What reduces poverty — and what doesn't.

Direct cash transfers produce broad, durable improvements in wellbeing without evidence of waste or dependency.

GiveDirectly's Kenya experiments — including a long-run universal basic income trial — found that large unconditional cash transfers increased assets, consumption, and psychological wellbeing, with no evidence of increased alcohol use or reduced labor supply. The BRAC graduation program in Bangladesh and similar programs across six countries confirmed that transfers to the ultra-poor produce lasting income gains years after transfers end, provided they also include livelihoods support. The accumulated evidence from Mexico's Progresa, Brazil's Bolsa Família, and direct transfer programs in 40+ countries consistently finds positive effects on education, health, and consumption.

Basic investments in health — deworming, bednets, oral rehydration — produce the highest returns per dollar of any development intervention.

Kenya's bednet distribution experiments found near-universal adoption and significant reductions in malaria when nets were provided free; subsidized nets reached far fewer households. The deworming Kenya study found that school-based deworming produced long-run labor market gains at very low cost — effects detectable 10+ years later. The Kenya WASH Benefits study found that water treatment dramatically reduced diarrheal disease in children under 2. These health investments show high returns because they address bottlenecks that prevent all other human capital development.

Microfinance improves financial access but does not reliably generate business growth or poverty exit.

Six simultaneous randomized evaluations of microcredit programs across Bosnia, Ethiopia, India, Mexico, Mongolia, and Morocco (published in the American Economic Journal: Applied Economics, 2015) found consistently modest effects. Business investment and profits increased for some households with existing businesses; consumption improved slightly for some. But the hoped-for transformation of the poor into entrepreneurs through credit did not materialize at scale. Compartamos Banco in Mexico showed positive take-up of credit products but no transformative income effects.

Impact of Microcredit — Mexico

Rural Mexico · 2015

Expanding credit access to poor households improves financial flexibility but does not reliably generate business growth or poverty exit. The credit constraint hypothesis — that lack of credit is the binding constraint on poor households — appears weaker than the broader poverty trap hypothesis.