Poor Economics: Rethinking Poverty & the Ways to End it

by Abhijit V. Banerjee

On the other hand, if people are subject to a sunk-cost effect, for example, these subsidies can backfire—usage will be low because the price is so low. In The White Man’s Burden,31 William Easterly seems to suggest that this is what is going on. He points to examples of subsidized bed nets being used as wedding veils. Others talk about toilets being used as flowerpots or, more graphically, condoms being used as balloons. However, there are now a number of careful experiments that suggest that such anecdotes are oversold. Several studies that have tested whether people use things less because they got them for free found no evidence of such behavior. Recall Cohen and Dupas’s TAMTAM experiments, which found that people are much more likely to buy bed nets when they are very cheap or free. Do these subsidized bed nets actually get used? To figure this out, a few weeks after the initial experiment, TAMTAM sent field officers to the homes of people who had purchased nets at the various subsidized prices. They found that between 60 percent and 70 percent of women who had acquired a net were indeed using it. In another experiment, over time usage went up to about 90 percent. Furthermore, they found no difference in usage rates among those who had paid for them and those who had not. The same kinds of results, which rule out the possibility that subsidies are to blame for low usage, have now been found in other settings. But if subsidies are not the cause, what is?