What this covers

Michael Clemens of the Center for Global Development speaks with Russ Roberts about the comparative effects of foreign aid and migration on poverty reduction. Clemens surveys the empirical evidence on aid's growth effects, drawing on meta-analysis of the major cross-country studies—including Burnside-Dollar, Rajan-Subramanian, and Boone—to show how sensitive their conflicting results are to arbitrary data choices. He then pivots to migration, arguing that the wage gap for identical work across borders creates a vastly larger poverty-reduction opportunity than aid can offer. The conversation moves through the feared labor-market effects of low-skill immigration in receiving countries and concludes with the question of whether migration policy itself can be coherently designed.

The argument proceeds through several layers of economic evidence. Clemens establishes that aid produces only modest growth—roughly 1 percentage point of additional annual growth per 10 percentage points of aid-to-GDP—an effect that diminishes as aid levels rise, while remittances from migrants dwarf total aid flows several times over. The place premium quantifies the labor arbitrage opportunity: the same worker earns 300–1000% more in real terms doing identical work in a wealthy country, implying tens to hundreds of trillions in unrealized welfare gains. On the displacement question, Clemens draws on labor market segmentation theory and historical examples—apartheid's end, century-long US population quadrupling, historical low-skill migration patterns—to argue that native wage losses are marginal and offset by demand shifts through task specialization, complementarity, price effects, and sector creation. The discussion turns contentious on methodology: how to attribute causation in noisy data, whether growth regressions measure what they claim, and whether work-visa caps are coherently designed policy or Pareto-improving destruction.

Sharpest takeaway

Clemens argues that foreign aid has positive but modest effects on growth, while removing barriers to migration represents a vastly larger, first-order opportunity to reduce global poverty because labor earns radically different wages for the same work across borders and the feared wage-displacement effects on natives are tiny and offset by demand-shifting forces.

  • Aid produces only modest growth effects (roughly 1 percentage point of growth per 10 points of aid/GDP) and diminishes as aid rises, while migrant remittances alone dwarf aid flows several times over.
  • The 'place premium' shows the same worker earns 300-1000% more doing identical work in a rich country, implying tens to hundreds of trillions in potential welfare gains.
  • Low-skill immigration depresses native wages only marginally (single-digit percent over decades) because labor markets are segmented and immigrants shift out labor demand through complementarity, specialization, and sector creation.

The argument · threads2 threads · 30 claims
0.85

Labor markets adjust to immigration restrictions through wage increases, mechanization, or reduced output, not collapse.

1 pointcentrality 5/5
  • If immigrants were not allowed in, work like crop-picking would still get done—by higher-paid workers, by machines, or with less output—because the economy adjusts to labor supply restrictions much as it would to a plague or other exogenous shock; the 'we wouldn't get our food picked' argument is therefore not a compelling case for immigration.

    “We'd get it picked; it would just be by people who were now paid a lot more. There might be less of it picked; it might be picked by machines now.”

0.85

Education policy, not immigration exclusion, should address wage harm low-skill workers face.

1 pointcentrality 5/5
  • Since the modest wage harm from low-skill immigration falls mostly on high school dropouts, the appropriate policy response is to keep people in high school rather than to exclude immigrants—and propping up the wages of those who do not finish high school may itself be antithetical to the goal of encouraging high school completion.

    “instead of keeping out immigrants, we ought to try to keep people in high school. That that would be the appropriate policy response to that empirical finding.”