Dambisa Moyo, author of Dead Aid and one of Time's most influential people in 2009, posted this on twitter yesterday: "Villagers using mosquito nets as wedding gowns: Yet another unintended consequence of a 'good' deed."  As someone trained in policy from Harvard and economics from Oxford, she should know better.  As someone with an increasing amount of influence, she should really know better.

We shouldn't be using anecdotes as support for any policy formulation.  This one especially concerns me because it is misleading.  Yes, bed nets intended for fighting malaria used as fishing nets or wedding gowns are wastes of resources.  That is unfortunate, but it is a consequence of free distribution.  And recent studies provide compelling evidence why we should do so.

To begin, the presence of externalities associated with malaria (i.e. a person falling ill has a negative effect not just on themselves but on society as a whole) justifies government intervention to align private and social costs.  So bed nets and medication and eventually vaccines should be subsidized by the government.  The question is, by how much?

Neoclassical economic theory would tell us to subsidize the difference between private and social costs.  This assumes perfect information and rationality, both of which often do not hold in the real world.  People may not understand the benefits of the nets.  You could do an information campaign, but that might be more costly than just giving away the nets for free.  More interesting and nebulous, people might not behave rationally even with perfect information (i.e. they don't properly calculate private cost).  This is the fascinating emerging field of behavioral economics I'll have much more to say about later.  But even if neoclassical assumptions held, knowing people's individual utility functions and calculating private costs is not a realistic endeavor for even the most adept governments.  So what then, should we charge, if anything?

The argument against giving away nets for free goes like this:  charging reduces waste by those who don't need or don't value the nets.  By charging some small fee, we can ensure that the nets get to those who need them.  There is also the argument that once you have paid for something, you value it more and hence are more likely to use it (e.g. I'm really not in the mood to go to this play tonight but I've already bought tickets).  This is known as the sunk cost fallacy.

Recent empirical evidence does not support these arguments.  A randomized field experiment in Kenya by Jessica Cohen and Pascaline Dupas found no evidence that charging for nets get them into the hands of those who value them most: they measured no difference in anemia, which is an important indicator of malaria, between those who paid for nets and those who received them for free.  The study also found no evidence of the sunk cost fallacy: those who received free nets were no less likely to use them than those who paid for them.

What they did find was a steep drop off in demand for nets when any positive price was charged.  A 87.5% subsidy saw uptake at 25% of the 100% subsidy levels.  Let's make these numbers clear: 12.5% less subsidy per net = 75% less people sleeping under them.  It is also important to note that much of the cost of net distribution is fixed, meaning average cost per net goes up significantly when demand drops from charging for nets.

Why we see this enormous sensitivity around zero is unclear, I think this is a behavioral economics question that we don't understand yet.   Regardless, the study provides compelling evidence for free distribution of bed nets.  Here's a link to their paper if you'd like to read more.

This is of course just one study, we need more to draw any conclusions.  But it's a scientific one, and a much more substantial basis to begin thinking about policy formulation than the misleading anectodes Dambisa Moyo is using to support her case.