PRRAC Poverty & Race Research Action Council
Home About PRRAC Current Projects Publications Newsletters Resources Contact Us Support PRRAC Join Our Email List

"New Study Disregards Racial and Gender Dynamics in Predicting Child Poverty,"

by Danielle Hayot is an article on a new study that disregards race and gender when predicting child poverty.July/August 2001 issue of Poverty & Race

A new report by Robert Rector, Kirk Johnson and Patrick Fagan of The Heritage Foundation, “Understanding Differences in Black and White Child Poverty Rates” (May 2001), argues that child poverty is caused by welfare dependence and single motherhood, and that black children are more likely to live in poverty than white children because they are more often raised by single parents who are long-term welfare recipients. The Heritage Foundation paper concludes that race is not an important factor in predicting child poverty, because race effects disappear when controlling for other factors. None of these claims about poverty and race are supported by the data.

The authors’ conclusions are unsound for a number of reasons. First, the analysis excludes important predictors that research finds are consistently associated with transitions in and out of poverty. Second, the analysis does not acknowledge the systemic disadvantage that links race and poverty, and neglects a body of literature on discrimination’s effects on employment outcomes. Third, the authors do not adequately address the reasons that many women must rely on welfare for economic support, and the implications of these factors for the possibility of achieving self-sufficiency through employment. Fourth, the authors’ claim that welfare and single parenthood cause poverty is not supported by their analysis, and shows a gross misunderstanding of when causation can and cannot be imputed through statistics.

Neglect of Root Causes of Poverty

The authors use a subset of data from the National Longitudinal Survey of Youth (NLSY), but not nearly enough variables to thoroughly consider a full range of likely poverty indicators. Their narrow selection of variables is significant, given the wealth of information available in the NLSY. The independent variables studied in the Heritage analysis are: percent of life lived in a single-parent family and in a family receiving AFDC; mothers’ math and verbal skill level; average age of mother at first birth; number of children born to the mother; residence in rural area; race; and time spent living in the South, Northeast, North Central or West. The authors don’t measure variables assessing the prevalence of barriers to employment, such as job discrimination, lack of quality and affordable child care, domestic abuse, mental or physical health problems, lack of job training and lack of access to transportation, all of which have been shown to negatively affect employment prospects and earnings.

Rather than addressing why single-mother households are poorer on average than two-parent households, the authors attack single mothers for seeking government support and “falling prey” to a welfare system that “entices” women out of the labor market and into welfare, thus “causing” poverty. Their analysis overlooks the reality that many of these women turn to welfare as a last resort, when they are unable to make it in a system with limited employment options and low wages. Welfare leaver studies have consistently documented that the post-welfare employment experiences of women result in poverty or near-poverty conditions, and a 1999 study by the Urban Institute found that welfare leavers are earnings on average only $6.61/hour.

Race as an Indicator of Poverty

Black women are especially vulnerable to negative labor market experiences. Although the Heritage Foundation authors reject this argument, claiming that black and white wages are “virtually identical,” holding education, skill levels and work experience equal, they fail to acknowledge discriminatory practices that affect the ability of black women to acquire the skills needed to be placed in good-paying jobs. The earnings of all working black and white women, allowing for differences in human capital, show that white women are earning substantially more than their black counterparts. The assumption that discrimination does not exist “holding everything else equal” discounts the impact of generations of disadvantage accrued by black men and women in this country. The fact is that everything is not equal —assuming so presents a narrow and uninformed reality of race relations in America today.

The authors’ claim that black men’s earnings are 80 percent of white men’s earnings only partially explains black men’s labor market experiences. By limiting their presentation of earnings data exclusively to workers and excluding the influence of zero-earning non-workers, the authors over-estimate the black-white earnings ratio. Since black men experience significantly higher unemployment rates than white men, limiting the analysis to comparing the earnings of employed men presents an inaccurate view of blacks’ earnings potential. The perceived influence of marriage on earnings, therefore, is eliminated when one considers that the same forces that account for the difference in black and white men’s earnings potential exist regardless of the institution of marriage. Studies linking race and labor market experiences are not hard to find. The Urban Institute found that black males ages 19-24 years were three times more likely to experience discrimination than their equally qualified white male counterparts. A likelier explanation for the link between black male earnings and marriage is that poverty, driven in part by discrimination, negatively affects earnings, which leads to decreased rates of marriage as the earnings potential of two-parent families falls. Economists William Darity, Jr. and Samuel Myers, Jr. conclude that if blacks and whites were treated equally, blacks would have “a drastically reduced incidence of female-headed families, lower expected welfare income, and thus higher work effort and higher earnings.”

Heritage Misconstrues Reasons Women Resort to Welfare

Single women raising children can remain economically self-sufficient only with great difficulty, not only because of the influence of race on earnings potential but because of the specific difficulties women face in the labor market. It is not surprising that female-headed families with children under 18 made up 50 percent of the poor in 1997. The Heritage report explains this by assuming that poor women would rather be on welfare than work, citing a study that links an increase in the value of the welfare benefit “package” with an increase in welfare caseloads. As Jared Bernstein of the Economic Policy Institute points out, the correlation between welfare caseloads and the monetary value of welfare does not hold over all time periods. When the real value of AFDC benefits started decreasing around 1972, caseloads did not decrease. State data also refute the authors’ argument. If caseloads and the value of the benefits package were positively correlated, we would expect states with greater benefits to have higher caseloads, which is not the case. Considering that welfare provides income support equivalent to only 62 percent of the poverty line in the median state, we would expect policymakers and public officials to pause at the thought that women’s labor market experiences are so dismal that poverty conditions are more appealing than paid work. Women in fact turn to welfare because labor market and family support policies have not adequately addressed their needs.

Statistical Association Causation

Lastly, in applying welfare use as an explanatory variable, the Heritage Foundation report authors essentially claim that a program created to alleviate poverty is in fact causing poverty. One of the basic tenets of social science research is that association does not equal causation. If it did, the research community would be inundated with similarly dubious declarations claiming, for example, that Social Security causes retirement or that unemployment insurance causes unemployment. If the authors were truly concerned with determining predictors of poverty, a more thorough approach would have been to incorporate more variables from NLSY into their statistical analysis.

In short, the Heritage Foundation study lacks insight into the race- and gender-specific dynamics that affect women’s ability to move out of poverty. Rather than promoting policies that increase women’s earnings, the authors advocate shifting women’s dependency from one entity to another. The study also fails to address the systemic devaluation of mothering that has been integrated into policies targeting poor women. The emphasis the authors place on eradicating single motherhood and welfare dependency disregards the complexities of how people become poor and stay poor and the choices they make between raising children and working full-time without reliable child care; staying in physically or emotionally harmful marriages; and working in low-wage jobs with no employee benefits. Acknowledging these complexities would allow for a more profound understanding of poverty and serve as a better tool in formulating policies and programs that will make a difference in the lives of today’s poor women and children.

Danielle Hayot is the Research Program Coordinator at the Institute for Women’s Policy Research (1707 L St. NW, #750, Wash., DC, 20036, 202/785-5100)

Join Our Email List
Search for:             
Join Our Email List