Abstract
This study examines the extent to which the positive correlation between employment in an exporting company and wages reflects a causal relationship. In order to identify this relationship, I use several methods on the basis of three different datasets: controlling for observable and unobservable individual characteristics by means of a datafile that tracks workers over a number of years; controlling for observable and unobservable characteristics of individuals and companies by means of a datafile that tracks companies; and controlling for individuals’ skills by means of cross-sectional data from an international skills survey (PIAAC). I find that while each percentage point in an industry’s export rate is correlated with a 1-percent increase in wages, the causal relationship, which is similar across all of the methods, is much weaker and lies between zero and 0.2 percent. Finally, I find that the simple correlation between the export rate and wages in Israel is unusually large relative to other countries in the PIAAC sample, while the causal relationship is similar to the premium in other countries, although it is in the upper part of the distribution.