Feb 10, 2009

When It Pays to Pay

by Laura Starita

We have written in the past about the ongoing debate about whether asking clients to pay part of the cost of condoms, water purification tablets or bed nets improves the effectiveness of aid programs. Cost-sharing advocates argue that applying a nominal price to a good serves to filter out those who neither want nor need it, and imposes a “sunk cost” psychological effect on those who do buy, which encourages use. Together, the argument goes, these factors increase use above what would be achieved through free distribution, at less cost. Those in favor of free distribution, in contrast, say that applying even a nominal cost to these items puts them out of reach for the severely impoverished, many of whom are sick and more vulnerable to illness.

Economists have taken on this debate as a ripe subject for randomized controlled trials, and a number have produced interesting results. A year ago we referenced a study which tested use of a water purification tablet, Clorin, and showed that, indeed, demand declined as prices rose, but buyers who paid used the product more consistently than those who received the product for free. This seemed a strong indication that cost-sharing advocates were on to something.

But does it apply beyond water purification? Last week I came across a 2008 study conducted by economists Pascaline Dupas, from UCLA, and Jessica Cohen, from the Brookings Institution, on cost-sharing in malaria bed net distribution The researchers studied 20 health clinics in Western Kenya, four of which acted as the control, and the remaining 16 of which were provided with bed nets to distribute to prenatal clients at randomized prices ranging from free to 40 Kenyan Shillings ($.60)—note that $.60 is a 90 percent subsidy of the real costs of a bed net, and the prevailing cost-sharing price in Kenya is equal to $.75. The researchers found that demand declined by 60 percent when the price rose from zero to $.60 but that paying a price for the nets seemed to have no impact on use. If anything, women experiencing a first pregnancy who received the nets for free seemed more likely to use them than women who paid. The study estimates that cost-sharing, at least in the Kenyan context, would result in 75 percent fewer pregnant women receiving insecticide-treated bed nets than would under a free-distribution effort. This lower usage rate is particularly relevant as it pertains to bed nets, because they help not only the person who sleeps under it, but also, according to epidemiologists, result in a lower carrier mosquito population in the community once fifty percent of residents use them.

Dupas and Cohen caution that their results may only be relevant to Western Kenya, where malaria is endemic year-round, and where focused efforts to market bed nets for malaria prevention have been ongoing for years. Nonetheless, their results provide a lesson to those of us (myself included) who are often tempted to extrapolate. Kenya is not Zambia, and Clorin—a consumable which benefits only the person who uses it—is not bed nets. Context is everything in poverty interventions, and this study provides not only insight into what could work to stem malaria infection in Western Kenya, but a reminder that we have to test and re-test our assumptions in the real world.