Conditional Cash Transfers, Version 2.0

Although 75% of published evaluations on conditional cash transfer programs (CCT) are sourced from Latin America, newer CCT programs in Burkina Faso, Cambodia, Indonesia, Malawi, Pakistan, Tanzania and Nigeria are beginning to document, analyze and report results. This week, the World Bank brought a distinguished group together to discuss new findings, evaluation approaches and policy implications (see agenda here). Here are some highlights:

• Conditional cash transfers are winning the field… for now: Cash transfers are gaining increased popularity, and many studies are increasingly looking into issues of design and conditionality. The Malawi program showed that while CCT were more cost-effective than unconditional CT at increasing school enrollment, unconditional transfers proved more successful in reducing pregnancy and marriage rates – due to the fact that cash transfers were cut when the conditions were not met. This shows that unconditional cash transfers might be useful for obtaining other objectives.
• Don’t forget the supply-side: CCT programs are intended to create incentives for poor families to increase their investments in human capital, and most evaluations have focused on measuring these effects on the demand-side. However, less work has examined the factors that mediate the uptake and impact of CCT such as supply-side constraints, which are particularly binding on the health side. In Colombia, for example, supply-side issues were found to constrain CCT success: while designing programs and surveys, supply-side should be taken into consideration as well.
• Governance, systems and institutions matter: Beyond the technical capacity of education and health facilities, the overall institutional structure of countries matter. With the exception of Nicaragua, most rigorously evaluated CCT have been operating in MIC settings with relatively strong institutions; the importance of institutions will become ever more important as CCTs are slowly diffusing into low-income settings.
• Unintended results and spillover effects can be more significant than intended results: A very interesting result from the Cambodia program is that the transfer size did not matter for enrollment; giving $60 instead of $45 did not correspond to any significant increases in enrollment. Many other programs point to the relative inelasticity of enrollment on transfer size. Another spillover effect is on investment: in Mexico, Oportunidades increased consumption by 22% over the course of five years, due to the fact that households have reinvested their transfers. As beneficiaries used the transfer money to improve their housing and personal well-being, their health outcomes rose as well. These indirect, reduced-form effects have positive implications for both education and health.
• Poverty is multidimensional; CCT evaluations should be multidisciplinary: CCT have implications that go beyond increased human capital; such programs affect government behaviors (clientelism and rent-seeking), behavioral changes (academic success, decreases in crime, risky sexual behavior), and economic changes & investment habits.
• Transition into cost-benefit and cost effectiveness analysis: CCT research has tremendous policy implications, and the costs and benefits of these programs have to be analyzed more carefully – both in the design of the program and the evaluation of the outcome. This would inherently embody valuing the long-term impacts in human capital improvements due to health and education, as well as participation in labor markets – which would prove to be tricky given the supply-side issues on labor markets.

While many relevant aspects of CCT evaluations and design were addressed, a few were left out: Is it possible to change supplier behaviors through CCT? Is it possible to go beyond increases in enrollment and evaluate achievements instead? Finally, are CCT the best way to go for poverty reduction, or should the focus shift towards unconditional CT? These are questions yet to be answered through rigorous impact evaluations, and the workshop showed the potential of the field in generating evidence into policy.