Analysis: Unconditional money
Free money works
JOHANNESBURG, 2 August 2010 (IRIN) - A World Bank study that put money in the hands of girls and their parents in Malawi's poor southern region has caused ripples across the debate on cash transfers in academic and aid communities.
"Cash transfers enhance access to education whether or not they are conditional, so the determining factor is the income effect, not the conditionality," said development economist Stephen Devereux.
An average of US$10 per month was handed out to a group of girls participating in the study on condition that they attended school 80 percent of the time so to remain eligible for the stipend.
It made no difference to the outcome when no conditions were imposed - the girls participating in the study attended school 80 percent of the time, and in both cases the school drop-out rate also fell by 40 percent. The school attendance findings were among others
The debate on whether vulnerable people should be given cash as a social safety net without any conditions attached has run for decades, said Anna McCord of the Overseas Development Institute (ODI), a UK-based think-tank, and John Hoddinott of the International Food Policy Research Institute (IFPRI), who have both studied cash transfer programmes at length.
Nicholas Freeland, who heads the Regional Hunger and Vulnerability Programme in Southern Africa, said attaching conditions made cash transfers more complex and expensive because they had to be monitored and enforced; there was also the moral issue of the donor or government deciding on how people should behave.
Much of the enthusiasm for conditional cash transfers (CCT) has been built on the success of such programmes in Latin America, which is quite different to Africa, said Sheila Sisulu, deputy director of the World Food Programme (WFP). "A CCT in Latin America can ... [give money] to mothers on condition that she goes to the clinic, but many parts of Africa are not well-serviced - there may be no clinic or school for miles around."
Does the Malawi study in the poverty-stricken Zomba district add anything to the debate? ODI's McCord admitted that the findings were significant but did not think they resolved the question.
"This is a strong and important finding, both because it challenges the rationale for conditional cash transfers, and because it comes from the World Bank, which has been the strongest advocate of conditional cash transfers in the international community," Devereux said.
The Bank backs CCT programmes in 13 countries, and provided $2.4 billion to such initiatives during the global economic crisis in the 2009 financial year, according to a Bank statement.
|And cash transfers have been around
|From Philip White, co-author of the book, Social Protection in Africa as told to IRIN.
"The oldest cash transfers I know about would be the English Poor Laws which go back to Elizabethan times, when parishes were charged with raising special taxes to provide for their native poor. Such provision included cash as well as in-kind relief for the able-bodied poor for whom work could not be found or those unable to work through age or sickness.
In the South, large-scale cash transfers were initiated during the 1990s in South Africa, Mexico, Brazil, India, Indonesia and China. South Africa had in fact brought in non-contributory pensions for 'White' and 'Coloured' citizens in 1928, extended to 'Natives' in 1944 (1973 in Namibia), albeit at steeply differentiated rates according to racial categories until 1994. To cite just a few other examples, Botswana's non-contributory pension began in 1996, Lesotho's in 2004. South Africa's Child Support Grant began in 1998. India's Employment Guarantee Scheme began in Maharashtra in 1979 and went national in 2006. Brazil's Bolsa Escolar started in 1995 and went national in 2001. Mexico's Progresa programme began in 1997. And so on...
There was a "limited availability" of "rigorous evidence to measure the effects of 'conditions' separate from those of the cash transfer", noted Ariel Fizbein, the World Bank's leading authority on social protection and a global expert in the field, who maintained that the Bank did not have a policy on the issue and had supported both CCTs and unconditional cash transfers (UCTs).
Freeland commented, "The point is that the Zomba experiment represents the only 'ideal' experiment to date to determine whether it is the cash or the condition that causes the impact."
He quoted the World Bank's "Bible on CCTs" - Conditional Cash Transfers: reducing Present and Future Poverty - co-authored by Fizbein: "Ideally, to disentangle the effect of conditions from the income effect inherent in the transfer an experiment would be designed whereby a first group of households or villages receives a UCT, a second group receives a CCT, and a third group serves as a control group."
The Malawi study did just this. The researchers selected three groups of girls, the first of which were given stipends on condition that they attended school 80 percent of the time; the second were given the same stipends unconditionally; the third received nothing and served as the comparison group for the study.
The study was led by researchers Berk Özler, a senior economist at the World Bank's Development Research Group, Sarah Baird of the George Washington University, in Washington, and Craig McIntosh of the University of California, San Diego.
The World Bank noted in a statement that the findings were "in contrast to the conditional cash transfer (CCT) experience in Latin America, where the condition to attend school has been key to the programme's success".
Devereux agreed with Freeland, pointing out that "We do not have 'experiments' in Latin America, with control groups of conditional and unconditional cash transfer beneficiaries, so in that sense the Malawi case study is unique."
Fizbein was cautious in his endorsement. "The Malawi study is ... a welcome addition to this literature. In my view ... it is problematic to seek a resolution to the complex issue of human capital conditions based on a narrow interpretation of the Malawi study. The effects of conditions may vary, depending on a range of factors (target group, level of payment among others)," he said.
"My recommendation to Bank teams and their government counterparts remains the same: consider with care the objectives being pursued by the particular programme, the potential pros and cons of alternative designs (including political economy ones), and build a solid evaluation design to learn from what is being done."
Why the fuss about conditions?
Hoddinott suggested looking at the UCT versus CCT debate from a public as well as a private perspective, both of which provided "good rationales to impose conditions" on getting money from the state.
The case for CCTs
Governments may believe they know what is better for the poor "than do the poor themselves", and as a result impose conditions to bring about changes in their behaviour, Hoddinott said quoting from a paper he co-wrote with Alan de Brauw.
"For example, governments may place greater weight on the intrinsic value of educating girls than their families do," and give the family cash, provided they sent their daughters to school.
Governments could also use CCTs to create awareness or persuade people of the benefits of a course of action, such as being screened for a chronic disease.
Politicians could also use conditional transfers as a "useful tool to help them stay in office", Hoddinott said. Politicians were "often evaluated by performance indicators, such as changes in school enrolment or health clinic use.
"By conditioning transfers on behaviours that increase these indicators, politicians and policy-makers can provide useful evidence of accomplishments long before the indicators show more important evidence of poverty reduction (e.g. increased productivity or better adult health)."
Within households CCTs could empower or "strengthen the bargaining position of individuals whose preferences are aligned with the government's preferences," Hoddinott commented. Making welfare payments conditional could also have the affect of legitimizing the transfer and overcoming any stigma.
"Finally, recent work in behavioural economics emphasizes that when households are myopic (they make a decision today that they will regret in future), they are better off under constraints designed to reduce or limit their ability to trade future consumption for present consumption. Conditionality can be seen as such a constraint.
The case against CCTs
"Conditioning transfers can be perceived as being demeaning to the poor," Hoddinott noted. "Conditioning can be seen as implying that the poor simply do not know what is good for them."
Imposing conditions increased administrative costs and made cash transfers more complex, he said, citing studies from Latin America. It would be pointless if imposing the conditions became costlier than the actual benefits.
If the conditions imposed costs on the beneficiaries - such as getting to a school that was too far away - then it might also not be considered worthwhile. Some extremely poor households might find the conditions difficult to meet.
"If the preferences of the poor do not align with the conditions placed on their behaviour by the government, the restrictions that conditionality imposes on the poor will reduce their total welfare gains, thus decreasing the net benefits of the CCT," Hoddinott said.
Imposing conditions could also "create an opportunity for corruption, whereby individuals who are responsible for certifying that conditions have been met could demand payments for doing so."
ODI's McCord said many African countries, such as South Africa, ran social protection programmes without imposing conditions, and "I think there needs to be more discussion" of the issue.