Cash transfers

To condition or not to condition?

Cash transfers are an increasingly popular social protection mechanism throughout Latin America, where conditional cash transfers are dominant, and sub-Saharan Africa, where unconditional cash transfers are more common. How can this difference in approaches be explained, and what evidence exists on their relative effectiveness?

At the London Summit in April 2009, the G-20 countries pledged to mitigate the social impacts of the global financial crisis, by providing “US$50 billion to support social protection, boost trade and safeguard development in low income countries” (in the Global Plan for Recovery and Reform: the Communiqué from the London Summit). Later in the same month, the World Bank announced that its lending for social protection programmes will rise from US$4 billion to US$12 billion by 2012. Their press release said, “This lending includes rapid social response programs and conditional cash transfers, where families are granted money transfers in exchange for sending their children to school and for regular medical check-ups”.

In response to the food price crisis of 2008, the World Bank established a US$2 billion Global Food Crisis Response Program (GFRP), which supports social protection interventions such as food- or cash-for-work schemes and school feeding programmes. Following their July 2009 meeting in L'Aquila, Italy, the G-8 countries issued a Joint Statement on Global Food Security, which urged support for a range of social protection measures - including cash-based social protection systems - to assist millions of people affected by the global food crisis.

Initiatives such as these represent a massive escalation of social protection, already a highly fashionable development agenda. But conditional cash transfers, public works and school feeding programmes are three highly controversial social protection instruments. This issue of insights examines the case for and against the first of these – conditional cash transfers.

Unconditional cash transfers are given to poor and vulnerable people with no restrictions on how the cash is spent, and no requirements beyond meeting the eligibility criteria (for example, being poor, an orphan, or over 60 years of age). The primary objective is to protect current consumption or food security. By contrast, conditional cash transfers (CCTs) are delivered only on condition that recipients meet certain requirements, such as that their children should be enrolled in and attending school, and must be immunised. These programmes aim not only to alleviate current poverty through income transfers, but also to reduce future poverty by encouraging investments in human capital – education, health and nutrition.

Each variation of cash transfers is strikingly geographically concentrated. Large-scale national CCT programmes are currently operational in most countries of Central and South America. Conversely, although numerous small scale pilot projects and some national cash transfer programmes are found throughout sub-Saharan Africa, very few of these are conditional.

The debate over whether cash transfers should be conditional or unconditional is partly empirical and partly political. The empirical case for CCTs is based on evaluations in several Latin American countries, confirming that they deliver both wellbeing benefits to recipient households and improved education and health outcomes for children in these households (see John Hoddinott). CCTs also achieve significant impacts on poverty reduction, especially poverty gap and poverty severity measures (see Emmanuel Skoufias and Vincenzo di Maro). The empirical argument for unconditional transfers is based on evidence from several African countries, confirming that recipients invest some of their cash transfers in education and health anyway, so there is no need to compel them to do so (see the article on the Basic Income Grant and the piece by Michael Samson). The political argument for CCTs is that domestically financed social protection requires buy-in from the taxpaying middle classes, who typically object to ‘welfare handouts’ and prefer to support programmes that ensure poor children are educated and receive adequate health care. The political or ethical argument against CCTs is that conditionalities are paternalistic and interfere with the people’s right to choose how they allocate their resources.

There are other arguments against conditionalities. Linking social transfers directly to public services requires well-functioning services - but especially in rural Africa, education and health facilities are often weak, or even non-existent. This is not an argument against CCTs in principle, but it is an argument for prior investment in the supply of services, before stimulating demand with conditionalities. Also, there are complaints that the burden of adhering to conditionalities falls disproportionately on women, since it is mainly mothers and other female carers who take children to clinics and ensure they go to school (see Sarah Bradshaw). Unconditional cash transfers do not add to pressures on women, but they can contribute to intra-household tensions, unless the programme is carefully designed (see Rachel Slater and Matseliso Mphale).

Finally, the global food crisis highlighted a risk that is faced by all cash transfer programmes, whether conditional or unconditional – that their purchasing power will be undermined by inflation. In some cases this has undermined the popularity of cash transfers and led to a resurgence in beneficiary preferences for food aid (see Rachel Sabates-Wheeler).

This issue of insights asks three questions about CCTs:

• Do CCTs reduce poverty?
• Do cash transfers discourage work?
• If conditionalities do work, do women pay the price?

…and four questions about unconditional cash transfers:

• Is a universal Basic Income Grant feasible?
• Can cash transfers reduce intergenerational poverty?
• Can cash transfers improve gender relations?
• Are cash transfers susceptible to high food prices?

Stephen Devereux
Centre for Social Protection
Institute of Development Studies
University of Sussex, Falmer
Brighton BN1 9RE, UK
s.devereux@ids.ac.uk