In the social sciences, intergenerational transmission refers to the transfer of economic or social status across generations. These cross-generational transfers occur through a variety of means, including the inheritance of occupational status, educational attainment, earnings, and wealth. In a society marked by significant levels of inter-generational transmission, there will be a corresponding persistence in economic and social stratification, and a child’s socioeconomic outcome will be highly predictable. While immobility is not necessarily synonymous with inequality, intergenerational transmissions can generate unequal opportunities for children born into the society. Thus, the presence of intergenerational transmissions has strong implications for public policy, and many scholars have made efforts to improve the knowledge of how and to what extent these transfers exist.
RESEARCH ON LEVELS OF INTERGENERATIONAL MOBILITY
A central question in the analysis of social mobility and stratification is the extent to which advantages and disadvantages are passed from generation to generation. There is a rich history of scholarship attempting to quantify the level of intergenerational transmissions. In the 1950s sociologists improved both local data collection and mobility scales, and began to assemble intergenerational mobility tables to examine inflow and outflow percentages among different occupational classes. These classes were oftentimes highly aggregated, making it difficult to arrive at substantial conclusions on intergeneration mobility. Methodologies improved significantly in the 1960s, with the introduction of the socioeconomic index (SEI), a set of scores for occupational categories that allowed researchers to apply continuous data analysis techniques to occupational groups. In 1967 Peter Blau and Otis Dudley Duncan (creator of the SEI) published a groundbreaking work, The American Occupational Structure. In it, Blau and Duncan introduced their model of the "process of stratification," which allowed for the multivariate testing of direct and indirect effects (such as education and family background) on the earnings of children. In isolating the effects of various factors on a child’s occupational status, Blau and Duncan found that the ratio of education effects to father’s occupation effects on a son’s current occupation is approximately 2.9 to 1, suggesting that achievement tends to be far more important than ascription in occupational outcomes.
Many scholars from around the world began to apply status-attainment models to their own intergenerational transmission analyses in the 1970s. As methodological improvements continued to be made, including the transition from multivariate linear regressions to loglinear models, a growing consensus developed that intergenera-tional mobility was at high levels in industrial countries. Specifically, in several Western European countries and the United States, researchers in the 1970s and 1980s consistently found intergenerational elasticity coefficients of 0.20 to 0.25 (which implies that only 20 to 25 percent of earnings gaps between groups would remain after one generation). In a 1988 address, American Economic Association president Gary Becker’s comments reflected widespread sentiments at the time: "In every country with data that I have seen … earnings strongly regress to the mean[;] … low earnings as well as high earnings are not strongly transmitted from fathers to sons."
Yet the common belief that intergenerational mobility levels were high in industrial countries would change dramatically in the late 1980s and early 1990s. Regarding the United States, specifically, recent results for elasticity coefficients suggest that intergeneration mobility is more limited than previously thought. Gary Solon (1989) argued that the low correlation between a parent and child’s socioeconomic indicators was due to selection biases and poor data measurement in previous studies. By using an entire dataset from the Panel Study on Income Dynamics (as opposed to the numerous data omissions typical of previous research) and using a five-year average of fathers’ earnings data in lieu of the standard one-year data (to better represent lifetime earnings), Solon (1992) discovered that intergenerational elasticity coefficients rose significantly from 0.21 to 0.41. Using data from the National Longitudinal Survey, David Zimmerman (1992) arrived at a similar conclusion, and in combination these two studies provided compelling evidence that after methodological corrections, intergenerational mobility was more limited than previously thought.
Many scholars have since tried to replicate and expand on Solon and Zimmerman’s original findings, and most have arrived at an intergeneration elasticity of about 0.4 or higher for the United States. Bhashkar Mazumder (2005) argued that the usage of five-year averages on fathers’ earnings is flawed because data from such a short period of time can still reflect the effects of temporary shocks, which tend to be correlated in time. Mazumder is able to average fathers’ earnings over sixteen years using social security earnings histories, and finds that the inter-generational elasticity between a father and a child’s earnings is dramatically higher at 0.6. By contrast, researchers in other industrialized countries have typically found intergenerational elasticity coefficients that are less than in the United States. It is important, however, to note the difficulty of cross-country coefficient comparisons, because variations in earnings measures, age ranges, and other important study characteristics can be misinterpreted as differences between countries. Taking this issue into account, it appears that Sweden, Canada, and Finland enjoy more mobility than the United States, whereas Britain and France may be slightly worse. "American exceptionalism"—the perception that the United States is a uniquely mobile country—is largely discredited as a myth.
