Work over welfare:
lessons from America

What are the key US findings?

The intended consequence and leading indicator of success of the 1996 reforms was to be falling welfare caseloads – and in this respect they have been an unqualified success – see Figure 1. But to rely on this is to repeat the error made by Tommy Thompson, Governor of Wisconsin, in believing that the reduction of welfare rolls is the same as a reduction in poverty.14 In fact, caseloads began falling as early as 1994 and in 2005 reached the lowest they have been since 1966.15 It is certainly worth noting here that the fall started well before the 1996 reforms, probably due to an upturn in the economy.

Employment up, welfare down

Employment rates have also risen, but from a reasonably high starting point (see Figure 3). Before 1996, employment grew for both married and lone mothers to 65 per cent of married mothers and 63 per cent of lone mothers.16 In the 1970s lone mothers were more likely than married mothers to be employed. Rates for single (never-married) lone mothers and those with children under six are lower. Unfortunately, the numbers are not strictly comparable to UK employment rates for lone parents, because figures for single mothers in the US include cohabitating single mothers.

Ron Haskins of the Brookings Institute (recently quoted in an Observer article by Will Hutton17) concludes that falling welfare rolls, increased employment rates among single mothers and the increase in the proportion of income from earnings rather than benefits represents ‘a triumph for the federal government and states – and even more for single mothers’.18 He claims raised employment rates for low-income lone mothers, up from 58 per cent to nearly 75 per cent and from 44 per cent to 66 per cent for single lone mothers (he does not define low income). This, despite child poverty increases since 2000. But he acknowledges the role played by a strong economy and the childcare, Medicaid, tax reliefs and credits designed to help low-income families.

'No work, no welfare'

Parrott and Sherman acknowledge that outcomes are usually discussed in terms of falling TANF receipts and both rising employment rates and child poverty reduction in the 1990s.19 But, they argue, this ignores important factors, such as subsequent child poverty rises, particularly for those below half the poverty line. And, as fewer now claim TANF, it now fails to protect the most vulnerable. Employment rates for lone parents are now higher than in the mid-1990s but they have fallen since 2000. Rates rose from 61.7 per cent in 1995 to 73 per cent in 2000, falling back to 69.1 per cent in 2005. And, typically, these working lone mothers remain poor, or near-poor. Also, the number now both jobless and receiving no cash assistance has increased significantly, so despite one million more lone mothers in paid work, roughly one million mothers, with two million children, now fall into this ‘no work, no welfare’ group. And, TANF helps fewer families – by 2002 only 48 per cent of eligible families qualified, down from 80 per cent in the early 1990s (see Figure 4). So, it is not that fewer families are poor enough to qualify, but that TANF is no longer available to them. This drop in TANF participation accounts for more than half – 57 per cent – of the decline in caseloads since 1996. While lone parents in work but not claiming welfare rose from 32.2 per cent in 1996 to 46.6 per cent in 2004, the ‘no work, no welfare’ group rose from 16.4 per cent in 1996 to 32.6 per cent in 2004 – see Figure 5 below.

Added to this, many immigrants are no longer eligible for assistance such as Medicaid and food stamps during their first five years in the US, and many, including their children who are US citizens, continue to shun benefit programmes, partly for fear of deportation or loss of citizenship rights.20 Nearly one-quarter of poor children have an immigrant parent and they make up a significant portion of low-paid workers – benefit claimant rates are now well below that of the general population.

Mixed results

Studies find that most lone parents enter low-paid work, with limited earnings growth and employment instability. There is also a group of lone parents who remain unemployed after leaving assistance with little evidence overall of any improvements in child wellbeing.21 The results are commonly described as ‘at best mixed’. For example, Mark Greenberg of the Center for American Progress in Washington points out that it is clear in the 1990s employment went up and child poverty went down for lone-parent families, but he contends that there is neither any consensus on what role the welfare changes played, nor on how to disentangle the effects of changes in welfare from the role of the strong economy, and the expansions of other supports.22 In his view, it is also clear that during this time welfare caseloads fell much more than employment increased or child poverty declined (see Figure 6), and there has been an increase in the ‘disconnected’ population of families not in work and not in welfare. Also, those most likely to be sanctioned off the welfare rolls were those least likely to be able to get and keep a job – eg, the lone parents furthest from the labour market, without unemployment insurance, lacking skills and qualifications, who may be sick or disabled or have children who are, who may have large families and additional problems perhaps including mental illness or drug addiction.

According to Ron Haskins, the number of mothers in the poorest fifth of income who report zero earnings and zero welfare-income continues to rise year on year.23 Moreover, in the current decade, child poverty has grown and lone-parent employment has fallen, even as welfare caseloads continue to decline. The number and proportion of poor children getting any help through TANF continues to fall (see Figures 6 and 7).

