You are here

After 1996 Welfare Law, a Weaker Safety Net and More Children in Deep Poverty

August 9, 2016

careful assessments indicate that TANF has fueled a notable increase in “deep poverty”As we approach the 20th anniversary on August 22 of the 1996 welfare law that created the Temporary Assistance for Needy Families (TANF) block grant, some policymakers maintain that TANF has been a clear success and that other anti-poverty programs should adopt its block-grant structure and rigid work and other requirements.  As we’ve explained, however, the research shows TANF’s record has been mixed.[1]  The child poverty rate did fall in the years following TANF’s creation, though that decline was likely driven more by the strong economy and reforms to the Earned Income Tax Credit and Child Tax Credit than by the welfare law.  At the same time, a dwindling share of families in need received TANF benefits, inflation eroded the benefit levels, and states diverted limited TANF block grant funds away from basic income assistance.  As a result, careful assessments indicate that TANF has fueled a notable increase in “deep poverty” — that is, in the share of Americans who live below half of the poverty line.

A 2015 Center analysis examined changes in deep poverty over the welfare law’s first decade, from 1995 to 2005.[2]  Those are appropriate years to analyze because they had comparable overall economic conditions and represented similar points in the business cycle.[3]  Our analysis found:

  • The share of children below half of the poverty line rose from 2.1 percent to 3.0 percent between 1995 and 2005, and the number of children in deep poverty rose from 1.5 million to 2.2 million.[4]
  • This rise in deep poverty is directly attributable to a weakening of the safety net for the poorest families.[5]  If government income support programs had kept the same share of otherwise deeply poor children out of deep poverty in 2005 as they did in 1995, the children’s deep poverty rate would have fallen, not risen.

The rise in deep poverty in the decade after the welfare law contrasts with children’s experience in the next five years, a period that included the worst recession in decades.  Surprisingly, deep poverty rates for children fell somewhat in this period, from 3.0 percent in 2005 to 2.6 percent in 2010.  (Comparable data are not available past 2010.[6])  The decline is largely attributable to strong temporary expansions in food assistance and unemployment insurance that the federal government enacted in response to the Great Recession and illustrates the important influence of policy choices on deep poverty trends.

Our analysis goes beyond traditional Census data in several ways:  we count non-cash benefits and taxes in our income measure, and we correct for the undercounting of government benefits using the Urban Institute’s TRIM model, the best available method for making such corrections.  Counting non-cash benefits is favored by researchers and analysts across the political spectrum.[7]  Correcting for undercounted benefits is also very important, because households often neglect to fully report their income in Census surveys, and underreporting has a particularly strong effect on deep poverty levels and trends.[8]

Using these same data, we also recently extended the analysis to look more closely at single-mother families — those whom welfare policies affect most. 

  • In 1995, 2.8 percent of children of single mothers lived in deep poverty.
  • By 2005, this figure had more than doubled, to 5.8 percent.

We also explored alternative ways of measuring inflation.  When adjusting the poverty line from year to year, our analysis generally uses the same inflation adjustment that the Census Bureau employs: the Consumer Price Index (CPI).  The government has multiple inflation adjustments, however, and some analysts prefer an adjustment known as the Personal Consumption Expenditures (PCE) deflator.  When we use the PCE deflator to adjust the poverty line, the data show that the deep poverty rate among children of single mothers rose from 2.8 percent in 1995 to 5.2 percent in 2005 — slightly less dramatically than under the CPI — but still nearly doubled. 

Other studies confirm the rise of deep poverty in the wake of the welfare law.  For example:

  • Academic researchers Yonatan Ben-Shalom, Robert Moffitt, and Karl Scholz report a rise in deep poverty rates for all ages, from 4.5 percent of U.S. families in 1993 to 6.6 percent in 2004.[9] 
  • Harvard’s Christopher Jencks observes that economic well-being for extremely poor Americans (measured at the second-lowest percentile of the population) has fallen significantly in the last two decades — evidence, he says, that “the poorest of the poor were a lot worse off in 2012 than in either 1996 or 1999.”[10]
  • An authoritative 2015 literature review on the 1996 welfare law from the National Bureau of Economic Research concluded that “declines in welfare benefits arising from leaving welfare often cancel out the earnings increases, leaving income relatively unchanged (…with some families experiencing income increases and others, decreases).  Especially in more recent years, in addition, a significant number of single-mother families appear to have been made worse off and to have higher deep poverty rates.”[11]
  • Early 1990s welfare-to-work pilot programs — local forerunners of the work requirements in the national welfare law — also tended to fuel deep poverty even while lowering poverty rates overall.  In a federally funded evaluation of 11 programs, all of the programs raised employment rates somewhat in the short term and several reduced poverty.  Yet deep poverty rates rose significantly in more than half of the programs (and fell in none), and the majority of the pilot programs raised the share of families with income from neither work nor welfare.[12]  These findings reflect rigorous comparisons between program participants and control groups randomly assigned to a more traditional welfare program, thus revealing the pilot programs’ impact distinct from the shifting economy or other policies.
  • The number of U.S. children in families with monthly cash incomes below $2 per person per day doubled from 1996 to 2011, according to an analysis by H. Luke Shaefer and Kathryn Edin.[13]

The increase in financial hardship among the poorest families reflected the declining availability of TANF and the eroding value of TANF payments.  While 76 families received cash assistance through Aid to Families with Dependent Children (TANF’s predecessor) for every 100 poor families with children in 1995, by 2014, only 23 families received TANF cash assistance for every 100 poor families with children.[14]  Moreover, those benefits have lost value.  In 35 states and the District of Columbia, the maximum monthly TANF cash benefit payment for a family of three in July 2015 was at least 20 percent below its 1996 level, adjusted for inflation.[15]

TANF’s tough work requirements, time limits, and block grant structure contribute to limiting the amount of assistance the program provides.  TANF’s frozen block grant funding level erodes further with inflation every year.  The TANF structure also gives states broad latitude to spend TANF funds on priorities other than cash assistance.  States spent only 26 percent of federal and state TANF funds on basic monthly cash assistance in 2014.[16]

As more evidence of TANF’s effects has emerged, researchers have adjusted their assessments of welfare reform, and many are concluding that TANF is not the unvarnished success sometimes claimed.  As Ron Haskins, who led much of the drafting of the welfare law as a senior Republican congressional aide in the 1990s, testified in 2015, deep child poverty has increased since the law’s passage and “some poor mothers and their children have neither cash welfare benefits nor earnings.”  Said Haskins, “We should acknowledge this problem and try to figure out ways to help these disconnected mothers and their children.”[17]

End Notes

[1] Robert Greenstein, “Welfare Reform and the Safety Net,” Center on Budget and Policy Priorities, June 6, 2016,

[2] Arloc Sherman and Danilo Trisi, “Safety Net for Poorest Weakened After Welfare Law But Regained Strength in Great Recession, at Least Temporarily,” Center on Budget and Policy Priorities, May 11, 2015,

[3] Overall unemployment rates were similar in 1995 (5.6 percent) and 2005 (5.1 percent) and both years were the fourth full year of an economic recovery.

[4] The poverty line in this analysis was about $25,000 a year in 2010 dollars for a couple with two children renting in an average-cost community.

[5] For more on the erosion of TANF benefits, see Ife Floyd, LaDonna Pavetti, and Liz Schott, “TANF Continues to Weaken as a Safety Net,” Center on Budget and Policy Priorities, updated October 27, 2015,

[6] The Center’s 2015 analysis used estimates of work expenses and rental subsidies — broadly similar to those used in the federal government’s Supplemental Poverty Measure — that are not yet available for years after 2010.

[7] In this analysis of trends, counting non-cash benefits is particularly important.  For example, the Food Stamp Program (now known as the Supplemental Nutrition Assistance Program, or SNAP) protected somewhat fewer children from deep poverty in 2005 than 1995.

[8] Our 2015 analysis corrected for underreporting in every underreported program for which consistent corrections are available since 1995:  TANF, Supplemental Security Income, and SNAP.  TRIM does not provide an underreporting correction for other programs such as unemployment insurance (UI), but rough calculations suggest that including corrections for underreporting in these programs (if the data to do so were available for the 1995-2005 period) would not alter the conclusion that children’s deep poverty rose significantly between 1995 and 2005.

[9] Yonatan Ben-Shalom, Robert A. Moffitt, and John Karl Scholz, “An Assessment of the Effectiveness of Anti-Poverty Programs in the United States,” prepared for the 2012 Oxford Handbook of the Economics of Poverty, Chapter 22.  Figures are based on monthly income and are from the first waves of the 1993 and 2004 Survey of Income and Program Participation, adjusted for underreporting.

[10] Christopher Jencks, “Why the Very Poor Have Become Poorer,” The New York Review of Books, June 9, 2016,

[11] Robert A. Moffitt, “Introduction to Volume 1,” forthcoming in Economics of Means-Tested Transfer Programs in the United States, Robert A. Moffitt, ed. (National Bureau of Economic Research, 2015),

[12] Stephen Freedman et al.National Evaluation of Welfare-to-Work Strategies Evaluating Alternative Welfare-to-Work Approaches: Two-Year Impacts for Eleven Programs (U.S. Department of Health and Human Services and U.S. Department of Education, 2000), Exhibit ES-9, page ES-34),

[13] H. Luke Shaefer and Kathryn Edin, “Rising Extreme Poverty in the United State and the Response of Federal Means-Tested Transfer Programs,” National Poverty Center Working Paper Series #13-06, May 2013,

[14] Ife Floyd, LaDonna Pavetti, and Liz Schott, “TANF Continues to Weaken as a Safety Net,” Center on Budget and Policy Priorities, updated October 27, 2015,

[15] Ife Floyd and Liz Schott, “TANF Cash Benefits Have Fallen by More Than 20 Percent in Most States and Continue to Erode,” updated October 15, 2015,

[16] “Chart Book: TANF at 20,” Center on Budget and Policy Priorities, updated August 5, 2016,

[17] Ron Haskins, testimony before the Committee on Ways and Means Subcommittee on Human Resources, February 11, 2015,