Michael Clemens on Aid, Migration, and Poverty

Episode Summary

Executive Summary: Russ Roberts and Michael Clemens compare foreign aid and migration as tools for reducing global poverty. Clemens argues aid’s growth effects are hard to isolate and usually modest, while migration—by moving workers from low- to high-productivity locations—creates vastly larger gains, though political barriers remain high and empirical effects on natives are generally small or offset by broader economic adjustments.

Main Topics: Foreign aid and the difficulty of measuring growth effects (Priority: 5/5): Clemens explains that aid studies struggle with causality, timing, noisy GDP data, and crude aid measures. He argues many aid flows are not intended to raise growth, making broad regression studies especially hard to interpret. Reconciling conflicting aid-growth studies (Priority: 5/5): He describes his paper 'Counting Chickens When They Hatch,' which harmonized data, time periods, and assumptions across influential studies. After adjusting for timing and sample differences, the apparent conflict in the literature largely shrank, yielding only a modest positive effect. Aid types vary widely in economic relevance (Priority: 4/5): The discussion distinguishes humanitarian aid, health campaigns, education, roads, and institutional training. Some aid clearly helps welfare without necessarily raising growth quickly, while other aid is unlikely to affect GDP in the short run. Migration as a far larger anti-poverty force than aid (Priority: 5/5): Clemens argues that labor mobility creates enormous welfare gains because workers can earn far more in richer countries. He says remittances alone dwarf aid flows, and total gains from freer movement of people could reach trillions. Effects of low-skill immigration on native workers and the economy (Priority: 5/5): Roberts and Clemens debate whether low-skill immigrants harm native low-skill workers. Clemens argues the labor market is more complex than simple substitution: immigrants complement natives, lower prices, raise labor-force participation, and can expand demand for native labor. Policy design, labor-market knowledge, and 'open borders' (Priority: 4/5): Clemens questions vague calls for or against open borders and criticizes labor-market tests and quotas as inefficient. He suggests employers and sector-specific actors often know labor needs better than central planners.

Key Arguments: Foreign aid is very difficult to evaluate because poor countries receive aid when they are poor or after shocks, GDP data are noisy, and the timing of aid’s effects is uncertain. Many aid projects are not designed to raise growth at all; humanitarian, health, and democracy aid should not be judged primarily on GDP impact. By reconstructing and harmonizing earlier datasets, Clemens and coauthors found less disagreement in the aid-growth literature than previously claimed. The estimated growth effect of aid is positive but modest, and does not support viewing aid as a major growth strategy. Migration yields much larger welfare gains than aid because workers can earn several times more in richer countries while doing the same work. Aggregate migration benefits are not captured by wage effects alone; remittances, productivity spillovers, and sectoral adjustments matter. Simple models predicting that more immigrants necessarily depress native wages ignore complementarity, specialization, induced demand, and price effects. Empirical studies of low-skill immigration generally find small adverse wage effects on some native workers, but these effects are limited in size and often offset by broader gains. Immigrants can expand sectors like agriculture, caregiving, and other nontradable services, lowering costs for everyone and increasing labor-force participation. Policy should be shaped by actual labor demand and local knowledge rather than rigid caps and centralized labor-market tests.

Data Points: Annual foreign aid: a little over $120 billion - Clemens uses this to argue aid is small relative to migration flows. Annual remittances sent home by migrants: about $400–500 billion - He says remittances are roughly four times foreign aid. Aid effect estimate: 10 percentage point increase in aid/GDP associated with about 1 percentage point higher growth per year - Result reported from Clemens’s harmonized aid-growth analysis in subsequent periods. Aid truncation threshold in Boone paper: 15% of GDP in aid - Boone excluded observations above this level, which Clemens disputes. U.S. low-skill job growth projection: about 3 million more jobs over a decade - Based on BLS projections for occupations not requiring a high school degree. Projected increase in U.S. working-age labor force: 1.7 million - Clemens contrasts this with projected demand for low-skill work. Place premium for marginal migrant: typically 300% of home-country real income; over 1,000% in some cases - From Clemens, Lant Pritchett, and Claudio Montenegro’s estimate of the gain to the next migrant. U.S. population change example: about 75 million in 1905 to over 300 million in 2005 - Used to question a simple labor-demand-only model of migration effects. Unemployment rate example: about 5% in both 1905 and 2005 - Clemens uses this to argue population growth did not mechanically raise unemployment. White foreign-born adults with high school degree in 1940: 12% - Used to highlight that many immigrant ancestors were low-skill by today’s standards. Borjas low-skill wage effect: about 8% for high school dropouts; about 3% for the average American record - Russ Roberts cites Borjas’s estimate as small in magnitude.

Pivotal Quotes: "foreign aid is a little over $120 billion a year... the portion of migrants' earnings that they send home is about four times that" — Michael Clemens: Used to argue that migration-related flows are much larger than aid. "We're talking about very modest effects... foreign aid per se can be anything resembling a growth strategy per se for any developing country" — Michael Clemens: Clemens summarizes his view of the aid-growth literature. "the labor demand curve is downward sloping, with is in italics" — Russ Roberts: Roberts paraphrases a core labor-economics argument against immigration.

Implications: For poverty reduction, migration appears far more powerful than aid, though politically constrained. For policy, the evidence favors more flexible labor mobility and skepticism toward rigid visa caps, while recognizing that aid and immigration debates should be judged by broader welfare effects, not just narrow GDP or wage studies.