The distance from the country of origin to the United States (in thousands of miles).24 Boijas and Bratsberg (1996) have shown that the return migration rate of immigrants in the United States is negatively correlated with distance from the source country. This empirical finding suggests that immigrants who originate in far-away countries are more likely to view their migration to the United States as permanent, and have greater incentives to invest in U.S.-specific capital. In terms of the theoretical framework, longer distances decrease the probability of outmigration and increase the discounting factor p. Distance from the source country, therefore, should have a positive effect on the rate of wage growth.
Political conditions in the country of origin. Barro and Lee (1994) used the Banks (1986) Cross-National Time-Series Data File to calculate the number of revolutions (per year) that occurred in the various countries in the periods 1960-64, 1965-69, 1970-74, and 1975-79. Political instability in the country of origin would likely have an impact on the return migration rate, and should again affect the discounting factor. A higher degree of political instability, therefore, would presumably lead to higher rates of wage growth in the United States.
It is useful to begin by documenting that the source country characteristics have effects on the log entry wage that are consistent with those reported in the existing literature. The log entry wage is defined by the fixed effect vijk(t) calculated in equation (12), and is obtained from the 1970 Census for the cohorts that arrived in the 1960s, and from the 1980 Census for the cohorts that arrived in the 1970s. The main specification of the log wage regressions is reported in the first column of Table 6.
In general, the results are consistent with the evidence reported in existing studies. Immigrants who originate in richer countries earn more and immigrants who originate in countries with high levels of income inequality earn less. The regression also reveals that immigrants originating in open economies earn more, and that the immigrant groups who exhibit substantial geographic clustering earn less. The remaining columns of Table 6 show that the inclusion of educational attainment or of country-of-origin fixed effects weaken many of the coefficients.