We’re hosting a guest for the next few days, an undergraduate student from Notre Dame.  She studies political science and economics, and her advisor was a graduate student with me (TP) at Yale as well as a fellow statistics consultant.  (Note: it’s a little unsettling that my friends are now teaching undergraduates at real universities.)  She is doing some research on trade policy preferences in Malaysia, and needed a place to stay for a week after just arriving.  Her real survey research will be in Ipoh, not KL, but until she leaves for there, she’s with us.

Her research on trade policy preferences is interesting.  The literature on this is actually very new, and centers around the question of how much people care about tariffs and other types of protection for industries.  The basic idea is that if your job is hurt by free trade, you will favor protection, and otherwise you’ll prefer no protection.  There are two reasons for the second view: first, no protection in your country probably means that other countries will lower protective barriers as well, which helps you; second, protection in any form is generally inefficient, and hence raises prices.  Anyway, a former Yale political scientist and a Dartmouth economist found a neat way to test this hypothesis by digging through National Election Survey data in the US.  (Working paper here, final article here.)  The null hypothesis is that people just have random trade policy preferences that don’t depend on their job.  They reject that; the job matters.  There are two alternative hypotheses, on that focuses on factors (basically, class) and one that focuses on sectors (basically, industry).  In the first model, unskilled workers in the US should favor protection because they are the ones who suffer from free trade.  In the second model, only unskilled workers in sectors that compete with imported goods (like textile manufacturing, but not unskilled service industries) will favor protection.  The difference is only how intersectorally mobile these unskilled workers are.  In the first model, workers are mobile; in the second model, they are not.  Note that you could express this a different way: people in the first model have long time horizons in terms of policy decisions, but in the second model, they have short time horizons.  (Because of course, in the long run, all labor is mobile.)

Anyway, the authors find support for the first hypothesis.  Our guest is interested in Malaysia to see if this works in developing countries.  In developing countries, though, the predictions should be switched.  Unskilled laborers should be in favor of free trade because it benefits them to have access to markets in developed countries, while holders of capital (the scarce resource which competes domestically with international capital) should favor protectionism.  You can also tweak the predictions to test for the short-run, sector-specific factors model.

So that’s what our guest is interested in studying.  It’s really cool stuff.  I believe that someday I’ll have to come back and do this study as well, but with a twist.  I hypothesize that since politics in Malaysia is dominated by the relative positions of ethnic groups, different ethnic groups will have different preferences, controlling for industry.  So, controlling for sectors, Malays and Indians will favor trade (or protectionism, not sure which one yet) as a whole while Chinese will favor the opposite.  I have to work through this to get the hypotheses correct, but it could be an interesting way to study how trade policy gets formed in a developing country.