RESEARCH ON HOW INTERGENERATIONAL TRANSMISSIONS OCCUR
Scholarly evidence suggests that intergenerational mobility is highly limited, particularly in the United States, but why it is limited and how these outcomes are transmitted across generations is not clear. One theory that has garnered much attention is the human capital explanation, developed by Gary Becker and Nigel Tomes (1979, 1986). Becker and Tomes argue that the human capital acquired by children greatly affects their earnings outcomes, and that this capital is greatly affected by investment choices made by "utility-maximizing" families. Given a limited set of resources (time, income), parents will continue investing available resources in a child until their expected rate of return (determined by a child’s ability and earnings potential) is no better than the market rate of return on investments (i.e., interest rates). Some implications of this theory are as follows: (1) Given unlimited resources, the levels of investment in a child will be determined by the child’s ability; (2) the presence of credit constraints (in low-income families) will lead to reductions in child earnings and less intergenerational mobility. Several findings support the Becker-Tomes framework, most notably the evidence from cross-country comparisons suggesting that countries with high levels of educational subsidies (and therefore lower levels of credit constraints) tend to be associated with more intergenerational income mobility. However, as previously mentioned, cross-country comparisons are difficult given the variety of alternative explanations for differences, and it should be noted that the overall evidence on the human capital investment explanation is mixed.
A second economic explanation for the persistence in intergeneration immobility is the transfer of wealth in the form of inheritances. This explanation has received little attention due to the lack of panel data on respondents who have reached the age at which bequests are typically passed on. In studying the death records of wealthy Connecticut residents and their children, Paul Menchik (1979) found an intergenerational elasticity of 0.69 between child and parent wealth at death. This finding suggests that inheritances are highly important for the upper end of the income distribution. However, scholars have largely dismissed wealth as a major factor in overall intergenerational persistence of outcomes simply because most families do not pass along substantial inheritances. For example, according to Casey Mulligan (1997) only 2 to 4 percent of estates passed on from deaths in the United States between 1960 and 1995 were subject to inheritance taxes.
Several other avenues of intergenerational transfer deserve mention as alternative explanations of immobility, despite having received minimal scholarly attention. George Borjas (1992) alludes to the possibility of "neighborhood effects" in his assessment of the impact of parent and ethnic group successes on the wage rates and occupational statuses of first- and second-generation Americans (excluding African Americans and Native Americans). His results suggest that ethnic group effects are at least as high as parent effects on the economic successes of these immigrant children. While Borjas does not attempt to explain what underlies these group effects, several possibilities exist, including discrimination, cultural effects on behavior and personality, and unequal access to information.
If discrimination is in fact a significant explanation for intergenerational immobility, this poses a considerable challenge to a society’s claim to be a meritocracy. Indeed, scholarship since The American Occupational Structure has acknowledged the unique difficulties faced by African Americans in achieving economic and social mobility. Using data from the Panel Study on Income Dynamics, Tom Hertz (2005) finds a 25 to 30 percent difference in income between the adult income of blacks and whites who grew up in households with identical long-run average incomes, even when parental levels of education were taken into account. But is this due to discrimination, or other potentially confounding factors? For example, Dalton Conley (1999) asserts that it is not race that matters in the intergenerational transmission of poverty, but rather parental assets or the class status that is typically associated with race in America. Alternatively, Thomas Sowell (1981) argues that cultural deficiencies, not market-based discrimination, explain intergroup differences in economic success.Overall, controversy continues over the interpretation of racial and ethnic differences. Further analysis is required to determine the extent to which a "legacy of race" is reflected in economic outcomes.
While the evidence is not definitive, research suggests that levels of economic mobility across generations are low, especially in the United States. Explanations for this persistence are inconclusive, although several possibilities, including human capital, wealth, group effects, and discrimination, have received some attention. Samuel Bowles and Herbert Gintis (2002) attempt to provide a comparison of some possible causal channels of intergenerational status transmissions in the United States, and conclude that wealth, race, and schooling are important, whereas IQ and genetics are not. Though the authors acknowledge that much of the transmission process remains unclear (at least two-fifths is still unexplained in their work), their analysis serves as a model for future research. Experts hope that continued efforts in this direction will result in a better sense of how intergenerational transmissions occur, and perhaps how social inequality can be rectified.
