View: Development economists who go small have a point

A few weeks ago, much of the economics world came together to celebrate the Nobel prize awarded to Abhijit Banerjee, Esther Duflo and Michael Kremer for “their experimental approach to alleviating global poverty.” That approach is the randomized controlled trial, in which economists try out an anti-poverty program — deworming treatments, for example — on some region or population and then measure the results against the people and places where the intervention wasn’t tried. In recent years, this approach has rapidly become more popular in the field of development economics.

But not everyone was happy. Some jumped to point out the limits of experiments as a research technique: the sample sizes are typically small, and it’s difficult and expensive to verify if small-bore programs can work at larger scales. Others worried that experiments are being overused at the expense of other approaches.

One of the most negative reactions came from development economist Lant Pritchett of Harvard University’s Kennedy School of Government. In a ferocious Facebook post, he accused Banerjee, Duflo and Kremer of actively hurting the study of global poverty.

Pritchett reasons that the only thing that really alleviates a nation’s poverty on a significant scale is economic growth:

Changes in poverty…are overwhelming (sic) associated with growth in median income/consumption. We have seen…poverty go from high levels to low levels right in…China, Vietnam, Indonesia…Variation in the size and efficacy of poverty programs had little or nothing to do with poverty reduction.

By committing to the use of experiments, Pritchett argues, this year’s Nobel winners precluded themselves from addressing the important question of how poor countries become rich countries. The implication is that everyone else who uses experiments is limiting themselves in a similar manner. Pritchett’s post thus seems to be a warning to development economists: focus on the big questions instead of the small easy ones or become irrelevant.

Pritchett certainly has a point about growth. A quick glance at any map of global extreme poverty will reveal that the countries where poverty remains high are those where per capita gross domestic product is low — many African countries and so on. And the countries where poverty has fallen substantially are just those that have experienced rapid growth, such as China, India and a number of countries in Southeast Asia. The historical experience of rich nations bears this out; as their economies grew, poverty rates fell.

If there were some government policy that could transform a poor country into a rich one, it would be like a magic wand. Compared with that wand, things like deworming, microcredit and all the other humble programs that development economists tinker with would be almost irrelevant. To some economists, this means that they should ignore the small stuff and focus on finding that wand. Nobel-winning macroeconomist Robert Lucas once admitted that after he started thinking about the question, it became “hard to think of anything else.”

Unfortunately, Lucas never found that magic wand. Neither has anybody else. Lucas made some very simple models of economic growth in which human capital was the key. But beyond universal public education and health care — things every government and development agency already knows are good — it’s just not clear how to generate big increases in human capital. Other growth models tend to suffer from similar problems; they end up simply applying labels like “technology” and “productivity” to the mysterious magical stuff that causes poor countries to turn into rich ones.

Development economists, meanwhile, have tried to create their own magic wands, in the form of big foreign aid programs. Economist Jeffrey Sachs, for example, spent years implementing his Millennium Villages Project starting in 2006 but achieved few positive results. Sachs and his fellow economist William Easterly had a vigorous and long debate about the effectiveness of foreign aid, and there are partisans on both sides. Yet the fact remains that foreign aid has lifted few if any countries out of poverty so far. Meanwhile, economists’ recommendations of economic liberalization in the 1990s, commonly known as the Washington Consensus, also delivered underwhelming results.

Some countries have succeeded in going from rags to riches, such as South Korea and Taiwan. Others, such as Malaysia and Turkey, are on the cusp of completing their transition. A number of economists and writers believe that export-oriented industrial policy was the key to these countries’ success, but this has yet to translate into a simple formula that any poor country can adopt.

Then there’s the possibility that there simply is no magic wand. According to some economic theories, global development spreads naturally from one region to another, and all that poor countries can do is wait and hope their turn comes.

So although Pritchett has a point that growth is crucial, that doesn’t imply that every development economist should spend their careers in pursuit of big transformational growth policies that may not even exist. It makes sense for some economists to focus on the small things that are clearly within their power to change, such as helping kids in poor countries get a little healthier and a little better educated. To denigrate these humble economic plumbers, to use Duflo’s term, is an act of hubris. Instead, economists who insist on addressing only the biggest, hardest questions should reflect on their failure to find answers.