MY RESEARCH HAS FOCUSED ON THE INTERGENERATIONAL TRANSMISSION OF SOCIO-
economic inequality and the public policies that affect that process. My dissertation on race, class and wealth received the 1997 American Sociological Association Dissertation Award. A book that emerged from this work entitled Being Black, Living in the Red: Race, Wealth and Social Policy in America was published by the University of California Press in 1999. The book argues that traditional measures of social class and / or socio-economic status have been too dependent on labor market measures. This reliance on labor market indicators of social class resulted has its roots in 19th century social thought and resulted in the neglect of property in class analysis throughout the 20th century. I use data from the Panel Study of Income Dynamics to show that when property measures (i.e. net-worth) are included into our conception of social class, black-white differences in a variety of areas are eliminated - even those which had persisted over and above income, occupation and education controls. Among the generation born since the Civil Rights triumphs of the 1960s, class dynamics in general and property relations in particular appear to be the basis of persistent racial differences in areas of life ranging from the propensity to work or be on welfare, wage rates and educational attrition. The book concludes by suggesting property-based alternatives to the current affirmative action policy in the United States.

In 2000, my sociological memoir, Honky was released by the same publisher (the paperback edition was put out by Vintage Books in 2001). This book describes my own experience of growing up white in an area of predominantly minority New York City housing projects. Through narrative that results from in-depth interviews of my “truly disadvantaged” neighbors, it sociologically addresses the categories of “white” and “middle class.” By being socialized as a white minority in a community of color, my experience illuminates the racial category “white” that most Caucasian Americans take for granted.

The book also attempts to document how middle class status can assert its privileges even when a family has no money. That is, it describes how “middle class-ness” is less about current economic circumstances and more about life chances, opportunities and expectations.

I have just completed a major project that examined class differences among individuals from the same family of origin, using intra-family inequality as the counterfactual to processes of social stratification. This study used a combination of survey data (PSID, SAF & the US Census) and about 175 in-depth interviews of siblings to investigate the sizable portion of variance in socio-economic outcomes that exists within families. In this vein, the project sought to open up the “black box” of the family in stratification research, viewing background as a fluid, endogenous variable. This project was supported by a Robert Wood Johnson Investigator Award in Health Policy Research and a NSF CAREER Award.

REASEARCH ON ECONOMIC INEQUALITY AND HEALTH

One current line of research investigates the relationship between health and social inequality.  At least since the now-famous Whitehall Study, socio-economic status (SES) gradients in both mortality and health status have been taken as a given. Across time and place, study after study has reinforced the conclusion that groups with higher SES – as measured by educational attainment, job classification, wealth, or income, for instance – display lower death rates and overall better health than their lower status counterparts (McKeown, 1988; Wilkinson, 1992). Much scholarship has been devoted to interpreting these observed correlations as well as understanding the underlying mechanisms that lead to the observed disparities in the first place (e.g. Taylor, Repetti, & Seaman, 1997). Some of the main questions yet to be answered include whether differences in social status are indeed causal of better or worse health, to what extent causality flows in the opposite direction (poor health causing low SES); and to what extent both are predicted by an underlying third variable (such as success at dealing with stress, genetic endowment, toxic environments and so on). These question have been particularly pressing since SES-health associations persist even in societies where there is universal access to medical care, and they often appear at all strata of society (Marmot, Shipley, & Rose, 1984; Marmot, Smith, Stansfeld et al, 1991). Furthermore, status differences in self-rated health, morbidity and mortality rates are often observable at the upper end of the distribution where inadequate material resources would not seem to be accounting for these effects (e.g. Redelmeier & Singh, 2001).  Much of my past, current and future research has set out to answer this question of causality using different methodological strategies at different points in the life course.  These methods range from using family-fixed effects models to factor out unobserved heterogeneity, to using wealth, inheritance, and investment income as an instrumental variable, to developing endogenously developed hierarchies for adolescents using focus group and network methods, to using income changes induced by the Vietnam era draft lottery as an exogenous estimation strategy of income effects on mortality.  The following is a description of some of this work:

