# The cash-transfer evidence base
Primary-source notes on whether giving people money actually works — and on
why the strongest form of that evidence is the systematic review, not any
single trial.
## Why the systematic review is the right unit of evidence
Any single randomized controlled trial (RCT) of a cash-transfer programme can
be a fluke of context, season, or design. The defensible claim is built from
*synthesis*: dozens of independent RCTs across many countries pointing the
same direction. The anchor for this site is therefore a rigorous review, not a
hero study.
**Bastagli, F., Hagen-Zanker, J., Harman, L., Barca, V., Sturge, G. &
Schmidt, T. (2016). "Cash transfers: what does the evidence say? A rigorous
review of programme impact and of the role of design and implementation
features." Overseas Development Institute (ODI), London.** The review screened
the global evidence from 2000–2015 and analysed 165 studies covering 56
cash-transfer programmes across low- and middle-income countries, organising
findings across six outcome areas (monetary poverty; education; health and
nutrition; savings, investment and production; employment; and empowerment).
The pattern across that body of evidence: cash transfers consistently and
significantly **reduced monetary poverty** (total expenditure and food
expenditure rose; poverty headcount and gap fell), **increased school
attendance**, **increased the use of health services and dietary diversity**,
and **increased savings and productive investment** (livestock, agricultural
inputs, business assets). The effects vary in size with design — transfer
size, predictability, payment regularity, and targeting all moderate impact —
which is the review's central methodological point: design features, not the
existence of a transfer alone, drive results.
## The "do transfers make people lazy?" question
This is the single most common objection, and it is the one with the cleanest
answer.
**Banerjee, A.V., Hanna, R., Kreindler, G.E. & Olken, B.A. (2017).
"Debunking the Stereotype of the Lazy Welfare Recipient: Evidence from Cash
Transfer Programs." *The World Bank Research Observer* 32(2):155–184.** The
authors re-analysed the data from seven randomized controlled trials of
government-run cash-transfer programmes across six countries (Honduras,
Indonesia, Morocco, Mexico, the Philippines, and Nicaragua) using a common
specification. They found **no systematic evidence that cash transfers
discourage work** — neither hours worked nor labour-force participation fell
systematically among adults of working age. The "lazy recipient" prior is not
supported by the pooled experimental record.
## Conditional vs unconditional, and the cost of conditions
A persistent policy debate is whether transfers must be *conditional* (tied to
school attendance, clinic visits) to work. The cleanest single experiment on
this is:
**Baird, S., McIntosh, C. & Özler, B. (2011). "Cash or Condition? Evidence
from a Cash Transfer Experiment." *The Quarterly Journal of Economics*
126(4):1709–1753.** In a randomized programme for adolescent girls in Malawi
with both a conditional (CCT) and an unconditional (UCT) arm, the schooling
effect of the unconditional arm was real but smaller — roughly 43% as large as
the conditional arm's effect on dropout. But the condition carried a cost:
teenage pregnancy and early marriage rates were **substantially higher in the
conditional arm than the unconditional arm**, because girls who fell out of
the conditional programme lost support entirely. Conditions can sharpen a
targeted outcome while harming others — a genuine trade-off, not a free win.
## A benchmark unconditional trial
**Haushofer, J. & Shapiro, J. (2016). "The Short-Term Impact of Unconditional
Cash Transfers to the Poor: Experimental Evidence from Kenya." *The Quarterly
Journal of Economics* 131(4):1973–2042.** A village- and household-randomized
trial of unconditional transfers from the NGO GiveDirectly to poor rural
Kenyan households (transfer arms of roughly US$404 and US$1,525 at PPP) found
increases in consumption, asset holdings, food security, and psychological
wellbeing, with no significant increase in spending on alcohol or tobacco — a
direct rebuttal of the "they'll just waste it" objection. (Note: an erratum,
*QJE* 132(4):2057–2060, 2017, corrected multiple-hypothesis-adjusted p-values;
the main results survive.)
## What needs a caveat
**1. "Reduces work" is not uniformly zero — the population matters.** The
Banerjee et al. result is about prime-age adults in targeted poverty
programmes. Universal or near-universal transfers in high-income settings, and
transfers to the elderly or to households near retirement, can show modest
labour-supply reductions. The honest claim is "no systematic disincentive in
the studied programmes," not "labour supply is unaffected in every design."
**2. External validity and duration.** Most strong RCTs measure short- to
medium-term effects. Follow-ups of the Malawi cohort (Baird, McIntosh & Özler,
"When the Money Runs Out," *Journal of Development Economics* 140:169–185,
2019) found that several gains faded after transfers stopped — cash relieves
the constraint while it flows; it is not automatically a permanent escalator.
**3. General-equilibrium and inflation effects.** RCTs randomize *within* a
population and can miss economy-wide effects. If everyone receives transfers
at once, local prices can rise, eroding real value (see the Iran case
discussed in `counterarguments.md`). Egger et al. (2022), "General Equilibrium
Effects of Cash Transfers," *Econometrica* 90(6):2603–2643, found large
positive local multipliers in Kenya with only modest inflation — but that is
one setting, not a universal guarantee.
**4. Publication and programme-selection.** Reviews can over-weight
well-implemented, well-studied programmes. Bastagli et al. are explicit that
implementation quality (payment reliability, low transaction costs) is a
precondition; a badly run transfer system can fail even where the instrument
works in principle.
The defensible summary: across a large, multi-country experimental record,
direct cash reliably reduces poverty and increases human-capital and
productive investment, does not systematically reduce work, and is not
typically squandered — with effect sizes that depend heavily on design, that
can fade once transfers end, and that must be checked for economy-wide price
effects at scale.
## Sources
- Bastagli, F. et al. (2016). "Cash transfers: what does the evidence say?"
ODI.
(full review PDF: )
- Banerjee, A.V., Hanna, R., Kreindler, G.E. & Olken, B.A. (2017). "Debunking
the Stereotype of the Lazy Welfare Recipient." *World Bank Research Observer*
32(2):155–184.
- Baird, S., McIntosh, C. & Özler, B. (2011). "Cash or Condition? Evidence
from a Cash Transfer Experiment." *QJE* 126(4):1709–1753.
- Haushofer, J. & Shapiro, J. (2016). "The Short-Term Impact of Unconditional
Cash Transfers to the Poor: Experimental Evidence from Kenya." *QJE*
131(4):1973–2042.
- Egger, D., Haushofer, J., Miguel, E., Niehaus, P. & Walker, M. (2022).
"General Equilibrium Effects of Cash Transfers: Experimental Evidence from
Kenya." *Econometrica* 90(6):2603–2643.