MIT Center for International Studies

Back to Homepage
Audit Archive
Editor's Archive
Audits are short essays, published by CIS, that put the conventional wisdoms of U.S. foreign policy under close scrutiny.

April 2008
Cambridge, Mass.
Good and Bad News on Global Development
By Dani Rodrik

I START WITH SOME GOOD NEWS, because there is, I think, a lot of good news in the world of development. Then I want to present what I think is essentially a paradox. The paradox, to put it very crudely, is that while economic development is working, development policy is not.

Let me start with the good news.

If you look at the total number of people who live on below $1 a day and look at the trend, between 1981 and 2001, what you see is basically that there are now roughly 400 million fewer people who live below the $1-a-day line. So there actually has been not just a relative reduction in the number of the absolute poor; there has actually been an absolute reduction in the number of the absolute poor. This is in a period when, of course, the population of the developing world has increased quite significantly.

In terms of the somewhat higher poverty line, which is the $2-a-day line, the number of poor people below that threshold has actually increased somewhat, but it is still the case that relative to the population of the developing world, it has come down.

That is basically good news. In this period, there has been, in fact, significant poverty reduction around the world.

But if you look at where that has come from, it is also the case that much of it has actually been localized. China alone accounts for the full 400 million-person reduction in absolute poverty when measured by the $1-a-day line. If you take China out, basically, in the rest of the world, some countries have had an increased number of poor people, others have had a decline.

This matters for the rest of the account, because one of the big paradoxes of our time is how well China has in fact done.

Another way of looking at the good news is by looking at another indicator, such as life expectancy. If you don't like income-based measures, you may want to look at life-expectancy as being one of the key determinants of life chances of individuals.

There is a positive relationship between a country's income level and the average life expectancy of individuals in that country. What is less well known is how this relationship has shifted up over time. There has been a true health and life-expectancy revolution. That is, even in countries that are at the same level of income as 25 years ago, the average life expectancy now is easily 20 or 25 years longer than it was for countries at the same levels of income in the earlier period. So for a given level of income, you have much better health indicators, here proxied by life expectancy, than you did before.

Of course, there are a few outliers. These are countries in sub-Saharan Africa which have been badly hit by HIV/AIDS.

Where is the paradox that I am talking about? The paradox is that by the 1970s and early 1980s, the people who focus on economic development issues in North America-primarily the multilateral institutions in Washington, but also in leading centers of thinking about development policy in the United States-had converged on a set of ideas called the Washington Consensus and essentially gave a certain sense of faith that there were a number of policies which, if countries adopted them, would yield relatively good outcomes.

Here is the story for Latin America (Exhibit 1). What you see here is the average growth rate that Latin American countries achieved in the decades before 1980 and the average growth rate that they experienced since 1990. Of course, in the interval was the debt crisis, which is part of the reason why they jettisoned their old policies and took on these new policies that came to be called the Washington Consensus policies.

Spotlight Image
Exhibit 1

What is striking in Latin America's performance since 1990 is that compared with other parts of the world-in particular, compared to Asia and East Asia-Latin America has been doing quite poorly. Because of some recent growth, this may have somehow been shadowed, but it is still the case that if you look at the experience of Latin America since 1990, it is doing relatively poorly compared with countries in Asia. That's important because growth is a determinant of a lot of things-poverty reduction, as well as improvement in social indicators.

What is even more striking is not just that Latin America is doing poorly compared with Asia; it is that Latin America post-1990 did, in fact, worse compared to Latin America's own experience prior to 1980. The sense in which this is a tremendous paradox is that, of course, prior to 1980, Latin America had all those "terrible policies"-the macroeconomic populism, the protectionism, the import substitution, the infant industry promotion-all of those things that were supposed to be the roots of the trouble of the continent, the reason that those countries weren't going ahead sufficiently rapidly.

Yet it turns out that even after that mess was largely cleaned up and countries liberalized, stabilized, and privatized, and opened up themselves to the world economy like they had never been before-save, possibly, during the nineteenth century-these economies are still doing worse than under those bad old policies.

The Paradox of Success

The other side of the coin is that when you look at countries that have benefited the most from integration into the world economy, those are, it turns out, countries that follow highly nonstandard policies.

Here is the Heritage Foundation's Index of Economic Freedom (Exhibit 2). If you read the bottom, it essentially turns out to be a very close proxy for all the elements in the Washington Consensus. It covers things like how low your taxes and tariffs are, how little your economy is regulated, how little government intervention there is, and so forth.

