In search of the causes of the globalization backlash: methodological considerations on post-treatment bias

We study the implications of post-treatment bias in the context of the globalization backlash. We discuss the viability of horse-race regressions aimed at assessing the relative role of economic vs. cultural drivers of the backlash. First, we develop a contrived example in which exposure to an economic shock raises support for the radical right, while at the same time changing the cultural traits of a fraction of the population from libertarian to authoritarian, with authoritarian traits being positively related to radical-right support. Given this hypothetical, yet plausible, set-up, and using also simulated data, we show how controlling for culture leads to biased estimates of the economic shock’s effect on support for the radical right. Such horse-race regressions do not allow to estimate the relative role of the economy vs. culture, nor they allow to grasp the mediating role of culture. We complement the methodological discussion with an application to observational data from 15 western European countries.