Restoring culture and capital to cultural capital: origin–destination cultural distance and immigrant earnings in the United States

Abstract

ABSTRACTAn extensive sociological literature maintains that cultural capital is pivotal in perpetuating social inequalities. However, empirical tests of cultural capital theory focus on how culture influences educational outcomes, not earnings, and they mainly look for cultural differences across social classes within societies. We propose a direct test of economic returns to cultural capital based instead on differences in national cultures across countries. Using the American Community Survey and the National Survey of College Graduates, we analyze the relationship between immigrants’ lack of U.S.-specific cultural capital, proxied by cultural distance between the origin country and the U.S., and their earnings. Findings consistently indicate that origin – U.S. cultural distance is linked to immigrants’ lower earnings after controlling for numerous other factors, supporting cultural capital theory. Cultural distance earnings penalties are more pronounced for immigrants with at least a bachelor’s degree, those arriving in adulthood, and those with foreign degrees. Moreover, county-level analysis reveals more sizable cultural distance penalties in more competitive and unequal labour markets, highlighting how subnational receiving contexts shape origin-country disparities in immigrants’ economic incorporation at their destinations.KEYWORDS: Immigrant economic incorporationcontexts of receptionplace of educationreturns to educationcultural capital Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1 For example, Friedman and Laurison (Citation2019, 15) pithily summarize the role of cultural capital after acknowledging it is a ‘more complex’ and ‘hard to detect’ aspect of class privilege than economic or social capital: ‘simply by expressing their tastes or opinions, the privileged are able to cash in their embodied cultural capital in multiple settings.’ The only quantitative evidence they provide for such ‘cashing in’ on upper-class culture is an origin-class pay gap within elite professions. However, such a pay gap might stem from any unobserved factors that both affect earnings and also correlate with class origin.2 A related body of literature in business studies shows that cultural distance between countries affects corporate decision-making about whether to integrate foreign entities or form joint ventures, as cultural similarity lowers the economic uncertainties associated with integration (see Kogut and Singh Citation1988).3 ACS respondents with negative or zero earned income in the surveyed year accounted for only 0.06 percent of the overall analytic sample; excluding these individuals does not alter our results.4 We excluded from this analysis immigrants from member states of the former Soviet Union because the ACS did not further distinguish the individual countries, such as Russia, Estonia, Armenia, and Tajikistan. The ACS also lumped ‘North Korea’ and ‘South Korea’ into a single region called ‘Korea,’ and we used the Hofstede indices for ‘South Korea’ to represent this region, on the assumption that most Korean immigrants in the U.S. are from South Korea.5 We normalized the county-level standard deviation of earnings using the contemporaneous nationwide standard deviation of earnings as the base, generating the normalized standard deviation of individual earnings within the local county (ranged 0∼1).6 We tried additionally controlling for industries, but the results did not change substantively, so we omit them.7 We also report the origin-country-level raw cultural distances from the U.S. (see Figure A1) and the origin-country-level bivariate relationships between average immigrant annual earnings and raw cultural distances across those four cultural dimensions (see Figure A2).8 For a full descriptive summary of all the covariates, please refer to Table A1.9 The ‘shared official language’ dummy yielded insignificant or inconsistent results, from both the ACS sample (Table 1) and the NSCG sample (Table 2). On the other hand, while the ACS-based findings revealed a significantly negative association between sending-country GDP per capita and immigrant earnings, the NSCG-based results were insignificant. The two origin-country-level controls generally rendered inconsistent findings compared to the robust and consistent results about SCD earnings penalties.10 Adult/child immigrants denotes immigrants who were adults/children at the time of immigration.11 AIC and BIC support the inclusion of the interactions between SCD and the education categories.12 For highly educated immigrants who completed high school in the U.S., there are no statistically significant main SCD effects or main educational gradients even if the interactions between SCD and degree types are excluded (results available upon request).13 The tests suggest statistically insignificant SCD penalties for those whose highest credential is a US bachelor’s degree, regardless of the high school location.14 As Figure 4 shows, immigrants from culturally similar countries to the U.S. with foreign bachelor’s degrees actually have higher predicted earnings than immigrants from the same countries with U.S. bachelor’s degrees. This may seem counterintuitive, but it probably results from self-selection into adulthood immigration by those who received bachelor’s degrees abroad. Citizens with a college education in countries that culturally resemble the U.S. are likely to reap high average returns to their degrees in their origin countries, so those that emigrate to the U.S. are probably distinguished by especially high returns in the U.S. labor market due to individual idiosyncrasies (e.g. specific fields, institutions, or unobserved ability).15 It does not matter whether we use the raw scores or the US-centered scores (obtained by subtracting the US score from the corresponding origin-country score on a particular dimension) because this centering only affects the intercept, not the regression slopes.16 See Figure A4 for a sensitivity analysis using refined age-at-immigration groups, which yields the same conclusion as the binary age-at-immigration groups presented in the main text.17 The mitigating effect of a US college degree is also suggested by the minuscule SCD penalty for college educated child immigrants in the ACS analysis, most of whom probably attended US colleges. The linear combination of main and interaction effects (−.025), while statistically significant, is dwarfed by the implied effect for adult immigrants with college (−.150). The two sets of results indicate SCD penalties are strongest for those with bachelor’s degrees from non-US institutions, and college degrees from US institutions substantially, perhaps entirely, mitigate the disadvantage of coming from a culturally distant country.18 A third possibility, suggested to us by an anonymous reviewer, is that origin-country cultural values governing the extent to which individuals value earnings may also affect immigrants’ self-selection into some occupations. We leave the task of adjudicating among the three alternative explanations for future research.