Infants: Examining low birth weight would seem to provide an ideal heuristic for addressing issues of causality with respect to health and social status.  The infant’s future SES cannot be causal of her health.  Likewise, the infant’s health is not causal of the parents’ SES up to the time of birth (though, of course, it may be after that).  However, when we view social inequality and health as having strong intergenerational components that affect one another, the story gets more complicated.  Using the PSID, my published work on birth weight uses sibling comparisons to estimate an effect of income on birth weight and to estimate an effect of birth weight on adult educational attainment (19 years later).  Sibling comparisons provide a methodology to factor out unobserved heterogeneity to the extent that it is stable across pregnancies.  We find birth weight is a strong predictor of high school graduation.  We also find that income has no effect on birth outcomes for the vast majority of the population.  But it does have a salutary effect for those who are medically at risk – that is, whose parent(s) are also of low birth weight.  We interpret this as a genetic-environmental interaction.  We also attempt to develop a new methodology to estimate the genetic component of birth weight heritability that does not rely on twins, adoptees or other samples that might be of dubious generalizability. [1]   This work has been published in a series of journal articles and is now appearing as a book called The Starting Gate: Birth Weight and Life Chances (University of California Press, 2003).

Young Children: What is the relationship between wealth and health among children? There has been much epidemiological research done to investigate the association of children’s health and developmental indicators with parental SES as measured by education level, occupation and income. However, household wealth (i.e. net worth) - which displays a distribution that is more unequal than that for income - has received little attention with respect to child health. In fact, the income and wealth distributions are not very co-linear at all (some research shows a correlation between a multiyear income measure and net worth of around .45). Also, racial wealth differences are substantial even when controlling for income. Wealth and inheritance may be less associated with the attributes that positively predict both labor market success (i.e. earnings) and positive parenting and thus may yield a “purer” estimate of the independent effect of financial resources on child outcomes.

Adolescents:  Within the debate about the SES-health gradient, it has long been regarded as an epidemiological puzzle that the SES gradient in health – which appears particularly strong in early childhood – appears to weaken during adolescence before reasserting itself in adulthood. Various explanations have been offered for this paradox of youth (West, 1997).  One of the strongest possibilities is that both ill health and SES display a lot of measurement error for teenagers.  Parents’ SES may be a poor proxy for adolescent social standing for at least two reasons: 1. there may be significant intergenerational mobility which begins to assert itself in the teenage years.  2. the school may act as a “total institution” and have its own logic of social hierarchy.  As PI, but in conjunction with researchers at Rice University and the University of Texas Health Sciences Campus, I am developing a methodology to address these concerns and to test for status effects on health (as measured in a laboratory setting).

To capture the multiple social hierarchies that may affect the health of teenagers, we propose developing a network-based endogenous scale of self-perceived and other-perceived social status that captures at least three dimensions. Borrowing techniques from network analysis (Bearman, Jones, & Udry, 1997) that have not been applied to the investigation of class hierarchies among youth, we will draw a saturation sample (census) of a particular grade in two high schools and ask each respondent in that sample universe to nominate their five closest friends in that universe, i.e. the grade. Using a visual analogue scale (VAS) approach the respondent will then rank order his or her friends on a twenty-five rung ladder as to how close they are to ego. (For an example of this visual “ladder” approach, please see Goodman et al. 2001.)  Such an approach has been found to be effective with the measurement of subjective phenomena (Bond, Shine, Bruce, 1995; Gift, 1989) and to provide significantly greater variance than when the same scale is responded to using a discrete form, such as a Likert-type scale (Pfennings, Cohen, & van der Ploeg, 1996). [2]

We then ask him/her to rank these nominees and him/herself from top to bottom on a twenty-five rung ladder on the following three dimensions (in addition to others that may emerge from focus groups):