micro, I think there are a few interesting results, starting with credit markets. There are to many models to go over but all of them seem somewhat inadequate now that microfinance burst onto the scene, but they do help to think about the markets. There are lots of good papers on savings and insurance and to what extent the savings policies and insurance systems for the rural poor are rational/close to optimal. They're pretty good it seems, but not perfect so behavioral issues exist. Speaking of behavioral issues there are a couple of great papers on health and fertilizer about why people don't or do use health services and agricultural inputs. A lot of the beavhior is bizarre, which ISN'T obvious. We've gone from knowing the poor are doing something wrong to identifying and area where they are doing something stupid. Now if we want a silver bullet and that takes more time (forever).

    I think the micro stuff is boring too and would rather focus on "big" things even if I'm actually just spinning my wheels. But if you want to claim the evidence is too murky and we haven't learning anything important than you're talking about the whole field of economics (think IV-based studies, business cycles, the EMF).

  4. I live in a developing country, and by and large, the people who actually _do_ development policy do not read or take interest in the field of development economics, as defined by the mainstream journals.

    I can't recall the last time one of the papers there actually mattered to something that actually got done in a developing country. In contrast, there are things that happened in non-development-journals which actually mattered to what got done in developing countries.

    When I meet young people interested in development policy, I always start by saying: Do not read the field called development economics.

  5. At least you are in a top 30. I am stuck in a mapless school in Canada. My international development professor is a self proclaimed neo-marxist. He doesn't know what randomized control trials are.

  6. My response of sorts:

  7. Thank you, Steve. My thoughts exactly in terms of content and ideas for further reading in "development economics"

  8. You're getting at a very difficult practical dichotomy that's very hard to escape in Development studies.
    The "little picture" stuff which your class (and my job, incidentally) focuses on with complete tunnel vision has questions that can be rigorously answered, problems that an NGO or community-level policy can affect.
    But as you so eloquently point out, they also leave you feeling like you've totally passed over the "bigger picture", the "real issue".

    Then the "Big picture" stuff has enormous questions. Ones that are of monumental importance to UNDERSTANDING the issue, but are almost impossible to rigorously answer, and can usually only be addressed politically by Big Governments and the like.

    Obviously both are important, and both need to be addressed, and I'm really not sure which I'd rather spend time looking at. But an important thing to remember is that if you're really looking to do something, anything, that might have a lasting impact, it might be better to start with "you-sized" problems that could have "you-sized" solutions, rather than starting with "I'm the President of Madagascar..."

    Not that the president of Madagascar doesn't need good men with a lot of data and insight. And training to be the guy that guy looks to for advice is a worthwhile pursuit. But there only so many men making "big picture" decisions, and there are Oh So Many economists.

    I agree with you in so So many ways, and I'm working through my own cute little existenial crisis right now. But I want a social science degree with the tools to deal with a problem, one that I can *really* study and maybe-just-maybe one I can *really* address with my limited resources. And it looks to me like Ms. Duflo will get you that far with greater certainty than the best of Macro growth theorists. Because even with the best "big picture" research, "big picture" problems are just hard for us mere mortals to address.

    So sure, that syllabus is unbalanced in a bad, bad way. But I can certainly see why someone interested in development would do it that way.

  9. Re: Elliot -- nice comments. but, hey, i've had development professors who teach courses which just focus on the household production model and small issues, and it so happens that several of them have been called up by developing governments and asked for advice.

    I suspect they didn't give very good advice.

  10. Re: Steve. AJR 2001 is garbage. The Crosby/Sachs/Diamond/Kamarck stuff is good. I've been assigned only one Sachs paper in 40+ econ courses taken, one diamond book (in undergrad) and none of the other stuff... Was never assigned Krugman's geo & trade either... I have also never had a course which discussed any of this in any detail...

  11. Thorstein, thank you for an interesting take on development economics and particularly on learning about development economics at a graduate level. I sympathize with many of the points that you make, although I do not wholly agree with them all. At the same time, your post and the resulting commentary has given me an impetus to expand my reading list to include a broader range of topics as you and your guests recommend. Cheers!

  12. Wow, you must have a really bad development instructor. Plenty of development courses have lots of big ideas, and lots on geography and history. See Ted Miguel's syllabus here:

    You really need to learn something about inference from small samples.

  13. Perhaps "Anon" -- but it's not like I've only read/heard/experienced the thoughts of one Development economist. I've sat in on dozens of classes, hundreds of papers, and clicked on dozens more syllabi.

    Even though it's just one data point, Ted Miguel's syllabus is AWFUL. The days are gone when you can put three Acemoglu papers on your syllabus (and one paper by the fraudster Alwyn Young), and still ask people to respect your intellectual ability.

    Also, I don't see any good macro development papers on his syllabus, even though there are lots.