Cash Transfers and the Evidence

A re-analysis of seven government cash-transfer RCTs across six countries finds no systematic evidence that transfers reduce work effort among recipients. The single most useful citation when a finance ministry argues that cash will make people idle — and the cleanest entry point to the whole literature.

The book that made the conditional-cash-transfer model travel from Latin America to the rest of the world. It lays out the theory of conditions — using transfers to correct under-investment in children's health and schooling — and the early evidence base. Dated in places, but still the canonical statement of the CCT case.

The successor to the long-running State of Social Safety Nets series, and the best single snapshot of where cash and other transfers reach today. Coverage has expanded to 4.7 billion people, yet 2 billion remain uncovered or under-covered — the scale problem that all the micro-evidence eventually runs into.

The flagship GiveDirectly evaluation, randomising transfer size, timing, and recipient gender. Large unconditional transfers raised consumption, assets, food security, and psychological wellbeing — and did not increase spending on alcohol or tobacco, contradicting a durable stereotype. The design also teaches you how to learn from variation within a programme, not just its average effect.

The accessible front door to the 12-year Kenya basic-income experiment across roughly 195 villages — comparing long-term monthly payments, short-term monthly payments, and lump sums. Useful in the classroom because it forces the question of design: a lump sum and a monthly stipend of the same total value produce strikingly different behaviour.

The research paper behind the GiveDirectly UBI headlines, exploiting the fact that the experiment was already running when COVID-19 hit. Recipients of an ongoing income were more food-secure and less likely to fall sick, and the data speak directly to the insurance value of regular cash against aggregate shocks — a question short experiments cannot answer.

The reference systematic review on the conditionality question for education. Both CCTs and UCTs raise enrolment and attendance; conditions add a meaningful edge for schooling specifically, while the effect on test scores is small at best. The honest answer to "should we attach conditions?" — it depends what outcome you actually care about.

The definitive evaluation of the "graduation" approach — an asset, consumption support, training, coaching, and savings, bundled. Across Ethiopia, Ghana, Honduras, India, Pakistan, and Peru it produced lasting gains a year after support ended, with benefits exceeding costs in five of six sites. The strongest case that for the ultra-poor, cash alone may not be enough.

A Mexican experiment that randomised villages into food boxes, equivalent cash, or nothing — and then watched local prices. In-kind transfers, by adding supply, pushed prices down about 4%, while cash barely moved them. A sharp reminder that the cash-versus-kind choice is a general-equilibrium question, not just a question of recipient preference.

A 19-million-person randomised rollout of biometric Smartcards for NREGS wages and pensions in Andhra Pradesh. Payments became faster, more predictable, and less leaky — without cutting genuine beneficiaries out. The foundational evidence that how you pay can matter as much as what you pay.

The cautionary sequel. Tightening Aadhaar-based identity requirements cut some corruption but knocked an estimated 1.5–2 million legitimate beneficiaries out of benefits, driven largely by how the transition was handled. Essential reading for anyone tempted to treat DBT digitisation as an unalloyed good.

The most-cited Indian intervention in the cash-versus-PDS debate, grounded in field surveys of what recipients actually prefer. Where the Public Distribution System works, people want food, not cash; preferences track the quality of the status quo. A direct counter to the assumption that cash is always the more dignified, efficient choice.

The living hub for the cash-transfer evidence base: policy insights that synthesise across dozens of randomised evaluations on transfer size, targeting, conditionality, and delivery. The place to send a policymaker who wants the current state of knowledge rather than a single study.