Spotlight Image
Exhibit 2


Then you read World Bank documents and identify which are the countries that the World Bank calls its "star globalizers." Of course, that is not a big surprise there. The star globalizers are countries like China, Vietnam, India. These are countries which have experienced the most rapid growth in terms of volume of exports. These are countries that have experienced the most rapid increase in inward foreign investment, alongside very rapid economic growth as well. So they are the world's star economic globalizers.

The puzzle is how they have done it: It turns out, with policies that are very restrictive of economic freedom, compared to countries in Latin America, which are so much more in line with what the Heritage Foundation and conventional wisdom would identify as policies that are conducive to greater economic freedoms, greater market liberalization, and openness to international trade and investment.

So this is the paradox. The reason that I said development is working while development policy is not is that, on the one hand, we observe all these successes around the world-in China, in Vietnam, in India-with poverty reduction, and yet, as North American academics or as technocrats with multilateral institutions, when we go and talk to developing country governments, the kinds of policies that we actually advocate, the kinds of policies that we want them to have, are the policies that the evidence shows are not what, by and large, have produced both successful economic globalization on the part of individual countries and rapid economic growth.

Here is where I try to make some sense of this. I am going to present five points, very briefly.

(1) First is a bow in the direction of conventional wisdom, which is to say there are indeed these general principles of good policy. When people say that successful countries need to provide effective property rights, that they need to maintain macroeconomic stability, that they should try to integrate into the world economy, that they should ensure an appropriate environment for productive diversification and innovation, that they should provide effective regulation of financial intermediaries, and that they should maintain social cohesion and political stability, they are absolutely right. So at this level of broad general principles of what makes for economic success, these are absolutely essential.

You can go to each one of these individual countries and say that their success has, in some key ways, something to do with the speed with which they have moved towards the achievement of these objectives. So there are some general principles of good policy.

(2) These general principles actually do not map directly and uniquely into specific policies, into specific policy reforms, into specific policy agendas. Another way of saying this is that institutional function does not determine institutional design. You can define the useful functions that institutions need to achieve, but that doesn't lead you directly to a particular design, a particular blueprint, as to what that institution ought to look like. Each one of the previously listed goals can be achieved in a variety of ways. For instance, you can imagine achieving greater integration with world markets through a variety of policies, including subsidizing your exports, creating export processing zones, providing investment incentives to multinational enterprises, or simply through traditional, old-style import liberalization. Each one of these will get you more trade and more investment.

When you think about it that way, then I think you are led down a path where you start thinking about how local conditions and local situations can start to influence the particular way in which you design policies, where you take into account the second-best context in which you are actually working. When you are in the second-best context-and, by definition, these are poor countries, where markets and institutions and governments work poorly, so they are, by definition, in a second-best context-in those contexts, you need to apply the principles of second-best economics, which are a lot more complicated than your standard "just privatize, liberalize, stabilize" commandments. You need to take into account interactions with a preexisting set of market arrangements, preexisting constraints, and preexisting market distortions elsewhere.

That kind of program inherently leads you to rely much more on pragmatism and deep knowledge about local context, and to deemphasize what is very much in fashion once again, the role of "best practices" or rules of thumb of the Washington Consensus type.

It also requires a certain amount of policy experimentation, because you are not going to figure out what works locally until you start experimenting.

Another implication is that what you figure out will work in your own economy is not going to travel well, necessarily, because what works well in your own local economy is responding to a particular second-best context you find yourself in, in that economy, and doesn't necessarily correspond to the same situation elsewhere.

(3) Generating economic growth requires hitting the right targets and not doing everything at once. I think this is something that the Washington Consensus and subsequent ways of thinking about economic policy got badly wrong, thinking that there was a wide agenda of things that countries needed to do all at once.

An alternative and I think much more productive way of thinking about this is that the binding constraints on growth differ across countries and over time. To put it more colloquially, there are always different strokes for different folks. China at the outset was constrained by poor supply incentives in agriculture. Brazil is currently constrained by inadequate supply of credit, despite all the money that is coming in (but look at the real interest rates). El Salvador is constrained by inadequate production incentives in tradables; South Africa, by inadequate employment incentives in manufacturing; Zimbabwe today, by poor governance.

The point is that you get the biggest bang for reform when you hit the right target, when you target the binding constraint on growth. A lot of other things are at best unproductive, at worst potentially harmful, if they are not targeted on the direct constraint.

So a reform strategy should be selective. It will target these particular constraints rather than be a laundry list.

(4) Growth collapses occur when countries do not use high-growth periods to strengthen institutional underpinnings. I think two kinds of institutions in particular are important. One is institutions of conflict management, to enhance the resilience of the economy to external shocks. This is why the economy of sub-Saharan Africa collapsed in the 1970s and the Indonesian economy collapsed in 1997, and why China may still face extraordinary difficulties if it doesn't strengthen its institutions of democracy and rule of law.

