Introduction

Recent developments in policies towards lone parents in Australia have emphasised the role of employment in increasing income and self-sufficiency. The emphasis on employment is also the case in other OECD countries with a general trend towards benefits for lone parents being made dependent on participation in the labour market (Millar and Rowlingson 2001).

The United States of America (US) has undertaken substantial reforms, over the 1990s, to the ways in which social assistance is provided to lone parents. The most prominent of the reforms have followed the passing of the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) which, among other things, introduced a five-year maximum lifetime limit on the receipt of Federal benefits and made work a mandatory requirement. There have also been changes made to increase the incentives to take up paid employment by allowing an increased amount of earned income to be kept before benefits are lost, and an expansion of the Earned Income Tax Credit.

Following the reforms there has been a dramatic fall in the number of lonemother families in the US receiving welfare payments and increases in employment rates, particularly amongst groups of lone mothers who historically have had very low employment rates.¹ The extent of the changes to the US system, combined with their apparently dramatic impact upon the behaviour of lone mothers, makes the United States experience of particular relevance to those interested in the Australian income support system. The recent report of the Reference Group on Welfare Reform (the McClure Report) contained a discussion of the policies in a number of other countries including the United States.

One of the reasons Australian policy makers are seeking to increase the employment rates of lone mothers is concern about “welfare dependency”.² It is sometimes argued that some kind of compulsion or external pressure needs to be exerted on people receiving income support payments in order for them to find paid employment. The policy objective of increasing rates of paid employment and breaking the cycle of “welfare dependency” has dominated the welfare reform debate in the United States (see Saunders 2000).

Australia has a relatively low rate of employment amongst lone mothers, and it seems that lone mothers in Australia are spending long periods in receipt of government payments. Australia may well have a much higher rate of long-term benefit dependency than previously thought. Recent work by Professor Bob Gregory from the Australian National University suggests that when the movements from one payment type to another are taken into account, the total amount of time spent in receipt of welfare payments is much longer and that the potential for “welfare dependency” is greater than previously thought (Gregory 2002).³

While several reviews of the implications of the US experience of welfare reform for Australian policy have been written (Burtless 2002; Duncan 2002; Ellwood 2002; Howe and Pidwell 2002; Perry 2000), these papers have focused on a narrow range of outcomes; primarily the numbers of benefit recipients, employment rates and income levels. The main objective of this paper is to consider the impact of the United States welfare reforms on a wider range of outcomes, including the mental and physical health of mothers and children’s wellbeing. A further purpose of the paper is to explore possible implications for the way in which the Australian system of income support for lone mothers can be reformed if the policy objective is to increase employment rates.

The following section presents an overview of the US social assistance system for lone-mother families and details the changes introduced during the welfare reforms of the 1990s. The third section discusses the evidence on the impact of the US welfare reforms on a range of outcomes for lone-parent families. The potential implications for the nature of policies which may be effective in increasing the wellbeing of Australian lone-parent families are also discussed. Particular attention is paid to differences in the social assistance schemes for lone parents between Australia and the United States, and differences in labour market conditions and institutions.

Social assistance scheme for lone mothers in the US

The United States system of social protection can be characterised as a social insurance model together with a very limited social assistance scheme. Working age adults can be covered by up to three layers of Federally sponsored protection. First, those who are unemployed as a result of losing their job due to general economic conditions can receive up to six months of Unemployment Insurance (UI). Second, low-income lone-parent families with dependent children are eligible for cash assistance in the form of Temporary Assistance for Needy Families (TANF). Third, virtually anyone who is poor can receive coupons that can be used to purchase food (Food Stamps).

The United States has no universal children’s allowance, no Federal means-tested support for lone adults (although some States have limited programs), no universal housing program (although here too there is housing aid available in some States), and no guarantee of health coverage. A further form of federally financed assistance is the Medicaid program, which provides health care benefits to poor children⁴ and lone mothers who received TANF. In the US context, the term “welfare” generally equates with TANF and food stamps and is almost exclusively for lone-parent families. In addition, there is social assistance for adults with a disability. Further details of the US system are presented in Table 1.

TANF, the Food Stamp Program (FSP) and the Earned Income Tax Credit (EITC) are designed to work together: TANF gives States program flexibility to address general family needs and to assist in moving people into the workforce; FSP provides a real-time resource supplement; and EITC provides a special cash bonus for families working for low wages.

Welfare reform

The current US system is the result of an extensive process of welfare reform, including the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA), passed by Congress in 1996. While the Act is perhaps the best known point of welfare reform, the process itself began during the 1980s with the Federal Government increasing the flexibility available to States in the ways in which they provided social assistance to lone parent families. A number of States took advantage of this flexibility to trial a range of initiatives including voluntary training programs, time-limits and mandatory work requirements. By 1996, over half the States had obtained some form of Federal “waiver” from the national law to experiment with the ways in which they provided benefits to lone-parent families. The PRWORA made quite major changes to the social assistance system for lone mothers including introducing a lifetime-limit on the receipt of cash assistance from the Federal Government.

Prior to the passing of the PRWORA in 1996 poor lone-parent families with dependent children and some two-parent families who did not qualify for Unemployment Insurance could receive aid indefinitely under a means tested program called Aid to Families with Dependent Children (AFDC).

