Introduction to the Development Economist

The four things I work on most

Professionally I am a development economist. I started preparing to be a development economist when I was 21 years old and in 2025 I am 66 years old and am still active and engaged in the same field. This is not to say I have done the same thing or stayed in the same place. In the course of my career I have worked both in academia (Harvard Kennedy School, Oxford Blavatnik School of Government, London School of Economics School of Public Policy) and in the World Bank, off and on, from 1986 to 2007 and was affiliated with the think tank Center for Global Development for many years. I have lived in India twice, 2004-2007 and 2012-2013, and in Indonesia, 1998-2000, in Argentina, 1978-1980, and in the UK, 2018-2022. I have visited and/or worked in more countries that I am years old.

Development economics is an extraordinarily broad domain as it is the economics of pretty much everything that happens in most of the world—as only about 1 billion people live in “developed” countries. For a committed disciple of an academic discipline, economics, that believes in specialization, I have written and worked on far too many issues and places, but there are four distinct topics I currently work on: Labor Mobility, Economic Growth, Education, and State Capability.

Labor Mobility. This is not the same as “international migration” which covers all movement of people across national borders and all of the many reasons people move. My focus in research and in advocacy is on people moving to make higher incomes, whether they intend that movement to be temporary, rotational, or permanent. I co-founded with Rebekah Smith a small, but growing, non-profit Labor Mobility Partnerships (LaMP) to promote more and better legal pathways for people to work in developed countries.

Economic Growth. I have done empirical research on three aspects of economic growth. One, showing that rapid, sustained, inclusive (enough) growth in developing countries is important, and in many cases, an empirically necessary and sufficient condition, for large improvements in human material wellbeing. Two, showing that economic growth in developing countries is not a single exponential trend but rather is largely episodic, with periods of booms and busts, accelerations and decelerations. Three, trying to move beyond broad platitudes and ideology and create growth diagnostics that give specific, contextual, practical recommendations about what countries can do to create and sustain accelerations and avoid decelerations.

(Basic) Education. The field of basic education is both one of the great success stories of all human history and, at the same time, one of the biggest development challenges. Achieving universal basic education has always been central to the idea of national development. And in the decades since 1950 the world has been fantastically successful at getting kids into school and keeping them there for more and more years. By 2015 the typical adult in Haiti (a development failure in nearly every aspect) had more years of schooling than adults in France in 1970. Huge, history changing success. Yet the core idea was never really universal schooling, it was universal education: that children would leave their youth with the skills, knowledge, capabilities, beliefs, dispositions that would allow them to be successful—by their own values and goals—adults. One of the biggest development challenges is that “Schooling Ain’t Learning” and due to a ubiquitous, but not universal, learning crisis most children (and in many countries nearly all children) leave basic schooling without basic education. For eight years I was the research director of the Research on Improving Systems of Education (RISE) a large, long-term, multi-country, multi-disciplinary, research project into the deep determinants of—and hence leading to scaled, systemic, solutions to—low levels of learning in basic education.

State Capability. Much of development research, and policy work generally, tends to focus on “what” governments should do. This tends to ignore the question of how governments can create and sustain public organizations that actually have the capability to implement with the fidelity those actions that would be effective in accomplishing the tasks they are assigned. There are domains in which there is little disagreement the government legitimately needs to be engaged, like law enforcement. All countries have police forces, but how effective these police forces are at actually reducing crime is an open question. It is not at all empirically clear that what countries formal policies are has as much impact on outcomes and whether the public sector organizations are capable of implementation. One of my most famous papers that was never published was “Is India a Flailing State?” which points out that the Indian state can do some things impressively well, like holding elections, and even do things of impressive complexity, like send a rocket to the moon, and, yet core responsibilities like basic education, primary health, water and sanitation, and policing were mediocre (at best). This led to the collaboration with Matt Andrews and Michael Woolcock on the book "Building State Capability” and an organizational approach to increasing capability called “Problem Driven Iterative Adaptation” (PDIA) for which HKS provides on-line training.

That said, I have not been so disciplined as to work on only these four topics, I have also worked on and written influential (or at least high citation (Google Scholar link) works (with co-authors) on: social capital, population policies/family planning, foreign aid, health policy and systems, international trade policy, poverty, safety nets and others (strangely perhaps my single most cited publication is a technical paper about how to estimate a household’s economic status with just simple questions about asset ownership, without income or consumption).

One last thing in an introduction. I am sometimes known as a “critic” of the use of randomized controlled trials (RCTs) in development. I have been engaged in many big picture debates and I win some, have no impact on some, and lose some. Some might take the fact a trio of economists won the Nobel Prize in economics for promoting RCTs in development economics is a sign I decisively lost. However, I heard a song from Dido the other day that has the lyrics:

But I will go down with this ship
And I won't put my hands up and surrender
There will be no white flag above my door

The way in which, and the reason, I am a “critic” of RCTs is that they have never proved particularly useful in answering any of the important questions I have had in my career in a practical and definitive way. While the randomistas are correct (enough) about one narrow methodological question of creating unbiased estimates of the causal impact of specific interventions there are two big problems. One, the issues of external validity and construct validity are, in development economics, just empirically huge and make the results of any specific RCT just one piece of information in any context specific decision and relying on RCT results can be demonstrated to lead to worse decisions that using context specific but weaker methods (here). Two, the methods of RCT require a randomized separation of units into “treatment” and “control” and hence are mainly useful in answering questions for which the individual (or household) is the relevant unit, but the most important questions about development are not reducible to individuals. Most of the world’s poverty is not because the world is full of poor people but because people are living in poor places (countries) and without shifting the reality of the country/place there is some, but limited progress to be made (here). I am not “against” RCTs I am “for” things (labor mobility, economic growth, high learning quality basic education, more effective public sector organizations) or which RCTs mostly don’t help, even though they vacuum up enormous intellectual and financial resources.