Previous work on Hispanic men finds that earnings rise with English proficiency. Several studies have found, in fact, that previously found non-Hispanic/Hispanic earnings differentials can be accounted for completely by differences in English proficiency. We question existing work on two grounds. First of all, the more able are more likely to be proficient at any point in time, so that English proficiency may simply be a proxy for unmeasured ability. Ignoring this role of English proficiency will understate non- Hispanic/Hispanic earnings differentials and overstate the return to learning English. Our second question of existing work causes biases in the opposite direction. At least some Hispanics who are proficient in English at any point in time acquired their proficiency after entering the U.S. labor market. Such individuals cannot be expected to have had the same wage growth as those who were proficient from the start. Ignoring the history of when proficiency was acquired will overstate differences between non-Hispanics and Hispanics, and understate the return to learning English.
We use data from the 1976 Survey of Income and Education to construct a model of the acquisition of English proficiency. Using predicted as opposed to actual levels of English proficiency, we are able to avoid some of the pitfalls of previous work. We find that Hispanic men we predict to have been proficient prior to entering the U.S. labor market do not earn less than non-Hispanics with the same measured characteristics (schooling, work experience, etc.). This confirms previous results, but apparently because the above mentioned biases approximately cancel each other out. We also confirm that there is a positive return to learning English. Individuals who enter the U.S. labor market with low predicted English proficiency, but then experience rapid improvement in their English skills, earn less than you would expect of individuals with their English skills at the time of the survey. This is consistent with our hypothesis that an individual’s history of English proficiency, and not just his current level, is important in determining earnings. We also find, contrary to previous work, that Cubans in the Northeast earn considerably more than other Hispanic groups in that region. We suspect this is due to selection. That is, Cubans in the Northeast are atypical of Cubans in general, the overwhelming majority of whom have chosen to locate in South Florida.