The funding arrangement for the AFDC system was a Federal–State hybrid. The Federal Government determined what types of people were eligible (mostly lone parents) and how their income would be assessed, but States were free to set benefits at any level they chose, and States were responsible for administering the system. The costs of the scheme was shared between the States and the Federal Government. There was no limit on the duration of aid, and no person who met the eligibility criteria in the State could be denied aid. In 1988 the Family Support Act was passed. This legislation required single mother recipients of AFDC to work in return for receipt of social assistance. However, the work requirements were not always enforced.

The PRWORA abolished the AFDC program and replaced it with block grants to State governments called the Temporary Assistance for Needy Families (TANF) block grants. The objectives of the Act include: reducing the number of single parents receiving government cash assistance (the caseload); increasing employment rates; decreasing the number of out-of-wedlock births; and increasing marriage rates. The PRWORA provides broad policy parameters, funding levels and general guidelines as to how the Federal funding can be spent, but allows States a great deal of flexibility. The States could largely spend it on anything “consistent with the purposes of the Act”.

The funding arrangements of TANF are very different from those of the AFDC program, which it replaced. Under the AFDC program, funds provided to States were almost entirely used for the provision and administration of cash assistance to low-income families. The Federal Government matched State expenditure. Under the TANF block grants, States were given a fixed amount of money based upon their caseload in 1996. The level of funding does not vary with changes in caseloads and is not adjusted for inflation.⁵ States were also allowed greater flexibility in what Federal TANF funds were spent on.

Perhaps the best known change introduced by the PRWORA is the imposition of a 60-month maximum lifetime-limit on Federally funded assistance for most families. Time-limits were first discussed in 1992 when presidential candidate Bill Clinton promised to “end welfare as we know it” by placing a two-year time-limit on AFDC benefits. An interesting feature of this promise was that subsidised jobs were to be provided, if necessary, to recipients whose benefits were terminated due to a time-limit (Ellwood 2002). In 1993 and 1994 some States were allowed to impose time-limits on benefit receipt if they wished, and several States experimented with time-limits. By August 1996 a total of fifteen States had received waivers authorising a termination time-limit.

Following the passage of the PRWORA, all States have implemented the Federally imposed five-year time-limit on benefits, except for ten States which adopted even shorter lifetime-limits. While the time-limit may seem very blunt, in reality States are allowed to give some recipients exemptions or extensions which means that the way in which the time-limits operate is complex. It seems that most recipients who have reached a time-limit and who are not employed but have made a genuine effort to find employment receive an extension of the time-limit.

Incentives for State authorities to decrease caseloads

In addition to the attempts to modify the incentives faced by individual welfare recipients to take up paid employment, a key component of the PRWORA was to create strong incentives for States to implement the reforms and develop policies which are effective in reducing the number of their residents receiving welfare payments. These incentives were created by requiring States to have enrolled 50 per cent of their welfare recipients in work programs by 2002. Failure to achieve that would result in a reduction in the block grant received by the State.

In response to Federal work-oriented requirements in PRWORA, States took steps to shift to “work first” welfare systems by requiring welfare applicants to participate in work-related activities. On the whole, the welfare programs have been designed to emphasise rapid labour force attachment. While work requirements were a component of the pre-PRWORA system, these requirements have been strengthened and more rigorously enforced post-PRWORA through the use of sanction policies that reduce or eliminate benefits if the program rules are not followed. In addition, States overwhelmingly adopted narrower exemptions from work requirements than had existed under AFDC.

Financial incentives for paid employment

The financial incentives for taking-up paid employment instead of relying on welfare payments will be partly determined by the interaction between the level of welfare payments, earnings in the labour market, and the taxation system. In the United States context a key factor is the Earned Income Tax Credit (EITC). This is a refundable tax credit, which families can apply for when they file their annual tax returns. It essentially subsidises low-paid workers’ wages as a way of increasing work incentives. It also allows the wages paid by employers to be kept low while maintaining income levels for workers.⁶ At low levels of earnings the credit rises, up to a specified limit. Parents who have no earned income are not eligible for the EITC. Thus the EITC increases the incentive to take up paid employment. While the EITC generates a financial incentive for non-working single mothers to take up paid employment, it generates a disincentive to increasing hours of work once family income exceeds the threshold level of income beyond which the credit is reduced

A key feature of the US welfare reforms was the increase in the financial benefits to welfare recipients from taking up paid employment. Several policy changes outside of the direct welfare reforms impacted upon the financial gains of taking up paid employment.

First, the minimum hourly wage was increased from $4.25 to $5.15 in 1993. Second, in 1993 there was also a major expansion of the Earned Income Tax Credit (EITC).⁷ Third, in sixteen States the Federal EITC was supplemented with additional income tax credits, which provide extra resources to working families. Fourth, prior to the PRWORA, when a lone mother moved from welfare to work she would often lose health insurance since the jobs that mothers leaving welfare typically find do not have employer provided health insurance. In order to reduce the effects of this, Transitional Medicaid was introduced, covering adults leaving TANF for 12 months. Perhaps even more significant has been the passing of the State Child Health Insurance Program. The State Child Health Insurance Program and changes in Medicaid rules have expanded the coverage of health insurance for both children and parents in low-income families, thereby increasing the incentive to take up low paying jobs which do not offer health cover.

Fifth, most States further increased financial incentives by modifying the rules regarding the impact of earned income on benefit receipt. In thirty-one States the earned income disregards were expanded and in many States the rate of withdrawal of benefits was reduced so that benefits could be received over a much larger range of earned income. (See Appendix A for details of the TANF program rules for each State.)

Child care

With the moves to mandate employment, the availability and affordability of child care has become a pressing issue. This has led to increasing amounts of money being put into providing subsidies to help with the costs of child care for low-income families, most often provided in the form of vouchers which help parents obtain care from a range of child care providers. The PRWORA reformed and expanded funding for child care assistance to low-income families by combining several existing programs into a single block grant, the Child Care and Development Fund (CCDF) and by authorising States to transfer funds from TANF to the CCDF. In addition, a number of States had increased State-level funding for child care in the mid-1990s via means-tested subsidies. The combined effect of these changes has been to increase the level of funding for child care assistance. The States have introduced varying eligibility rules, types of subsidy coverage and the level of subsidy provided.

As caseloads and expenditure on providing TANF payments to single-parent families have fallen, States have increasingly used TANF funding to subsidise child care for low-income families. To give some idea of the scale of the increase, in 1997 Federal child care funding through Federal TANF and CCDF funds was $2.1 billion as compared to $7.4 billion in 2000 (Adams and Rohacek 2002a).⁸ Despite the substantial increases in funding levels, there is evidence that there is excess demand for subsidies. This means that a substantial proportion of low income families who are eligible for child care assistance are not receiving child care subsidies (Adams and Rohacek 2002a).⁹

Summary

As outlined, the changes to the United States system of social protection for sole parents have been substantial and have involved almost every aspect of the system. A key dimension of the reforms has been to provide States with increased flexibility in the ways in which they deliver social assistance to lone mothers. The funding arrangements of TANF have been designed to provide States with increased incentives to reduce the number of lone mothers to whom they pay TANF. The combination of these factors has resulted in States adopting a wide variety of approaches and rules. This makes providing a simple summary of changes difficult.

However, the following commonalities can be identified:

• work requirements have been strengthened and enforced through the use of sanctions;
• lifetime-limits have been imposed;
• financial incentives to take up paid employment have been enhanced by changes to benefit levels, free areas and taper rates, and by the expansion of the EITC;
• subsidies for child care have been expanded;
• changes in the “culture” of welfare offices have taken place: some States have dramatically increased the pressure on benefit recipients to move off welfare: States have also sought means to divert eligible recipients from receiving any aid; and
• education and training programs have been introduced and expanded

In addition to the changes highlighted above, other changes include: unmarried teenagers being required to reside with their parents if seeking assistance; benefits to non-citizens who are legal residents of the US being restricted; the narrowing of eligibility for supplemental security income for children with disabilities; and provisions that target the use or sale of illegal substances by limiting or removing eligibility for individuals convicted of drug related crimes.¹⁰

In aggregate, the United States Government is now spending more on low income families than they did prior to the welfare reforms. However, the expenditure has been significantly redirected with support being withdrawn from non-working sole mothers and increased for working sole mothers and other lower income families with children (Ellwood 2002; Kamerman and Kahn 2001).

As David Ellwood (2002: 21) argues: “The moral legitimacy of a time-unlimited cash for non-working parents was eliminated by the changing nature of family structures and the patterns of mother’s labour market work. But the moral legitimacy for aid for working single parents was heightened with the new focus on the enormous pressures they face in trying to nurture and provide for their children.” (Emphasis in the original.)

Work requirements and sanctions

A central component of the welfare reforms has been the strengthening of the work requirements associated with receipt of TANF. Compliance with these work requirements is enforced through the use of financial penalties, known as sanctions. While sanctions have long been used to enforce work-related requirements for welfare recipients, the changes in the 1990s have been in the severity of penalties and the frequency of their use. The 1996 reforms required States to terminate or reduce benefits “pro rata” when recipients failed to comply with work requirements, but the amount and duration of sanctions were not specified. In addition, unlike in the past, the food stamp rules were changed so that benefits are no longer increased when the cash grant is cut.

The approaches adopted by States can be divided into three categories. Fullfamily sanctioning involves cutting off TANF payments at the first instance of non-performance of required work or other activities. Graduated sanctions involve reducing the level of TANF payments at the first instance of nonperformance and then progressively reducing the payments for further instances of non-performance and finally cutting off payments entirely. Partial sanctions involve cutting only the adult portion of the TANF payment even after repeated instances of non-compliance. This enables recipients to retain the bulk of their TANF benefits even if they fail to take up paid employment or other required activities.

Across the US, thirty-six States use full-family sanctions and eighteen of these impose them on the first instance of non-compliance. In the other eighteen States, partial sanctions can escalate to full family sanctions with repeated or continued non-compliance. In seven States repeated non-compliance can result in life-time ineligibility for benefits (Bloom and Winstead 2002). It is important to note that in several States with large TANF caseloads (California, New York and Texas) there are no full-family sanctions.

Most States also have specific criteria that provide some discretion as to whether a sanction is applied¹¹ In fact, most States grant exemptions when a recipient is ill or incapacitated, is caring for an incapacitated family member, or lacks transportation. Bloom and Winstead (2002) suggest that sanctions are often imposed on clients who do not understand the program rules or who have good cause for their failure to comply.

Time-limits, extensions and exemptions

What seems to be the most radical feature of the United States welfare reforms was the introduction of 60-month lifetime-limits on receipt of Federal assistance. However, there has always been an implicit time-limit generated by the fact that once an adult no longer has a dependent child they could no longer receive AFDC, and in the absence of a general social assistance scheme, in general, they would no longer have been eligible for cash assistance.

The States have adopted widely varying approaches to time-limits. The time-limits imposed by the PRWORA only apply to Federally funded benefits and are a maximum time-limit so States can impose a shorter time-limit if desired. Variations between States in the time-limits imposed on receipt of TANF can be seen in Appendix B. Currently forty States have time-limits that can result in the termination of families’ welfare benefits. Of these forty States, seventeen have limits of less than 60 months. However, nearly half the national welfare caseloads are in States that either have no time-limits (two States which use State funds to provide assistance) or a time-limit that reduces or modifies benefits when the time-limit is reached (Bloom, Farrell and Fink 2002).

While on the surface the time-limit on benefit receipt may appear very simple, the actual implementation has been complex. The reality is that all States provide exemptions or extensions from their time-limits for certain groups of families, although the policies can differ dramatically from State to State. Exemptions are most often for families in which the mother is incapacitated or is a victim of domestic violence. Most States allow for time-limit extensions in cases where the parent was unable to find work despite diligent efforts.

States are allowed to provide extensions under the funding conditions of TANF which allows Federally funded TANF assistance beyond 60 months to be paid to up to 20 per cent of the State caseload. States are not required to impose time-limits on assistance provided with State funds. This means that States can stop the Federal time-limit clock by paying for family’s benefits with State funds or they can use State funds to assist families who pass the Federal limit.

In practice, the Federal time-limit is not a limit on benefit receipt for individual families, rather it is a funding constraint that impacts on State policy decisions. In general, there are four reasons why a family can receive assistance for a given month without it being counted toward the Federal time-limit: it is a child only case¹²; the family is living on an Indian reservation that is experiencing an unemployment rate greater than 50 per cent; the family is exempt from having months counted under a State waiver policy; or the State is funding the assistance exclusively with State funds.

Families have begun to reach the Federal time-limit of 60 months as have families in some States where shorter time-limits have been applied. So what actually happens when families reach time-limits? As at December 2001, about 231,000 families have reached a time-limit and at least 93,000 families have had their welfare case closed due to a time-limit and another 38,000 have had their benefits reduced. Nearly 29,000 had their TANF case closed but were receiving alternative benefits. The remaining 71,000 cases that reached a time-limit received assistance in the following month. Some of these families received an extension and were later terminated; some left TANF voluntarily; and others continued to receive assistance (Bloom, Farrell and Fink 2002).

Based on the experience of eight States¹³, Bloom, Farrell and Fink (2002) find that the implementation of the time-limit policies differed across the State and for some States, even within the State. They find that as recipients approach a time-limit staff in welfare offices make special efforts to engage recipients in welfare-to-work programs and carefully monitor participation in the program. The intensive efforts target services to recipients in need as well as providing evidence about whether recipients are willing (and able) to comply with work-related requirements.

Most of the eight States grant extensions or exemptions for recipients who comply with program rules but do not have jobs when they reach the time-limit. How States determine who has played by the rules varies considerably, although many States base the decision on recipients’ current willingness to comply with the program’s work requirements rather than on their past history. The bottom line is that it seems that when a lone mother “plays by the rules” and makes a genuine effort to find employment they will continue to receive some form of social assistance if they cannot find employment after they reach a time-limit.

Welfare-to-work programs

With the shift to a focus on employment, welfare-to-work programs have become very prominent in the United States. These programs are particularly important for welfare recipients who have difficultly finding employment as a result of low levels of education, lack of work experience, health problems and substance abuse. States have adopted a variety of approaches to the ways in which welfare-to-work programs are provided. Gueron and Hamilton (2002) categorise the welfare-to-work programs implemented by States into three broad approaches.

First, education or training first programs which focus on putting adults on welfare into education or training programs before requiring them to find work. These programs generally stress anti-poverty objectives and are based on the premise that many welfare recipients need to improve their skills so they can get a job and emphasise the need for some stability in employment and the earning of income which is sufficient to support a family. Second, job search first programs which require welfare recipients to start looking for a job and focus on skills such as resume writing and interviewing. Third, mixed strategy programs that provide a more flexible approach in which staff and welfare recipients have choice in the types of assistance.

The range of activities that can count as work activities is quite broad, and can include: paid employment; educational activities and training; job search; and work experience (work in exchange for welfare benefits).

An implementation issue however concerns how effectively the changes were communicated to caseworkers, welfare recipients and applicants. Most States required some form of orientation session for applicants before they could be approved for benefits. Staff in welfare offices are using the process of application for TANF to give the message that employment is expected and that the period of time for which benefits can be received is limited. In general, recipients are required to sign a personal responsibility agreement as a condition of receiving assistance. Depending upon the State, participants can be required to attend classes on parenting, money management, life skills, family planning, and substance abuse counselling and treatment. Employees in welfare offices are being asked to be increasingly involved in helping welfare recipients find solutions to their particular difficulties. It is interesting to note that the terminology has changed with welfare officers having their job descriptions changed to the term “case manager”.

Given the financial incentives for States to reduce their caseloads, a number of States have implemented diversion programs. These programs include providing a lump-sum payment to cover a particular expense which is limiting a person’s ability to work. In return the applicant is required to forego assistance for some specified period of time. Other diversion programs include referring applicants directly to housing programs, and charities. While there may be some advantages to short-term diversion programs these programs also appear to have generated some problems. One particular problem is that since eligibility for TANF, food stamps and Medicaid are determined using the same application, families who are diverted from applying for TANF did not apply for food stamps and Medicaid to which they would have been entitled.

Impact of the US reforms

This section discusses the evidence of the impact of United States welfare reforms on a range of outcomes including the number of recipients of TANF (the caseload); income of welfare recipients and welfare leavers; the mental and physical health of mothers; and children’s wellbeing.

Existing reviews of the literature have focused on the impact of welfare reform on caseloads, employment rates and income, with much less attention being given to other outcomes such as children’s wellbeing and mother’s health. This is, in part, a reflection of a great amount of evidence on caseloads, employment and income being available. However, there have been a number of recent high quality studies of the impact of welfare reform on a range of other outcomes.

Caseloads and employment

The most striking change has been the dramatic decline in the number of recipients of TANF with the caseload being down nearly 60 per cent since their peak in 1994 (Figure 1). At the same time, there has been a substantial increase in employment rates among single mothers during the 1990s. Following gradual increases in the employment rates of single mothers in the twenty years to 1994, the employment rate has increased from 60 per cent in 1994 to 72 per cent in 1999. Among single mothers who have never been married the rise in employment rates over the same period was 47 per cent to 65 per cent (Moffitt 2002).

Figure 1: Number of families on AFDC and TANF, 1962-2000

Source: US Department of Health and Human Services, Administration for Children and Families, Office of Planning, Research and Evaluation, (Available online at http://www.acf.dhhs.gov/).

While the trends in the caseloads and employment rates are well understood and not really contested, the reasons for the decline are not as clear. The conceptual question is: what would have happened to caseloads and employment rates if the changes implemented in the process of welfare reform had not occurred? In other words, what is the counterfactual?

The coincidence of a major economic expansion in the US and a major shift in policy resulted in significant behavioural changes, with rapid declines in public assistance caseloads and rapid increases in labour force participation among single mothers. The key question is: how much of the reduction in public assistance caseloads is due to economic expansion versus policy change, and how much of it would be reversed by recession? The two are almost surely interacting with and reinforcing each other, so that a strong labour market has allowed States to put more energy into case management or move faster in placing recipients into welfare-to-work programs, without working as hard to help clients in these programs locate jobs. These interactions make it difficult to identify the separate effects of the economy and welfare reform. (See Blank 2001 for a discussion of these issues.)

After comprehensively reviewing studies of the impact of TANF on employment, earnings and caseloads, Moffitt (2001: 80) concludes that the evidence is reasonably strong that the TANF program has increased employment and earnings and decreased the caseload, on top of what would have occurred had AFDC remained in place. However, Moffitt (2001) notes that the separate effects of work requirements, time-limits, sanctions, family caps, and other individual features are essentially unknown.

It appears that the welfare recipients who have left welfare tend to be the most job-ready recipients and that the proportion of remaining recipients who face significant employment barriers has increased (Blank and Schoeni 2000). This means that it will be increasingly difficult for those recipients who have been receiving welfare for a long period of time and who remain on the welfare rolls to find employment, particularly as the rate of aggregate employment growth slows.

Income and poverty

A key measure against which the US welfare reforms need to be assessed is income and levels of poverty. The changes to the system which were designed to increase the financial benefits of paid employment (as compared to continued benefit receipt) are expected to have had an effect on income.

In aggregate, family incomes have risen since the passing of PRWORA and the number of single-parent families with an income below the US Federal poverty line has fallen.¹⁴ However, while the number of families with a family income less than the poverty line has decreased, the disposable income of the poorest single-parent families has fallen, indicating that the level of “deep poverty” has increased (Zedlewski 2002).

While there have been some falls in the number of families living below the poverty line, single-parent families continue to experience very high rates of poverty. In 1999, the rate for children living in single-mother families was estimated to be 42 per cent as compared to 17 per cent for all families with children (Waldfogel, Danziger, Danziger and Seefeldt 2001). The number of female-headed households living below the poverty line is estimated to have decreased from 3.8 million to 3.1 million between 1994 and 1999, a decline of 22 per cent. This, however, is much smaller than the decline in caseloads of 48 per cent (US Census Bureau 2001; US Department of Health and Human Services 2000).

Corresponding to the increasing rates of employment there has been a dramatic increase in the share of income, which is coming from earnings (see Haskins, Sawhill and Weaver 2001; Zedlewski 2002). However, any increases in income levels have been modest with many low-skilled former TANF recipients finding jobs that pay poorly and that do not last (Corcoran, Danziger, Kalil and Seefelds 2000).

Focusing in more detail on how families at different points in the income distribution have fared following welfare reform, Zedlewski (2002) analyses changes in disposable income between 1996 and 1998 using data from the Urban Institute’s National Survey of American Families. The results of this are summarised in Table 3. Perhaps the most important point is that for all families there have been very substantial increases in disposable income across the income distribution, whereas for single-parent families there has been a very major decrease in disposable income of 7.8 percentage points for the lowest income group. Overall, the rate of growth in disposable income is lower for single parent families.

While the increases in income levels are quite strong, it must be remembered that the income levels of single-parent families are much lower than for all families. For example, the income cut-off for the lowest income quintile for single parent families is below 70 per cent of the Federal poverty level, whereas for all families with children it is below 130 per cent of the Federal poverty level.

Notes: The income quintiles are calculated separately for all families with dependent children and single-parent family populations. Source: Zedlewski (2002), Figure 4.1. Derived from Urban Institute’s National Survey of American Families.

Amongst single- parent families in the bottom income quintile the proportion receiving cash assistance fell from 69 per cent in 1996 to 52 per cent in 1998. Three-quarters of these families received food stamps and Medicaid or State Child Health Insurance Program. The number of very low-income families receiving EITC increased (Zedlewski 2002).

A recent study by Danziger et al. (2002) uses data from the Women’s Employment Study, a random sample of welfare recipients living in urban Michigan county, which contains information on household income thus allowing a more accurate measure of living standards to be constructed. From a sample selected in February 1997, three-quarters of all respondents were working by 1999. Over the study period respondents typically moved first from being welfare reliant to combining work and welfare and then to wage reliant.

Interestingly, substantial numbers of the respondents co-resided with another household member who had earnings in the month prior to the survey. Many of the other earners were husbands or cohabiting partners, and they earned substantially more than the respondents did. Women who had left welfare for work and who were no longer receiving any welfare payments had a significantly higher income than women who were working and receiving welfare and those who were not working. The income levels of the working mothers however remained low, with 49.1 per cent of working mothers having an income below the Federal poverty line, 77.2 per cent of women combining work and welfare were below the poverty line, and 90.8 per cent of welfare reliant mothers were below the poverty line. Interestingly, among the mothers neither working nor on welfare, the poverty rate was 71.8 per cent. These women tended to be living in households in which other family members (often a partner) were receiving income from other sources

Other studies of the impact of TANF on income levels have used data on “welfare leavers”. These studies have generally found that there have been only slight improvements, and in some cases slight declines, in the financial wellbeing of single mothers who have moved from welfare to work (Bavier 2001; Cancian, Haveman, Meyer and Wolfe 2002; Acs and Loprest 2001). However, these studies have two main limitations in assessing the impact of welfare reform on earnings. First, they often do not include the earnings of other household members. Second, they do not separately consider the experience of women who leave welfare and are in paid employment and the experience of women who leave welfare but are not in paid employment.

Most of the studies of rates of poverty and living standards of TANF leavers do not take account of the additional costs of working, the largest of which is likely to be child care costs. This will lead to an overstating of the living standards of these families. A recent study by Meyers, Han, Waldfogel and Garfinkel (2001) of sole mothers in New York finds that the costs of child care to TANF leavers are considerable, and taking account of these effects increases the proportion of families living in poverty.

A key determinant of the effects of increased employment rates on financial wellbeing is the extent to which mothers moving off welfare are able to find steady employment and the extent to which they, over time, are able to find better jobs and hence have their income increase. While there are a number of studies of the employment and earnings experience of welfare recipients following leaving welfare, most of the studies are based on data prior to the 1996 reforms. These studies have generally found that the typical pattern is for a welfare recipient to find employment, then to lose or quit their job and to experience a period of joblessness (Acs and Loprest 2001; Campbell, Maniha and Rolston 2002; Loprest 2002; Rangarajan and Novak 1998; Rangarajan and Novak 1999). Using data from a national survey of the employment experience of mothers leaving welfare, Rangarajan and Novak (1998) found that only 30 per cent were in employment for more than 75 per cent of the five years the study followed them after they had left welfare. The studies have also found that the rate of growth in wages is low, although there is some evidence that earnings increase due to increased hours of work (Burtless 1995; Rangarajan and Novak 1999).

A high proportion of the jobs found by former welfare recipients have limited provisions for benefits such as paid sick leave and health insurance. Acs and Loprest (2001) find that only about one-third of the jobs have employer provided health insurance and between one-third and one-half provide sick leave. In addition, a relatively high proportion of the jobs are in the services sector and often involve working night shifts, or have irregular and unpredictable hours which can make arranging child care difficult.

Given the success of welfare recipients in finding employment, States have increasingly developed programs which have the objective of helping low-income mothers keep jobs they find and then to increase earnings over time (either through increases in hourly wages or increases in the number of hours worked).

Mother’s health

The mandatory work requirements, their enforcement and the imposition of time-limits may have created new stressors or exacerbated existing stressors for single mothers. In addition, the incentives for single mothers to be in a relationship have increased. On the other hand, increased participation in paid employment and training courses, combined with increased income for at least some lone mothers, may have led to improvements in mental and physical health.

There is strong evidence that having a low-income is associated with lower levels of physical and mental health. However, there is relatively little evidence about the effect that welfare reform has had on the physical and mental health of single mothers. While there is some evidence that single mothers receiving welfare payments and not working have lower levels of mental and physical health than do single mothers who are in paid employment (DeAngelis 2001a), the evidence does not exist yet to determine cause and effect. That is, it is not possible to disentangle the extent to which low levels of mental and physical health lead to lack of employment and the extent to which not being in paid employment and receiving welfare payments cause health problems.¹⁵

It is probable that, to the extent to which welfare reform has led to increases in income levels of lone mothers, there will be some improvements in the health of mothers. However, this must be balanced against possible increases in levels of stress associated with juggling child care and paid work (DeAngelis 2001b).

Immediately after the introduction of TANF there was a dramatic fall in the number of families enrolled in Medicaid health insurance (Weil and Holan 2001). The reasons for the decline in enrolments include: the strong economy meant that less people were dependent on government funded health insurance; administrative and other barriers to retaining Medicaid for those losing welfare meant that there were problems associated with take up; and State adopted diversion programs discouraged eligible families from applying (Weil and Holan 2001). However, these problems were recognised relatively early on and steps were taken to counter problems of low-levels of take up by de-linking Medicaid from welfare eligibility.

Impact of welfare reform on children’s wellbeing

One of the major reasons for wanting to increase the levels of employment of lone mothers in Australia is concern about the impact on the life chances of children growing up families in which no adult is employed (Gregory 1999). This makes the impact of the US welfare reforms on the wellbeing of children of great importance in assessing the overall success of the reforms.

There are a number of ways in which the US reforms may have impacted on the wellbeing of children. It has been argued that having a time-limited welfare system which results in increased employment rates of single mothers may lead to improved outcomes for children. It has been suggested that this may occur via:

• providing children with family role models who work and are self-sufficient;
• introducing a regular schedule into the family routine;
• increasing rates of non-maternal child care which may compensate to some extent for a poor quality home environment; and
• increasing income available to families. Others have argued that, on balance, the US welfare reforms will have negative impacts upon children’s wellbeing. It has been suggested that this may occur via:
• increasing stress in the family by reducing parents’ opportunities to spend time with their child and interfering with parents’ monitoring of their children’s activities, particularly in single-parent families;
• increasing rates of use of poor quality non-maternal child care; and
• for families unable or unwilling to comply with work mandates, work requirements may actually increase poverty.

Accurately estimating the causal impact of welfare reforms on children’s wellbeing is very difficult. The most reliable method for determining the impact of a policy change on the children of the adults affected is to use an experimental design – that is, studies in which parents were placed at random in either a program group, which had access to the new services and benefits and was subject to the new rules, or a control group, which received the benefits and was subject to the rules that had previously existed in the locality of the study site or sites. While most of the reliable existing empirical studies have been based upon changes in the welfare system prior to the 1996 Welfare Reforms, they are based upon programs that incorporated elements of the 1996 reforms. This provides some evidence as to the likely impact of the US reforms, however not enough studies using data after the 1996 reforms are available yet to allow any firm conclusions to be drawn.

A range of evaluations have been synthesised by the Manpower Demonstration Research (MDRC) Corporation (Morris et al 2001, Morris, Knox and Gennetian 2002). The MDRC reviews covered five studies which collectively examined the effects on children of eleven different employment-based welfare and antipoverty programs aimed primarily at single parent families.¹⁶ All of these programs where implemented prior to the welfare reforms in 1996. However, these programs tested three basic approaches that are currently being used in many State welfare policies to increase the self-sufficiency of welfare recipients–earnings supplements; mandatory employment services; and time-limits.

The findings from these studies suggest that welfare-to-work programs which involve earnings supplements produced more positive impacts on children’s wellbeing than other programs. The positive impacts were found to be largest for children in long-term recipient families and for elementary school-aged children. It is important to bear in mind that a substantial proportion of children living in single-parent families were not progressing normally in school.

The programs with mandatory employment services, all of which boosted parents’ employment without increasing income, had few effects on children. The one program with time-limits, which led to an increase in parents’ employment and a modest increase in income, produced few noteworthy impacts on children, and the impacts found were mixed. In the two studies that examined adolescent children, the program appeared to be less beneficial for adolescents than for children in middle childhood. Of particular interest is the finding of one study that living in a family in which the parent has left welfare but remains poor has a negative impact on the reading outcomes for children (Moore et al. 2002).

Overall, while specific welfare experiments have shown some positive results for children when parents go to work and increase the financial resources available to the family, and some negative results for adolescents, these studies do not allow unambiguous conclusions to be drawn about the relationship between welfare reform and child wellbeing. It is important to note that these results apply to particularly generous demonstration programs which received a great deal of attention and scrutiny and therefore were likely to have been well run (DeAngelis 2001a).

Many of the hypothesised effects of welfare reform on children, both negative and positive, are expected to be in the very long term. Given that only six years have passed since the passing of PRWORA, and that most studies are based on data over a relatively short period of time after the introduction of policy changes, it is too early to judge what the longer-term effects will be.

It is important to consider the impact of these programs for children of different ages. Parental employment may have a very different impact on children of different ages given the differences in the stage of development and need for care and supervision. The MDRC reviews of the experimental studies conclude that no impacts have been found on young children’s achievement and behaviour. However, it would be unwise to generalise as only one study assessed the impacts on young children.

For children of elementary school-age, programs that offer the most generous earnings supplements appear to have consistently achieved greater positive impacts on children’s school achievements than programs without these supplements. Although less consistently than for the school achievement outcomes, programs with earnings supplements appear to benefit children’s behaviour and health outcomes as well. By comparison, programs with mandatory employment services or time-limits had few effects across children’s behavioural and health outcomes. Concerns that the development of young children, especially elementary school-aged children, might be compromised by the stresses and disruptions wrought by welfare-to-work transitions receive virtually no support from these studies.

The evidence is unclear as to which components of family functioning appeared to have caused the beneficial changes in child wellbeing under the earnings supplement programs. Findings point to parents placing their children in formal child care or after-school activities as one important way earnings supplement programs may have affected the wellbeing of children.

The evidence for adolescents differs to that for elementary school-aged children. The results indicate that both voluntary and mandatory programs that promote work and programs with and without time-limits on benefit receipt lead to negative effects on adolescents school achievement and progress. The programs however are found to have no effects on suspensions, school dropout, or childbearing, nor do they appear to affect the school completion of older adolescents (Morris, Knox and Gennetian 2002). There is some evidence that these effects on adolescents are related to child care problems with adolescents being left unsupervised.

