evidoria

← Back to browse

Good practice Imported

Zomba Cash Transfer Programme — Conditional and Unconditional Transfers for Malawian Schoolgirls

Malawi · Zomba · See the Malawi profile

Evidence: Randomised controlled trial Top 45% 70/100 · Ask Evidence Copilot about this practice

A 2008–2009 randomised cash-transfer trial in Zomba, Malawi, paid school fees and stipends to 3,796 girls. Enrolment rose and early marriage/pregnancy fell among transfer recipients, but several effects, including marriage for in-school girls, faded once payments stopped.

45 pp
Increase in re-enrolment among baseline dropouts (CCT arm) (2008–2009)
12 pp
Reduction in marriage likelihood among baseline dropouts (CCT arm) (2008–2009)
8 pp
Increase in enrolment among baseline schoolgirls (CCT arm) (2008–2009)
8 pp
Reduction in marriage probability among baseline schoolgirls (UCT arm) (2008–2009)
7 pp
Reduction in pregnancy probability among baseline schoolgirls (UCT arm) (2008–2009)
0.6 years
Additional years of schooling, CCT dropouts (2 years post-transfer) (2011)
1.2 vs 3 %
HIV prevalence, transfer recipients vs comparison (during transfer period)
Zomba Cash Transfer Programme — Conditional and Unconditional Transfers for Malawian Schoolgirls

Details

Maturity
Discontinued
Promoter
World Bank / Abdul Latif Jameel Poverty Action Lab (J-PAL)
Period
2008–2009 (two-year randomised trial); follow-up surveys through 2011
Keywords
education, social protection, adolescent health, cash transfers

Context

Between February 2008 and December 2009, the World Bank and J-PAL ran a randomised controlled trial in Zomba district, Malawi, offering conditional cash transfers (CCT, paid only if the girl attended school) and unconditional cash transfers (UCT) to 3,796 unmarried girls and young women aged 13–22, split between girls still enrolled at baseline (2,907) and recent dropouts (889).

Objectives

The trial tested whether conditioning transfers on school attendance, versus giving them unconditionally, produced different effects on education, marriage, pregnancy and health outcomes.

Results

Among baseline dropouts, the CCT arm produced a 45-percentage-point increase in school re-enrolment and a 12-percentage-point reduction in the likelihood of marrying within the study period. Among baseline schoolgirls, the CCT arm raised enrolment by 8 percentage points, while the UCT arm reduced the probability of marriage by 8 points and of pregnancy by 7 points during the transfer period. Two years after transfers ended, CCT dropouts had completed 0.6 additional years of schooling on average, and HIV prevalence in transfer-recipient groups was roughly a third of the comparison group's rate (1.2% versus 3%) while payments were active.

Conclusions

Medium-term follow-up found that neither the CCT nor the UCT arm had a lasting effect on marriage among baseline schoolgirls once transfers stopped, the short-term HIV/HSV-2 gains did not persist, and no improvements in labour-market outcomes were detected. Time-limited transfers can shift schooling and health outcomes strongly while active, but do not guarantee durable change once the incentive is withdrawn.

Implementation
Indicative cost
Medium (€50k–€500k) — No published budget figures; band reflects two years of direct cash transfers to nearly 3,800 individuals plus a multi-round survey and research operation.
Time to results
Medium (1–3 years) — Two-year transfer period (2008–2009) with follow-up surveys through 2011.
Staffing & skills
Implemented by the World Bank and J-PAL research team, with local survey enumerators for baseline and follow-up data collection

Conditions for success

  • Conditioning transfers on school attendance (CCT) produced markedly larger re-enrolment effects among dropouts than unconditional transfers
  • Two treatment arms (conditional and unconditional) allowed effects to be isolated by transfer design

Common failure modes

  • Marriage-reduction effects among baseline schoolgirls did not persist once transfers stopped
  • Short-term HIV/HSV-2 gains did not persist in medium-term follow-up
  • No improvements in labour-market outcomes were detected

Where it fits

Governance type
donor-funded randomised trial implemented alongside government/school systems
Scale
district-level trial (Zomba district)

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful