The 1996 legislative reform, which replaced the Aid to Families with Dependent Children programme (AFDC) with the programme of Temporary Assistance for Needy Families, marked a significant shift in US welfare policy. From one perspective, this development can be seen as following a continuous path of step-by-step changes from temporary welfare to temporary work-conditioned welfare, starting from the passage of Title IV of the Social Security Act of 1935. This Act created what was originally called the Aid to Dependent Children (ADC) programme: the means-tested scheme of income support for women who were single parents. In 1935, the New Deal architects of ADC believed that they were designing a temporary measure, since the programme was expected to wither away in several decades as widows with children became eligible for survivors' insurance that was soon to come under the social security system (Witte, Reference Witte1962).

By the 1960s, however, it became increasingly evident that welfare was not as temporary as had been expected. The welfare rights movement aroused expectations that social assistance was a right rather than a provisional benefit. In 1966, mass welfare rights demonstrations were held in 16 cities across the country. A year later, the National Welfare Right Organization (NWRO) was formed, giving voice to a movement that sought to empower poor people, to increase access to assistance among eligible families, and to ensure respect for the legal rights of those on public aid. Welfare recipients were vulnerable to particularly onerous treatment under the ‘man in the house’ rule that mandated the termination of AFDC benefits to any woman with an able-bodied man in the house, and the investigative procedures such as ‘midnight raids’, which were sometimes conducted to check on who might be living with recipients. In 1968, the legal arm of the welfare rights movement was successful in getting the US Supreme Court to strike down the ‘man in the house’ rule (Trattner, Reference Trattner1999; Piven and Cloward, Reference Piven and Cloward1977).

Instead of shrinking, as the New Deal architects imagined, the ADC programme, renamed Aid to Families with Dependent Children (AFDC) in 1962, grew rapidly even in a strong economy with a relatively low rate of unemployment. Table 1 shows that between 1960 and 1971 the number of welfare recipients tripled, from 3 million to 10 million people (in 2.8 million families). By 1993, the number of recipients peaked at 14.2 million in over 5 million families. The 42 per cent increase between 1971 and 1993 was significantly greater than the 25 per cent increase in the US population over that period.

TABLE 1.

Average monthly families and recipients of ADC, calendar years 1936–2001

Note: Prior to enactment of the Personal Responsibility and Work Opportunity Reconciliation Act of 1998 (PL104–193), cash assistance for needy families was provided through Aid to Dependent Children (1936–1962), later called Aid to Families with Dependent Children (1962–1997).

Source: Administration for Children and Families (2002).

As the growth of the AFDC caseload accelerated after 1960, a series of welfare reforms was enacted over the next four decades, which placed increasing emphasis on the recipients' responsibility to become self-supporting. The most significant of these welfare reforms include (Gilbert, Reference Gilbert, Gilbert and Parent2004):

• • the 1962 ‘service amendments’, which aimed to reduce the welfare caseload through intensive social casework services to rehabilitate the poor, changing their behaviour in ways that would facilitate economic independence (it did not work, as 1 million new recipients were added to the rolls between 1962 and 1966, at a time when unemployment rates were relatively low);

• • the 1967 amendments, which provided services and financial benefits under the work incentive programme (WIN) to encourage voluntary participation in work training and employment (the number of recipients continued to increase, more than doubling, from 5 million to 11 million, by 1988);

• • the 1988 Family Support Act, which required AFDC recipients with children over three years of age to participate in work training programmes (the number of recipients kept climbing to over 14 million in 1993).

Following these measures, the Temporary Assistance for Needy Families (TANF) programme of 1996 may be seen as one more step, perhaps more decisive than the other reforms, in the journey to workfare. Taken as a step-by-step progression, the history of welfare reform conforms to the theoretical view that social policy changes are ‘path dependent’, which means that where they go is largely determined by the strength and momentum already created by their core constituencies: wherever it is headed, the direction is usually difficult to reverse.

It is hard to imagine that poor people – the core beneficiaries of welfare programmes – wielded much in the way of political influence on welfare reform, or were staunchly behind the progression to workfare. But since welfare grants flow through state governments and result in an infusion of cash into the states' economies, governors and state representatives can also be seen as an important constituency of these programmes. In capping the federal contribution to welfare programmes, the TANF legislation fixed the level of federal funding based on the amount states spent on AFDC in 1994. Between 1994 and 1966, the number of welfare recipients had declined by almost 2 million and was still falling. Thus, the federal allocation under TANF gave the states a considerable surplus, which encouraged their political support.

From another perspective, however, rather than one more step along the pathway of reform, it has been argued that TANF constitutes a transformative measure, one that jumps the track, so to speak, and sets welfare on an alternative path.Footnote ¹ It might be noted that for a brief period under the Nixon administration between 1969 and the early 1970s, policy-makers proposed another kind of transformative measure that would have taken welfare reform in a very different direction. Nixon's Family Assistance Plan (FAP) called for a guaranteed annual income implemented through a negative income tax. Despite the fact that the minimum guaranteed level of support for families with no income was quite low ($1,600 for a family of four), this proposal got bogged down in committee, where it died (Trattner, Reference Trattner1999). In contrast to the FAP, which would have dramatically extended the AFDC entitlement to cash assistance, TANF was a historical reversal of AFDC's entitlement to cash assistance (Haskins, Reference Haskins2006). As such, it breached the clause of the social contract under which the state provided a safety net for families struggling to survive outside the labour force.

Policy transformations under TANF

TANF policy revised the essential character of the social safety net provided by US public welfare through the 1990s. Under the six decades of AFDC (and ADC), poor families could rely on public aid that provided a modest cash benefit as a social right to which they were entitled for as long as it was needed without the requirement to work or behave in any particular manner. All this changed with the passage of the TANF legislation, which cut four core strands of the safety net by making benefits conditional on work, time-limited, linked to responsible personal behaviour, and by ending the federal guarantee of support.

Forces of change

What accounts for this fundamental change in US welfare policy? There are no pat answers. Some have suggested that the change was driven by racial politics. Welfare could be seen as a veiled but obvious reference to race. Those who opposed welfare ‘handouts’ were attacking the right to cash assistance that was particularly helpful to people of colour (Edelman, Reference Edelman2007). This is not entirely persuasive. Although African-Americans were disproportionately represented on the welfare rolls, they were never a majority of recipients. In 2001, 39 per cent of welfare recipients were African-Americans (Karger and Stoesz, Reference Karger and Stoesz2008). While cash assistance from AFDC was no doubt helpful to people of colour, the majority of people to whom it was helpful were Caucasian women and children.

It was indeed the case that for decades racial stereotypes contributed to negative views of welfare. In the late 1950s and early 1960s, as Charles Murray notes, ‘the statistics might show that whites have always been the largest single group of welfare recipients, but the stereotypes that enraged the critics was the family of four, five, six, and more children reared at government expense, and somehow the stories about such families always seemed to talk about black families’ (Reference Murray1984: 18). And in the late 1970s and early 1980s, Ronald Reagan often repeated the story of a ‘welfare queen’, apparently an African-American woman, ‘who supposedly drove a Cadillac, bought thick steaks with food stamps, and vacationed in resorts on taxpayer funds’ (Wilson et al., Reference Wilson, Schram, Soos and Fording2003: 194). Although the facts were difficult to pin down, this well-known anecdote imparted a virulent stereotype.

Thus, welfare-related racial politics had been around for almost half a century before the arrival of TANF. While they may have influenced some policy-makers when voting for TANF in 1996, the racial explanation of this fundamental policy shift ignores several more compelling reasons for the transformation of welfare policy. In trying to understand how and why this came about, it is important to appreciate that welfare policy in the US was built upon legislative foundations informed by political ideas, norms and scientific knowledge, which prevailed more than half a century ago. When Harry Hopkins, Frances Perkins, Rexford Tugwell and other New Deal architects devised the blueprints for social protection in the 1930s, public assistance under Aid to Dependent Children was initially designed mainly for the dependent children of widows. The New Deal policy-makers fully anticipated that ADC would fade away as soon as the ‘deserving’ widows with children could qualify for survivors' insurance to which their working husbands had contributed under the newly emerging social security system. In those days, mothers were expected to stay at home and care for their young children; divorce was relatively rare, as was child-bearing outside of marriage. Much has changed since then, and these changes help account for the transformation of welfare policy.

The 2005 reforms: adjusting course

The 1996 TANF reforms settled the big political issues of work requirements, personal responsibility, length of benefit and federal entitlement, which had reached a level of critical concern by the 1990s. As David Ellwood (Reference Ellwood1996) pointed out, the settlement of these issues represented a compromise among four distinct strands of conservatism: work-oriented reformers, who emphasised the critical importance of jobs and training; ideological critics, who emphasised the importance of incentives to encourage moral behaviour; budget-cutters, who viewed the welfare problem as one of excessively generous support; and devolvers, who saw the welfare problem in terms of excessive federal ‘command and control’, with red tape and regulations impairing the ability of the states to create welfare solutions crafted to their own circumstances.

Although when the time came to re-authorise TANF in 2002 there was considerable political haggling about the rules, this was more a matter of adjusting the course than fundamental reform. The fundamental issues of reform had already been resolved in 1996. By 2005, the political debate was concentrated mainly around how much to tighten the work requirements. Under TANF, 50 per cent of all adults receiving assistance and 90 per cent of two-parent households must participate in a set of work activities. However, the states were allowed to adjust these work-rates downward by one percentage point for each percentage-point decline in their TANF caseload since 1995. The Deficit Reduction Act of 2005 modified this ‘caseload reduction credit’ so that adjustments to the work participation rates are now based on a state's caseload decline after 2005. Failure to meet these rates can result in financial penalties of up 5 per cent of the federal family assistance grant to states, which can grow by two percentage points for each subsequent year of non-compliance up to 21 per cent of the federal grant (Baider et al., Reference Baider, Ganzglass, Greenberg, Minoff, Parrott, Schott, Sweeney and Turetsky2007).Footnote ⁶ Since the state caseloads declined substantially over the last decade, leaving some of the least qualified recipients still on the rolls, this modification of the work participation formula places considerable pressure on the states. These pressures are magnified by a significant new set of bureaucratic requirements for state monitoring and documentation of work participation. For example, unpaid work activities must be supervised daily, and homework time for those in vocational educational programmes must be monitored. Finally, the DRA sharpened and narrowed the federal definition of activities that counted toward the work participation rates. These activities included vocational education for 12 months, unsubsidised employment, subsidised employment, community service programmes, child care to enable another TANF recipient to participate in a community service programme, and up to six weeks of job readiness and job search assistance.

In all, the main implications of the most current reforms are to confirm and intensify the work requirements of TANF. These new more stringent measures come at a time when the states' caseloads have already shrunk, leaving them to deal with a pool of hard-to-place recipients in a faltering economy.

Conditionality and entitlement: vying for the new social contract

TANF and the DRA of 2005 have reinforced conditionality and restricted entitlement to welfare. After the last round of reforms, there does not appear to be much further room for tightening the federal conditions linked to public assistance under TANF, nor does there appear to be any political leverage for reversing the conditions currently imposed. Yet there is still room for progressive and conservative policy-makers to manoeuvre as they seek to redefine and rebalance the rights and responsibilities of a new social contract for public assistance. In the course of vying to rewrite this contract, there are at least four avenues along which issues of conditionality and entitlement are likely to be played out.

First, advocates of conditionality may invoke the relative success of TANF work requirements in lowering caseloads and increasing employment to justify extending conditionality to related means-tested programmes of public assistance. If it works in one programme, why not make eligibility for food stamps and public housing subsidies, for example, conditional on work requirements for able-bodied adults?

This question raises the issue, of course, of whom to include among the able-bodied and what conditions might be strengthened for people who are disabled. The political salience of TANF has declined significantly as the number of welfare recipients plummeted from 14 million (or about 5.4 per cent of the total population) in 1995 to 5 million (1.7 per cent of the total population) in 2005 (Crouse et al., Reference Crouse, Douglas and Hauan2007). TANF is one of two national means-tested programmes that offer cash assistance to the poor. The other programme, SSI, serves blind, elderly (not covered by Social Security) and disabled people. In 1995, TANF had 8 million more recipients than SSI. By 2005, the SSI recipients had climbed to surpass TANF by 2 million. Over the decade from 1995 to 2005, the number of SSI recipients has increased relatively from 43 per cent to 140 per cent of the number of TANF recipients. Most of this increase in due to the rising number disability recipients in the SSI programme. Several studies suggests that as many as 30 per cent of the people in the remaining pool of TANF recipients would be eligible for SSI under the administrative category of ‘mental retardation’ (Ramey and Keltner, Reference Ramey and Keltner2002). As the TANF regulations stiffen, states have initiated efforts to assist TANF recipients in applying for SSI, which is still an entitlement programme.Footnote ⁷ As progressives who support expanding social entitlements seek to boost SSI enrolment, advocates of conditionality may focus great attention on the ways to tighten the screen on claimants' ability to work and to introduce TANF-style work requirements into some portion of the SSI programme.

Second, as the social obligation to work has been intensified, advocates for social entitlements are moving to shift the emphasis on policy reforms from the responsibility to work to the social rights to earn a living wage for low-income workers, whose ranks have expanded as welfare recipients move off the rolls and into the labour force. The right to a living wage attracts a broader spectrum of political support than a guarantee of cash assistance for people who do not work. It is easier to mobilise public opinion in support of government intervention that helps people who work hard and play by the rules to provide their families with a decent standard of living, than programmes that pay poor people to stay home and care for their children.

There are several approaches to expanding income supplements for working families. One way is to increase the level of income disregard – the amount of pay for their jobs that recipients are allowed to retain without having their welfare allotment reduced – set by the states. Over a dozen states currently permit recipients to make up to $1,000 in earnings before losing TANF payments (Gilbert and Terrell, Reference Gilbert and Terrell2005). The Center on Income and Budget Priorities suggests that it is to the states' advantage to increase the earned income disregard because it not only provides a powerful incentive to accept low-income jobs, but it extends the period that working families qualify as TANF recipients, which helps the state achieve the federally mandated work participation rate. (Of course, the longer the low-income workers remain on TANF, the more months they accrue toward their 60-month time limit on federal support.)

