Last weekend, I went to Cape Breton to give a talk at a TEDx event. In case you don’t know, Cape Breton is wayyyy on the eastern part of Nova Scotia, and isn’t particularly flush with convenient flights into the area. So my weekend looked something like this:
This observation is actually kind of relevant- the theme of the conference was how to think intelligently about how to revitalize the area, since Cape Breton has suffered from a declining population and loss of industry in recent (and some not so recent) years. I couldn’t really give a talk on how to move the island in order to be more conveniently located for business purposes, and I’m not so familiar with the specific issues facing the island, so the best I could do was to give a talk about how to correctly measure the impact of potential social policies and economic development initiatives.
Not surprisingly, my talk was an attempt to steer people away from the correlation versus causation problem, or, in this case, specifically the “post hoc ergo propter hoc” problem. In other words, people have a tendency to look at initiatives and draw conclusions of the form “well, things got better once the program was implemented, so the improvement must have been because of the program.” This reasoning is incorrect since there is no comparison to what would have transpired had the program not been implemented.
One way to get around this problem is to revert back to the middle-school science project approach and conduct controlled experiments. The logic is that if a researcher constructs an experimental group and a control group so that the only difference between the groups is the treatment, any differences in outcomes for the two groups can be properly attributed to the treatment. In fact, this is what Professor Esther Duflo does in developing countries in order to assess the impacts of different social policies such as malaria bed net distribution and immunization incentives:
You can see comments and even download a version of the talk here. If you are the reading type, Prof. Duflo co-authored a book about her experiments called Poor Economics, which is also pretty interesting.
An economist friend of mine informed me that $7 malaria bed nets can reduce infant mortality by 20 percent, so it’s pretty clear that figuring out how to get these things to the people who need them (and getting those people to use them properly) is of the utmost importance. In fact, economists do, quite literally, put their money where their mouths are. For example, TAMTAM is a nonprofit organization founded by economists, and its main goal is to distribute malaria bed nets to rural communities in Africa in accordance with the methods that have been shown through the research of Professor Duflo and others to be the most effective and efficient at reducing the incidence of malaria. Similarly, the Bill and Melinda Gates Foundation hires economists to evaluate how to allocate philanthropic resources efficiently.
Research into program effectiveness is beneficial on multiple levels. The obvious level is that resources can be used more efficiently and for greater impact than if people have to guess about what works and what doesn’t. The less obvious level is that people are likely more willing to donate to fund these sorts of programs if they are confident that their contributions will have a significant impact.
I must admit, however, that Cape Breton, at least visually, doesn’t exactly seem like it needs my help: