Civic Report No. 35 March 2003 Gaining Ground, Moving Up: The Change in the Economic Status of Single Mothers Under Welfare Reform ENDNOTES See June E. O’Neill and M. Anne Hill, Gaining Ground? Measuring the Impact of Welfare Reform on Welfare and Work, Manhattan Institute Civic Report No. 17, July 2001. That gain is less than the gain in single mothers’ own cash incomes because non-cash benefits declined slightly and the incomes of other household members did not rise as fast as those of single mothers. The poverty rate for single mothers in 1996 noted here (40.1%) and throughout the report differs somewhat from the rate shown in Figure 1 (41.9%) because it is restricted to families with a single mother, ages 18–44, living with her own children. Figure 1 includes all female headed families with any related children but no husband present. In addition, our definition of single mothers includes heads of subfamilies as well as heads of primary families as long as they are living with children of their own. The only annual poverty data available for early years refer to the broader category of female-headed families. The proportion of single mothers living with a male partner increased from 7.6% in 1995 to 11.3% in 2001. (See Appendix Table C-1 for the distribution of single mothers by living arrangements.) We identify single mothers with a male partner in the Current Population Survey as a single mother living with one adult male aged 18–60 and with no other adults present. The maximum value of the EITC credit increased substantially over time, particularly for families with two children, for whom it rose by 130% in real terms from 1992 to 1996, reaching a level of $4,015 in 1996 (in 2001 dollars). For valuing non-cash benefits we use the estimates of the Census Bureau provided in the Current Population Survey. Non-cash benefits include Food Stamps, Medicaid, Medicare, free and reduced-price lunches and rent subsidies. The Bureau estimates a “fungible” value for Medicaid and Medicare which assigns no value to these health benefits if family income is below the level they determine is needed to cover food and housing requirements. At incomes above that need standard, the fungible value is the difference between the market value of Medicaid (or Medicare) and the need standard. The Bureau gives cash equivalents for other non-cash benefits essentially based on estimated market value. Estimates are also given for the full cash value of health insurance supplements to wage and salary income provided by employers—a non-means tested part of compensation. See the discussion below and Appendix B. See O’Neill and Hill, Gaining Ground, op.cit. The rise in poverty from 1995 to 1997 is pronounced only for white non-Hispanic women (Fig. 5). The SIPP surveys are complex in their design. We use the panel first interviewed in April, 1996. Members of the panel are divided into four groups, each of which is interviewed in a different month in a four-month wave. (There are 12 waves, three a year.) But in each interview the panelists are asked to recall their incomes, employment and other information for each of the past four months. Thus the first group of the first wave, interviewed in April, 1996, was asked to report on each month, from December, 1995 through March, 1996. By the 12th and last wave of the survey there was a loss from the sample of approximately 30% of women who had children and were ages 16–50 in the first survey interview. Attrition reduced the size of the sample and imposed some distortion since certain groups were more likely to leave the survey than others. The Census has attempted to correct such bias by adjusting the population weights and we use that weighted data in our analysis. For several reasons, we have broadened the number of months for defining welfare status beyond the usual one or two-month standard. One is that we are primarily interested in women with children who have had a real attachment to welfare, and who then actually leave the program. Welfare recipients can be reported off welfare for one or two months, however, due to an administrative problem, such as a temporary sanction or even a clerical error. In that case, their work participation may not change, particularly if they expect to go back on welfare quickly. Second, monthly changes cannot be accurately determined in SIPP because of what may be termed a “seam” problem. Panel members provide data for the past four months based on memory. When they can’t recall monthly detail, they tend to report the same status in all four months, whether for welfare receipt, employment or other items. Changes in reported status, therefore, tend to occur only at the next interview—i.e., at the seam. Our examination of the data suggests that the seam problem occurred frequently. As noted, it is not uncommon for those who leave welfare to return, particularly in the first year after leaving, and then to leave again. We have not excluded returnees from the income analysis. Since their poverty rates are generally higher than those of leavers who do not return, their inclusion in the data raises the overall poverty rate for leavers. However, the proportion of leavers who are returnees declines with time. Few returnees are found on welfare in the last year observed among the 1996 cohort of leavers. Based on cash income alone, the percent in poverty in the last six months before exiting welfare ranged from 58% to 74%. In most cases, poverty was already lower in the last months before leaving welfare than it had been earlier. As we show below, the work experience and incomes of women on welfare in recent years have increased in the last six months on welfare. This pattern likely reflects the effects of both the work requirements of welfare reform and the enhanced earnings disregards that in many states allow recipients to keep a larger portion of their benefits while they work, but only for a limited time. Only one cohort experienced little or no immediate improvement—those leaving after the second half of 1997. But the poverty rate of that cohort had already declined in the last six months on welfare from an earlier, higher poverty rate. Moreover, this cohort, like the others, experienced additional declines in welfare as time went on. In 1999, the CPS poverty rate for single mothers who were not on welfare was 26.3% based on cash income and 16.9% based on household cash and non-cash income post-tax (Figure 3). The poverty rate for SIPP leavers in their fourth year off welfare was 31.8% based cash income only and 19.2% on a household basis (Table 1). See Gaining Ground, op.cit. Between 