The Challenges to Creating Country-Specific Policy


This page in:

Editor's Note: Jishnu Das and Quy-Toan Do are Senior Economists in the World Bank's Development Research Group. They are authors of a working paper U.S. and Them: The Geography of Academic Research.

Sit in on policy discussions and dialogues with governments and development institutions and two themes recur time and again: “What policies, specific to our country, can improve _____ outcomes” (fill in the blank from a large list including growth, poverty, education, and health) and; “We are not sure that the policy in Y country is really relevant for us because of differences in _____ (fill in the blank from options including institutions, legal regimen, culture, and diversity among others).

Leading researchers now agree. The recently concluded Growth Commission, mandated to understand the sources of growth and appropriate policy, argues in its final report for country and time-specific growth strategies: 

Wedded to the goal of high growth, governments should be pragmatic in their pursuit of it. Orthodoxies apply only so far. This report is the product of two years of inquiry and debate, led by experienced policy makers, business people and two Nobel prize-winning academics, who heard from leading authorities on everything from macroeconomic policy to urbanization. If there were just one valid growth doctrine, we are confident that we would have found it.

—Growth Commission Report, Introduction, Page 4

If policies need to be country-specific, we clearly need a body of work on a country before we start advocating policy. This usually leads into the third recurring theme: “We need to re-position ourselves as a key knowledge institution and use that knowledge to generate specific policy for the country”.

Our new working paper suggests that producing that knowledge may not be an easy task.

Between 1985 and 2004, there were 76,046 empirical economics papers published in the top 202 economics journals, of which 36,649 were on the United States alone. That compared to the 4 papers published on Burundi during that same time period or the 9 published on Cambodia or the 20 on Niger.

This vast disparity is explained almost entirely by strong correlation between GDP and publications—a doubling of GDP per capita leads to a 37 percent increase in the publications per capita in the country. The US is bang on the regression line relating GDP to publications—a lot more is produced on the US because it is big and rich. Surprisingly, we are unable to find clear evidence of a link between the availability and quality of data and the number of publications.

(Click to enlarge)


Furthermore, there is a large premium in the top economics journals for papers focused on the US. For example, the American Economic Review has published one paper on India (on average) every 2 years and one paper on Thailand every 20 years. Neither are these numbers particular to this prestigious journal. Over a 20-year span dating from 1985 to 2004, the top-five economics journals together (American Economic Review, Econometrica, The Journal of Political Economy, The Quarterly Journal of Economics and The Review of Economic Studies) published 39 papers on India, 65 papers on China, and 34 papers on all of Sub-Saharan Africa. In contrast, they published 2,383 papers on the US.

Controlling for authors’ affiliations, and hence partially accounting for unobserved quality suggests that papers on the US are 2.6 percentage points more likely to be published in the top-five journals. This is a large effect as only 1.5 percent of all papers written about countries other than the US are published in the top-five journals.

We would love your feedback. Do you think:

  1. That policies for improving education in (say) Burkina Faso can be based on research done in South Africa or the US?
  2. That the US-premium in the top-5 journals leads researchers to divert their attention from other countries to the US? This may be important because in low-income countries typically a third or less of the research is produced by “in-country” institutions and the remainder by researchers in the US or Europe.
  3. What policies would work to “incentivize” knowledge production in low-income countries? Given that there were only 4 papers on Burundi over the 20 years, we may think that every additional paper has large value-added in terms of adding to the knowledge stock about the country.

Finally, we would love to hear your stories. Appendix 2 in the paper lists all the countries, the number of papers written on them and the number of papers in the top-5 journals. Do you have a specific story about economics research in your country? Write in, and let us know!

Join the Conversation

January 13, 2010

Jishnu and Quy-Toan: I know that you guys have tried to address the issue of data quality and availability, which can be a major source of bias toward high-income country research. One additional wrinkle you can consider about the data dimension is to think about how the dimensionality of the data can affect the quantity of research. Sure, the release of a new, high-quality household dataset can (and should) affect the number of research papers written at the intensive margin, but given the fact that there are only that many questions one can ask of household data (along with how widely non-household data are distributed), there will be less of an impact at the extensive margin. Concretely, think about how many applied microeconomists (a category for which I include trade and HD) there are, versus the number of empirical quantitative finance, labor, monetary, IO, and international finance guys out there. Thus, if (say) I'm interested in looking at the relationship between equity and bond returns, I really am limited by the availability of core and control data for countries that I can study, even if I have a prior interest in an emerging market like Vietnam or Botswana. It's not a problem for which I have a ready solution, but I'm just pointing out that disentangling researcher interest from constraints is really very hard. Increasingly, however, the saturation of papers using the same datasets (and the publishability of papers employing a unique dataset) has led to more primary data collection. As a consquence, we're seeing empirical researchers who are doing/have done their doctoral work at a U.S. university going back to their home countries and collecting primary data. This is a positive trend and bodes well for development research (and research on developing countries more generally).

Gaurav Datt
January 14, 2010

This is very interesting. But a couple of quick comments.

i) The US-bias of AER and indeed the "top-5" is well-known, though it is nice to see that documented. So, very glad that you have made the effort to do that. But for many of us "development practitioners", the "top-5" are certainly not the first source we are eager to look up for development-oriented empirical research. We would rather turn to Journal of Development Economics, J Pub E, EDCC, Review of Income and Wealth, et cetera, et cetera ... I think you know the potential list. So, it would of interest to do the analysis with some sort of a "development filter". Probably, the rich-country bias will still be there but it may be less pronounced. It may also influence your "surprising" findings on the non-significance of the availability and quality of data.

ii) When looking up recent work on India or China, I often find myself looking at papers published in journals such as the Economic and Political Weekly or the China Econmic Review. Not sure if these make it to your list of journals. These may be less prestigious, but a lot of good, relevant material is published in such in-country or country-specific journals. Again, a consideration of these sources will almost certainly temper your findings.

giulio quaggiotto
January 14, 2010

What a brilliant post! Let's hope it will generate a more mature debate about the myth of transferring lessons learned that still permeates so much of development work. "Hindsight cannot lead to foresight"

Thanks for sharing


Nilgun Gokgur
February 23, 2010

I find this posting very interesting and I only want to emphasize what is already said above... That is the "non-significance of availability and quality data." Any economist is often thought of being as good as the data he/she works with. Yet more research networks are needed among industrialized and less industrialized countries of Africa and Asia.