The second kind of institutions that you need is those that promote productive diversification. You might call these by the dirty term "industrial policy," if you will. But all successful countries, in fact, have used industrial policy to promote productive diversification.

The important point is that deep institutional reform is hardly ever a prerequisite for economic growth. Good institutions sustain growth; they do not ignite growth.

(5) Finally, the role of globalization. I would argue that global institutional arrangements help, but not in the standard ways that you think about them. I like to distinguish between two kinds of cases. One is when what is actually on offer for individual countries is truly deep integration. I think about that as the European Union model, where what is on offer is true institutional harmonization, full labor mobility, full integration of capital markets, and a system of interregional transfers to deal with regional inequities. When you have that, you have really created the prospect of economic convergence, and in that case, the cost of external discipline can be outweighed by the benefits of market integration. For countries in the immediate periphery of the E.U., this is a wonderful model; it makes a lot of sense.

For most other countries, what is on offer is necessarily a shallow integration model. In those cases, external constraints and disciplines constrain precisely the kind of domestic policy experimentation that I was talking about before that is required for successful policy formulation that targets the binding constraints, because by the time you are in the WTO today, you are talking about restrictions on governments' abilities to subsidize enterprises, to impose disciplines on multinational enterprises, and to carry out a whole range of productive diversification policies.

What that means for most of the developing countries, those that do not have the luxury of having a European Union right next door to which they become a member, is that a set of thin international rules, as in the GATT system, is actually a lot better than the extensive intrusions that something like the WTO entails.

China's Success

When you look at a case like China, then it starts to make a lot more sense as to how China actually managed to do it. To put China in the context of those five principles that I enumerated:

First, China's reforms followed a strategic and sequential approach, targeting one binding constraint at a time. Reforms started in agriculture, then moved to industry, then to foreign trade, then to finance. Their reforms were characterized by pragmatism, and they were often heterodox policies that were targeted at overcoming political constraints and second-best complications. Each one of these policy innovations that the Chinese government used-two-track pricing, a household responsibility system, township and village enterprises, special economic zones-each one of them can be seen as domestic institutional innovations that overcame a particular second-best problem. These are innovations that Western economists would never have come up with on their own, and the Chinese leadership would not have come up with on their own either had it not been for experimentation and trying things out.

Of course, China did not join the World Trade Organization and did not submit to those kinds of disciplines until after its policy innovations had, in fact, resulted in a strong tradable sector. After all, if WTO disciplines were effective in China in the 1980s and early 1990s, when a lot of these policies were in place, there is no way China would have been able to follow the kinds of policies that it did, simply because those policies would have been illegal under the WTO.

As I mentioned before, though, there are remaining important institutional challenges that China faces, especially with regard to the building of political democracy and the rule of law.

Let me end by just putting this into the broad context of the kind of thinking that Albert Hirschman was engaged in, and then make a comment on that.

There has always been a tension between two different perspectives in thinking about development and development policy. The perspective which probably has always been the dominant approach to development policy over the years has been the one that takes the comprehensive approach. First, it was the big push in the 1950s, then it was the balanced growth of the 1960s-and, by the way, Albert Hirschman was a critic of both of those, the big push and balanced growth-and then, of course, it was the Washington Consensus. Then you had the second-generation reforms. Now you have the governance agenda at the World Bank and the United Nations, and you also have the "poverty trap" thesis and the Millennium Project of the United Nations.

The intellectual traditions behind these are very different, but what is common behind all of them is the idea that you just need a very wide and multifaceted, across-the-board approach to deal with problems of development.

The alternative is a much more strategic approach, which says:

  • Do the best with what you have, instead of wishing that you could transform yourself wholesale.
  • Identify priorities and opportunities and work off them.
  • Seek sequential and cumulative change over time, rather than comprehensive change at the outset.

  • Albert Hirschman was firmly, firmly in this camp. The way that he conceptualized and thought about development was as being opportunistic, as taking advantage of what you have. He called this "a passion for the possible"-a wonderful term. He believed that countries that have the capacity to undertake comprehensive programs would not be underdeveloped to begin with. This is a point that he kept on making, saying that if you really believe in these comprehensive reforms, you are just daydreaming. This is not how change happens in the real world.

    Dani Rodrik is professor of international political economy at Harvard's Kennedy School of Government. His most recent book is One Economics, Many Recipes: Globalization, Institutions, and Economic Growth (Princeton, 2007). This Audit is adapted from his acceptance of the Albert O. Hirschman Prize, presented on November 1, 2007, by the Social Science Research Council.

    Massachusetts Institute of Technology