We all know what the benefits of open international capital markets are. Just like free trade in other goods, capital account liberalization means money flows where it’s needed, which makes for economic growth and development. We know this not just because of some verbal arguments, but because economists can write down a model of this process, and we can see it happening in the data. Right?

Wrong. A fascinating paper [pdf] by Pierre-Olivier Gourinchas and Olivier Jeanne shows that this isn’t the case at all. (This blog isn’t the New York Times, but in the spirit of full disclosure I should say that Pierre is a friend of mine.) They find that, for most countries, the average welfare gain from financial integration is only 1.24% of current consumption. This is tiny compared to the potential gains from increased productivity growth or a reorganization of the domestic market. Rather than helping them converge with developed economies, the best that capital mobility can do is to accelerate developing countries towards some steady-state (good, bad or very bad) largely determined by other factors.

What makes the paper particularly interesting and (I think) important, though, is that the authors don’t just do a bunch of cross-national time-series regressions to estimate the relative effect of some capital liberalization variable. That’s a worthwhile approach, but a significant result would not unduly upset most economists. This paper, by contrast, takes the standard model used to explain why financial liberalization is a good thing, calibrates it with the data, and shows that it fails on its own terms. That’s a much more serious challenge to the prevailing wisdom. The authors conclude that, based on this calibrated neoclassical growth model,

less developed countries do not benefit greatly from international financial integration. Less developed countries have far more to gain from improving their own domestic allocative efficiency than from an improvement in the allocative efficiency of the international financial system.

Pierre and Olivier are of course well-aware of ways in which financial liberalization might benefit developing countries, over and above the elusive gains from increased allocative efficiency at the international level. For instance, it might help developing countries if foreign banks (who know what they are doing) were allowed in; or liberalization might matter because it signals that the country is serious about protecting property rights, and so on. But these are not empirically settled questions and, as the authors say themselves, these “channels are not captured by the basic neoclassical framework.” And it’s that framework that’s supposed to give the original argument its real force.

Fascinating stuff. Read the paper yourself.