The University of Michigan
555 South Forest Street
Third Floor
Ann Arbor, MI 48104-2531

T 734-936-9842
F 734-998-6341


Women with less than a high school education experienced a decline in health insurance coverage over the 1990's, with the share uninsured rising from 27.0% in 1988 to 35.0% in 2000. Did welfare reform contribute to this loss? Thomas DeLeire, Judith A. Levine, and Helen Levy, in a study funded by the Economic Research Initiative on the Uninsured, find welfare reform partially offset the decline in coverage for women with less education. Without welfare reform, more women would have been uninsured.

Descriptive statistics

  • Among all women, the percent uninsured rose from 14.5% in 1988 to 18.5% in 1998 and fell to 16.5% in 2000. Trends among those with a high school education (63.6% of adult women) and less than a high school education (14.7% of adult women) drive this increase. College educated women experienced little change in uninsurance (increasing from 7.3% in 1988 to 8.6% in 2000). Percent uninsured among those with a high school education increased from 13.6% in 1988 to 16.1% in 2000, and the percent uninsured among those with less than a high school education increased from 27.0% in 1988 to 35.0% in 2000.
  • The decline in coverage among women with less than a high school education was largely a decline in private coverage, as the share with private coverage dropped from 49.6% in 1988 to 43.0% in 2000 while public coverage fell from 23.4% to 22.0% over the same period.
Effect of change in demographic composition
  • Controlling for demographic characteristics (including age, marital status, employment status, race and ethnicity, and state of residence) explains roughly half of the increase in the fraction uninsured among women with less than a high school degree, leaving a 4.4 percentage point drop over 1988-2000 unexplained by the characteristics controlled for.
Effect of welfare waivers and TANF implementation on health insurance coverage.
  • Among women with less than a high school education, being in a state with a welfare waiver in effect was associated with a 2.3 percentage point higher probability of having coverage (SE: 1.0) and a 3.6 percentage point higher probability (SE: 1.4) when in a state which had implemented welfare reform (TANF), relative to states with no reforms.
Effect of welfare reform on subgroups of women with less than a high school education.
  • Estimating the baseline equation across subgroups (single mothers, single with no children, and married with children; white non-Hispanic, black non-Hispanic, and Hispanic), waivers are associated with a 2.2 percentage point increase (SE: 1.3) in the probability of being insured among married mothers and TANF is associated with a 10.4 percentage point increase (SE: 2.6) in the probability of being insured for unmarried women without children.
  • Both waivers and TANF implementation were associated with an increase in the probability of being insured for Hispanic women but not for non-Hispanic black women.


In an era when the health insurance prospects of less educated women were declining, welfare reform provided a modest offset. Welfare reform, beginning with state waivers, set off a large increase in labor supply by less skilled women which may have led many to obtain jobs with health insurance. Policy steps that increase labor supply may reap higher levels of private health insurance as a side benefit.

Linear probability models with insurance status, private coverage, and public coverage as dependent variables.

The Temporary Assistance to Needy Families (TANF) program was implemented as a successor to Aid to Families with Dependent Children (AFDC) in all states in either 1996 or 1997. This narrow window provides only a small range of variation for estimating the effect of welfare reform, making the discernment of TANF's effect - whether positive or negative - less likely.

The welfare waivers are described at a gross level - either "on" or "off" - leaving just what it was about welfare waivers that had an effect an unanswerable question.

CPS as a data source means variables about marital status, education, and presence of children are measured as of the time of the survey while health insurance and employment reflect the previous year.

A large portion of the decline in coverage remains unexplained. For example, while the descriptive statistics show the largest decline in private coverage occurred over 1988 to 1992, the first welfare reform waivers included in the analysis took effect in 1993.

The coefficients measuring the effects of reforms on public coverage fell by about one half and became statistically insignificant when health insurance variables were coded as of the date of the survey rather than the previous year reference concept in the CPS. Other results were not affected by this recoding and were robust to a range of other sensitivity checks.

March Supplement to the Current Population Survey, 1989 to 2001. Sample restricted to women ages 18 to 64. The overall sample is 580,364 women; 91,124 have less than a high school degree. Variables coding implementation of welfare waivers and TANF implementation from Department of Health and Human Services (HHS.)

Is Welfare Reform Responsible for Low-Skilled Women's Declining Health Insurance Coverage in the 1990's?
Thomas DeLeire, Michigan State University, Judith A. Levine, University of Chicago, and Helen Levy, University of Michigan and University of Chicago

Conference paper presented at ERIU Research Conference, July 2002.

The final version of the paper is forthcoming in the Summer 2006 issue of the Journal of Human Resources.

ERIU Working Paper #22

Back to top

Funded by The Robert Wood Johnson Foundation, ERIU is a five-year program shedding new light on the causes and consequences of lack of coverage, and the crucial role that health insurance plays in shaping the U.S. labor market. The Foundation does not endorse the findings of this or other independent research projects.