Beyond using TANF and state funds to supplement low-income working families, progressive advocates argue for the right to a living wage. They propose lifting the federal minimum wage to 50 per cent of the average wage, its level relative to the average wage in the 1950s. This would have brought the federal minimum wage to $8.40 an hour in 2008 (the 2008 minimum wage was $5.85 per hour and is set to rise to $7.25 by July 2009). In addition, several federal tax reforms are seen as also advancing the right to a living wage. These involve expanding the Earned Income Tax Credit (currently over $4,000 a year to a minimum-wage worker with two children) and making the current federal child-care tax credit a refundable credit so that low-income workers who do not owe federal taxes can benefit. These proposals are made bearing in mind the need to balance the minimum wage paid by employers and income supplementation through government assistance (Greenberg et al., Reference Greenberg, Dutta-Gupta and Minoff2007).

The third issue concerns care work and the extent to which it meets the demands of conditionality. Under the DRA regulations of 2005, the core activities that count as work include providing child care which enables another recipient to participate in a community service programme, but not providing child care for their own children. A movement is afoot to recognise care work as equivalent to other work, advanced by a new strand of feminist thought in the US – sometimes referred to as ‘care feminism’ or relational feminism – which is on the rise.Footnote ⁸ This strand of feminist discourse emphasises the value of care work and the need for women to have meaningful opportunities to choose how much of their lives to invest in paid work and childrearing. Supporting this position, Alstott (Reference Alstott2004) proposes the development of caretaker resource accounts that would pay an annual grant of $5,000 to assist individuals who provide continuity of care for a child, and who sacrifice their own opportunities in doing so.

The debate about care work goes well beyond the US. In Germany, Leipert and Opielka (Reference Leipert and Opielka1999) have argued for a new social contract based on the understanding that parental child-care work is functionally equivalent to gainful employment. Calculating the monetary value of child care, they propose a monthly salary of approximately $1,360 for the first child and $680 for additional children up to age seven. Although this proposal stirred some debate in the late 1990s, it did not carry the day. Many countries provide a cash allowance for home care for elderly people, which can be used to pay relatives. And cash payments for home care of children are already offered in several European countries, including Norway and Finland (Gilbert, Reference Gilbert2008).

Finally, at the same time that TANF imposed federal requirements regarding work and time limits, it also allowed states great discretion to organise their welfare programmes in their own fashion and to adjust the balance between conditionality and entitlement in light of local priorities. TANF, for example, permits the states to limit assistance to particular categories of poor families. States, if they wish, can prohibit cash payments to otherwise eligible legal immigrants, or to two-parent families, or to families failing to meet certain procedural or programmatic rules and regulations. Cash, indeed, does not have to be provided at all: the states can structure ‘assistance’ in the form of vouchers, services or other benefits that meet ongoing basic needs.Footnote ⁹

The states have responded to this flexibility in a variety of ways. Some have developed requirements that are more stringent than those in the federal regulations. Idaho, for example, has established a lifetime eligibility limit of two years and refuses to offer exemptions to special need groups, such as victims of domestic violence. Mississippi requires welfare recipients to accept any job offered or else lose all benefits. Employers hiring welfare recipients are provided a TANF subsidy of $3.50 an hour to which they must add $1.00 an hour. In this fashion, public assistance is turned into a wage subsidy programme.

Other states have adopted more lenient rules, in some cases guaranteeing public support over the five-year federal limit with state funds. New York State, for example, provides a constitutional guarantee that obligates local governments to help all the needy, irrespective of how long they have been on welfare. Vermont, which also refuses to impose a life-time limit on welfare support, describes its approach as a ‘kinder, gentler’ form of welfare reform. While requiring single parents to work, the state provides a generous cash supplement to poor workers; it has also created its own nearly universal health care programme for children. Most recipients must do community service after receiving benefits for two and a half years. If a recipient is unable to secure a private sector job, the state provides ten months of employment in a public or non-profit organisation.

In sum, a new social contract is being forged about the right to public assistance, which seeks to change the balance between entitlements and conditionality. Politically, this contract is driven by two agendas. The responsibility agenda emphasises the obligation to work and socially acceptable personal behaviour; the entitlement agenda emphasises the right to work training, employment opportunities and a living wage. Although the federal framework has been codified, the states are still experimenting with various measures – not all of the clauses of this contract are entirely worked out yet. It is quite possible that between federal and state initiatives the political balance that is achieved will vary in different localities, reflecting the spectrum of political constituencies ranging across the 50 states. Whatever balances emerge, the federal role in the welfare arena has fundamentally changed from that of providing direct income maintenance to the poor to that of creating public support to promote private responsibility.