Sustainable jobs?

The US evidence on job retention suggests that although a high proportion (71 per cent) are in work at some point in the year after they left TANF and three-fifths are working in both the first and fourth quarters of the same year, very few (37 per cent) were in work for every quarter of the year.24 This suggests a lot of people must be cycling in and out of work. The evidence below for Milwaukee confirms this pattern for every year since the 1996 reforms. Claimants may be getting jobs and leaving TANF, but they do not seem to be staying in them for very long.

Child wellbeing

The debate also raises concerns about child wellbeing and the impacts of maternal employment. Pamela Morris analysed seven random assignment trials across thirteen programmes looking at children aged two to nine.25 She found positive impacts on school achievement for children five and under but these were confined to work programmes that increased family incomes through earnings supplements. The impacts faded after three years. These positive impacts were related to attendance at centre-based childcare during pre-school years. But data for adolescents (aged ten to sixteen) is not so encouraging, showing below average school performance and slightly increased likelihood of repeating a grade or needing special education classes.26 The adolescents concerned seemed to be those with younger siblings, suggesting they may be taking on early childcare responsibilities because of their mothers’ work. Documented levels of hardship for TANF applicants have also significantly increased – see Figure 9 below.

TANF re-authorisation

The 1996 Act came up for review and re-authorisation in 2002, but only after considerable debate did the Deficit Reduction Act finally officially re-authorise TANF last year. It gave states new requirements to raise work participation rates while narrowing the range of activities that can count towards this. It is now harder to tailor welfare-to-work packages to meet individual requirements, therefore, giving strong incentives to help fewer families and further reduce caseloads. Key academics are arguing that the troubling trends of increased child poverty and ‘no work, no welfare’ families stand to increase following re-authorisation. This all illustrates how hard it is for the block grant approach to respond in times of economic downturn.

The re-authorisation debate has been beset with rows about large state budget deficits that impede innovation and an inability to reach agreement over the nature of future work requirements, childcare spending and the role of government in promoting marriage. The Duncan-Smith ‘Social Justice Commission’ report from UK Conservatives, published earlier this year, with its emphasis on the importance of marriage, is redolent of last year’s TANF re-authorisation debate in the US, with George Bush’s emphasis on his ‘Healthy Marriage Initiative’ and promoting responsible fatherhood as part of the block grant subsidy. Haskins, indeed, recommends a ‘National Marriage Movement’.

Next steps

Mark Greenberg concludes the research is clear that new approaches need to focus on building skills, job quality and increasing family incomes.27

• ‘The most effective programs do not simply focus on job search or basic education, but blend a mix of individualized employment, training and other services with attention to local labor markets;
• retention and advancement are affected by wages and other aspects of the quality of initial job placements, the premise that any job is as good a starting point as any other job is not true; and
• to see improvements in child well-being, it’s important to connect families with stable quality child care and ensure that employment translates to increased family income.

Unfortunately the reauthorisation Bill enacted by Congress reflects almost none of these key lessons.’

And, following reauthorisation, the focus remains on cutting caseloads.

Notes

14. See note 2
15. See note 1
16. See note 5
17. W Hutton, ‘Britain Would Benefit from Clinton’s Tough Love’, The Observer, 3 September 2006
18. B Haskins, Welfare Reform, Success or Failure? It worked, Policy and Practice, American Public Human Services Association, 2006, available at www.aphsa.org/Publications/Doc/PP/0603ART1.pdf
19. See note 8
20. See note 8
21. See note 5
22. M Greenberg, Welfare Reform, Success or Failure? With Mixed Results, Policy and Practice. American Public Human Services Association, available at www.aphsa.org/Publications/Doc/PP/0603ART1.pdf
23. See note 1
24. See note 10
25. P Morris, L A Gennetian and G J Duncan, ‘Effects of Welfare and Employment Policies on Young Children: new findings on policy experiments conducted in the early 1990s’, Social Policy Report 19, pp3-22, 2005
26. L A Gennetian and others, How Welfare and Work Policies for Parents Affect Adolescents: a synthesis of research, MDRC, 2002
27. See note 22

Contents

Executive summary

Section 1
The Clinton reforms
Lessons for the UK
Americal welfare benefits
The 1996 reforms: the Personal Responsibility and Work Opportunity Reconciliation Act Key changes
The broader context in which TANF was implemented
US poverty measurement

Section 2
What are the key US findings?

Setion 3
The UK context
What are the key UK findings?

Notes

Print the entire document