About Ceteris Non Paribus

This blog is a place for me to put the ideas I have, and the stuff I come across, that I’ve managed to convince myself other people would be interested in seeing. It also serves as a way for folks who care what I’m doing (hi, mom!) to keep track of what I’m up to when I’m on another continent or otherwise out of touch. I’ve done stuff like this in the past, but my previous blogs have been trip-specific (Jason Does India) or joint efforts with other people (Positive Impact Institute, Methodlogical). This time, it’s a solo effort, and one I intend to last. That means that I am aiming a bit lower – at one-off opinions without tons of research or effort, rather than the kind of thorough, thoughtful analysis that the bloggers at Development Impact put up each week. In this I am inspired by Chris Blattman’s thoughts on staying power and blogs. He was probably being facetious when he said that “sometimes I even read part of the papers I reference”, but in my posts here I may do that quite literally. Carefully thinking things through what my day job is for; this blog is intended for stuff that’s just a bit too highbrow for facebook.

“Ceteris Paribus” is Latin for “other things equal”, and is the implicit assumption in all causal analyses. Suppose I observe groups A and B, seeing that group A does X while B does not, and therefore deduce that the different outcomes in A are attributable to X. This is true so long as ceteris paribus holds – if X is the only difference between the groups. Scientists can use randomized experiments to ensure that all else is in fact equal, but in the absence of an experiment it is almost always the case that ceteris non paribus: tons of other stuff is going on. I’ve named my blog after my obsession with this basic inferential error, because, while people make it everywhere, my research areas (public health and development economics) are especially plagued by failures to hold all else equal. Articles on Yahoo! News treat people who don’t drink red wine as a valid comparison group for people who do, and find all kinds of spurious health benefits. NGOs and advocates select the best schools located closest to roads for a pilot intervention, then tout the big advantage in test scores between the intervention schools and the rest of the country. 90% of my academic work boils down to saying “not so fast – what’s your comparison group?”, and one major purpose of this blog is for me to vent my frustrations about people constantly getting it wrong.

9 thoughts on “About Ceteris Non Paribus

  1. Jason Kerwin is a Ph.D. candidate in economics at the University of Michigan with interests in development economics, economic epidemiology, and labor economics. His research focuses on using randomized trials in developing countries to address first-order questions about decisionmaking behavior, while at the same time generating results that are valuable to local policymakers. His most recently-completed trial studied how individuals react to their subjective beliefs about disease risks by adjusting their health behaviors. His ongoing projects include research in Malawi on how exposure to temptation affects people’s ability to optimally smooth consumption and meet their savings goals, as well as a project in Uganda that studies the broader “spillover” benefits of an education program to entire schools as well as pupils’ families. Prior to starting his Ph.D., Mr. Kerwin received a BS in physics, a BA in international relations, and an MA in international policy studies from Stanford University.

Leave a Reply

Your email address will not be published. Required fields are marked *