Aid - free or cost-sharing?

One of the many open questions discussed in the aid community is whether aid is more effective if offered free of charge or if the poor have to pay for it. In White Men’s Burden. Why the West’s Efforts to Aid the Rest Have Done So Much Ill and So Little Good, development expert William Easterly describes how Mosquito nets often don’t make it to the poor: distributed by aid agencies for free (for example by the UN), they end up on the black market or get used for purposes other than intended – as fishing nets or wedding veils. According to a study of free mosquito distribution in Zambia, 40% of the nets were not used for their intended purpose.

On the other hand, the non-profit organisation Population Service International in Malawi sells insecticide treated bed-nets for 50 cents to pregnant mothers though birth clinics. The nurses selling the nets are allowed to keep 9 cents themselves and thus have an incentive to always have them in stock. The programme itself is financed by the profits from regular net sales (5 US$ a piece) to more wealthy Malawians. It was a huge success, resulting in 55% of pregnant women and children under 5 using bed-nets (up from 8% in 2000).

The rationale of the cost-sharing proponents is that people paying for goods and services will give aid institutions valuable feedback, whereas when delivered for free, the poor don’t have any power to complain or reject the goods offered to them.

“Charging the poor modest fees for health care is a way to increase accountability for delivering health services. If the villagers don’t get a good service after they have sacrificed to pay for it, they loudly complain”, says the founder of Gonoshasthaya Kendra (People’s Health Center), a Bangladeshi NGO charging a small fee in return for their support of pregnant women.

A new study by the Brookings Institute Free Distribution or cost sharing?Evidence from a randomized Malaria Prevention Experiment comes to a different conclusion.

The study, based on a randomised experiment in Western Kenia, shows significantly more pregnant women using insecticide treated bed-nets when distributed for free. The women receiving bed-nets without payment were consistently more healthy than their cost-sharing sisters in the control group.

Randomly assembled test groups – in which one group receives an intervention wheras another doesn’t – are the most effective evaluation methods for another aid expert, Abhijit Banerjee, director of the Poverty Action Lab at MIT. In his book Making Aid Work, he describes how randomized tests can determine what works and what doesn’t in aid policy. Some of the results are counterintuitive. Thus also in Western Kenia, such tests were used to evaluate whether the use of flipcharts in schools was beneficial or not. Surprisingly the study showed that students being taught with flipcharts learned less than those using more conventional approaches.

What accounts for the discrepancy in malaria prevention in Malawi and Western Kenia? I would imagine that every intervention has to be adapted to its specific local context. In the world of development there are only few globally valid solutions – what worked in one context, let’s say in public health clinics Western Kenia in 2007, isn’t necessarily THE answer. Different circumstances, whether cultural, economical or socio-political, can lead to different results and necessitate different solutions.

The conclusion of Next Billion thus seems realistic: development experts will get used to the idea that in some cases cost-sharing market solutions are appropriate, whereas in other contexts free distribution is more effective. There are no ready-made answers, fitting all situations. Instead interventions will have to be adapted to local contexts.