J-PAL studies demand, forgets about supply

I’m back in Mbale after a weekend trip to Bududa, a small town that was devastated by mudslides in March. Over the course of two days I learned new things, acted like a total cliché, and was greatly humbled by meeting some incredible people. More on that over the next day or so.

Right now, I want to comment on a nice little summary of health intervention pricing studies from the Poverty Action Lab (aka J-PAL). Each study looked at how the quantity demanded of various interventions (deworming drugs, insecticide-treated bednets, chlorine for drinking water, soap) differed as the price of the product was varied. These studies found that access to the intervention was higher when it was free; so was utilization of the product. These were great studies, and I’m generally a big fan of J-PAL.

But: I disagree with the inference from these studies that the health products should be provided free of charge. Duncan Green (on whose blog I saw the study, and whose analysis I generally like) calls this an “unequivocal conclusion“.

The problem: these studies looked at demand, found it to be high when the price was low (duh), and called that access to the health interventions. They didn’t look at the other curve that every Econ 101 student knows so well: supply. That’s okay, these studies are still great for what they are. However, we can’t conclude from them that a price of zero will ensure the products reach those who need them. The researchers who ran the studies had outside funding to supply the products at any price. The real world is not so simple.

A price of zero only works when a government or donor runs the supply operation. We know that such service delivery often fails to reach the efficiency and penetration of private sector supply chains. Here in Mbale, every third shop sells pre-paid airtime cards for MTN, Warid, or other networks. This is even true in nearby rural villages that barely have electricity. There is an incredibly extensive supply network because those doing the supplying have a chance to profit. Obviously, bed nets and other health products are different from cell phone minutes (for one thing, they’re heavier). But even still, we think too much of ourselves to assume that governments/donors can match this kind of efficiency and penetration.

If setting a price allows people to make a profit, we may actually generate more total access to the health intervention than if we try to give it away for free. We can’t know unless we study intervention supply chains as much as we study demand.