1992 and 1996 the maximum annual EITC benefit for a single mother with two children increased from $1,747 to $4,015 (in constant 2001 dollars), a gain of 130%. The benefit for a single mother with one child increased by only 45%. The basic data source for the regression analysis is the monthly Current Population Survey (CPS) micro-data files for the years 1984–2001, merged with data on the maximum EITC for one and two-child families, the maximum combined state benefit from AFDC/TANF and Food Stamps, hourly wage rates and unemployment rates by state, and the month of implementation of welfare waivers and TANF by state. The regression specifications are detailed in Appendix Table A. Work participation is measured here as the percent currently employed, based on monthly reporting of employment status. For the presentation of results the data were averaged over six-month periods. The regression analysis indicated a statistically significant positive effect of the EITC for women with two or more children, offset by a negative effect of the EITC for those with one child (see Appendix Table A). The net effect of the EITC on employment was smaller than we had anticipated. However, the EITC has a complicated structure that may discourage additional employment as earnings rise above particular levels. In part this may reflect moving off welfare and on to work during a part of the year, since welfare receipt as measured in the CPS only refers to receipt at any time during the year. It also may reflect increased work in anticipation of leaving welfare as well as increases in earnings disregards that have occurred in several states. Data from the National Longitudinal Survey of Youth, 1979 cohort (NLSY79), show that among a cohort of women ages 33–41 in 1998 who had ever been on welfare, 17% were high school dropouts. There is a vast literature on this point. In particular see the fundamental works of Jacob Mincer: “Labor Force Participation of Married Women” in H.G. Lewis, ed., Aspects of Labor Economics, 1962, Cambridge: National Bureau of Economic Research (NBER); and Schooling, Experience and Earnings, 1974, NewYork: Columbia University Press for NBER. Considerable attention was drawn to the finding of a decline in the income of single mother families in the lowest percentiles of the income distribution between 1995 and 1997 and that decline was attributed to the after-effects of welfare reform. (See Wendell Primus, L. Rawlings, K. Larin and K. Porter, The Initial Impact of Welfare Reform on the Income of Single-Mother Families, Center on Budget and Policy Priorities, Washington DC, August, 1999.) Our analysis suggests that reporting of income in the lowest quintile is unreliable and subject to random fluctuations. Moreover the change in income from 1995 to 1997 is not appropriate for evaluating welfare reform since in many states it had only been implemented for a portion of 1997. In any event, both the own cash income and total household incomes of the poorest quintile of single mothers increased between 1997 and 2000 and were higher in 2000 than they were in 1995. While on welfare, mothers and their children are typically covered by Medicaid. However, an increasing proportion of families have maintained Medicaid coverage after leaving welfare through programs that provide transitional benefits to welfare leavers and through provisions that have allowed states to cover low-income families with no ties to welfare. The State Children’s Health Insurance Program (SCHIP) provides federal matching funds to states and has given them the leeway to extend eligibility to families above the poverty line. Initially, states expanded coverage only for children. More recently many states have extended coverage to parents, although usually subject to stricter income limitations. See Matthew Broaddus, et al, Expanding Family Coverage: States’ Medicaid Eligibility Policies for Working Families in the Year 2000, Center on Budget Policy and Priorities, February, 2002. The statistics on welfare duration in this paragraph were calculated from the NLSY79 and refer to women who remained in the sample through 1998. Welfare use was reported each year for the prior year starting in 1978 when the women were ages 14–22. In years when the interview followed a skipped year information was obtained retrospectively for the past two years. This estimate is based on regression analysis of data from the National Longitudinal Survey of Youth, 1979 cohort (NLSY79). See Appendix Tables D-1 and D-2 which provide details on differences in characteristics and the underlying regression results on which the estimate is based. Additional analysis indicates that the returns to both general work experience and tenure are about the same for the two groups of women. Thus one can have greater confidence that narrowing the work experience gap will narrow the wage gap as well.

EMAIL THIS | PRINTER FRIENDLY CR 35 PDF (370 kb) WHAT THE PRESS SAID: Reforming Welfare The Daily Oklahoman, 4-14-03 Mike: I'll Sue 'Em New York Post, 4-14-03 Welfare Reform Works by June O’Neill, New York Post, 4-14-03 Moving up by Linda Chavez, Townhall.com, 4-9-03 Equal Time: To help single mothers, strengthen welfare reform by June O'Neill, Atlanta Journal-Constitution, 4-1-03 SUMMARY: This report examines how the economic circumstances of single mothers changed after welfare reform. The primary author, Dr. June O’Neill, found that the poverty rate for single mothers dropped by roughly 20% following reform, reaching an all-time low in 2000. The proportion of single mothers who were employed also surged during this period, as did their income. Single mothers’ own-cash income rose 21% overall, and rose by similar margins even in the most disadvantaged demographic groups. The report also found that single mothers’ earnings rise significantly over time as long as they remain in the workforce. TABLE OF CONTENTS: Executive Summary About the Authors Acknowledgements Index of Figures and Tables Introduction I. POVERTY Differences Between Welfare Recipients and Non-Recipients Education, Race and Marital Status Did Poverty Decline for Welfare Leavers? II. WORK PARTICIPATION Accounting for the Increase in Employment III. EARNINGS, WAGE RATES AND OTHER ASPECTS OF EMPLOYMENT Hourly Wages of Single Mothers Occupations and Sector of Employment The Employment and Earnings of Welfare Leavers: Findings from Panel Data Will the Wage Rates of Welfare Leavers Rise with Experience? IV. CHANGES IN TOTAL INCOME AND ITS COMPONENTS Changes in Household Cash and Non-Cash Income in the SIPP Panel V. CONCLUDING COMMENTS Endnotes Appendix Tables