Lessons for Australian policy

The dramatic fall in caseloads and increases in employment rates amongst lone mothers following welfare reform raises the question as to what lessons the United States experience holds for Australia.

When considering this question it is important to keep in mind the key differences in the income support and welfare systems in Australia and the United States. Government income support provisions in Australia differ from those in most other developed countries and are very different from the US system. In the Australian system, benefits are flat-rate and paid from general government revenue; there are no earnings-related features in the government benefit system. Benefits are subject to income and assets tests, but these are generous compared to the means tests in most other OECD countries. Benefits (in one form or another) are effectively available on an indefinite basis, subject to means tests. In addition, benefits for the unemployed are subject to an activity test.¹⁷

The principle of “mutual obligation” is a recent feature of the Australian system, where certain benefit recipients are required to participate in an activity of value to the community. This has led to moves for sole parents with a youngest child aged 13 to 16 years being required to meet a part-time activity requirement, and for all lone parents with school-aged children who are receiving government income support to attend an annual interview which is designed to start recipients thinking about taking steps to become ready to enter the paid workforce.

So what can Australia learn from the United States welfare reforms? The US experience has been that when time-limits are imposed on benefit receipt in combination with increased financial incentives to take up paid employment, and increased support for working mothers in terms of subsidies to child care and active labour market programs, a substantial proportion of lone mothers are able to find employment.

The extent to which lone mothers in the US were able to find employment surprised many commentators who believed that many lone mothers with low levels of education, lack of work experience and higher rates of health problems would find it impossible to find employment. In addition, there were concerns that the US labour market simply would not be able to generate enough jobs. The consensus now is that sufficient jobs were available to absorb welfare recipients entering the labour market (Burtless 2000).

It is unlikely that the Australian labour market would be able to generate enough jobs if there was a sudden increase in the number of lone mothers entering the labour market. There are several reasons for this. First, real wage levels in Australia are higher than in the US. This is likely to make it very difficult for a substantial number of lone mothers with relatively low levels of productivity to find paid employment. Second, the US reforms were implemented during a period of extremely strong and sustained economic growth.

While the longer term increased employment rates and resulting taxes and lower numbers of recipients may lead to reductions in expenditure, in the short to medium term, and possibly the longer term, the increases in employment rates have occurred at a time when total expenditure on assistance to lowincome families with children has been substantially increased (Ellwood 2001).

Another factor central to the success of the US reforms is that there appears to have been a congruence with general community attitudes that women with children, even very young children, could justifiably be expected to be in paid employment. This provided an apparent moral legitimacy to efforts to increase employment rates. While employment rates of Australian lone mothers are lower than those in the United States, the employment rates of partnered mothers are also much lower in Australia, particularly for full-time employment. Many Australians would be uncomfortable with the idea of compelling mothers with young children to participate in the labour market. For a discussion of the attitudes of Australians towards paid work for mothers of young children see Evans and Kelley (2002).

On balance, the US evidence suggests that it was a combination of all the changes introduced by welfare reform that was instrumental in the success in reducing caseloads and increasing employment rates. However, the relative importance of the changes is unclear. Similarly, it is unclear as to what would have happened if some of the changes had not been introduced.

The Australian system already has many of the features that were introduced into the US system by welfare reform. First, in Australia, concerted efforts have been made to ensure that financial incentives to be in paid employment exist using earnings disregards and having a rate of withdrawal of benefits of substantially less than 100 per cent (Whiteford and Angenent 2001).¹⁸ However, the financial incentives to be in paid employment continues to be an issue of interest (Ingles 1998). To address this issue proposals have been made in Australia for the introduction of a form of earned income tax credit. (For a discussion of such proposals, see Keating and Lambert 1998; Dawkins 2002; Duncan 2002.)

The combination of relatively high minimum wages, relatively generous income test tapers, and the provision of in-work benefits means that many lone mothers in Australia combine part-time employment with continued receipt of government income support (Whiteford and Angenent 2001). Over the last two decades, changes to the income support system have increased financial incentives for part-time employment compared to non-employment or full-time employment. In June 2001, 26.2 per cent of lone parents receiving a pension payment (Parenting Payment Single) reported having earnings (Department of Family and Community Services 2001). This allows many lone mother families to sustain a modest but reasonable standard of living. This is a clear strength of the Australian system as compared to the United States where many lone-parent families have very low levels of income, including a substantial number of lone mothers in paid employment.

The relative attractiveness of part-time employment for lone mothers can be seen in Figure 2 which shows the part-time and full-time employment rates of lone mothers with dependent children for the period 1985 to 2000. The picture is one of dramatic change, with the part-time employment rate increasing from 13.3 per cent in 1985 to 26.0 per cent in 2000. At the same time, the full-time employment rate has fallen from 26.7 per cent to 23.2 per cent.

Figure 2: Part-time and full-time employment rates, lone mothers, 1985–2000

Source: ABS (various years), Labour Force Status and other Characteristics of Families, Australia, Catalogue No, 6224.0, Australian Bureau of Statistics, Canberra.

Australia also has relatively affordable high quality child care and an extensive system of government subsidies (Press and Hayes 2000). In addition, Australia has adopted individualised case management and has a long history of using labour market programs, including for lone mothers.¹⁹ The Jobs, Education and Training (JET) scheme was initiated in Australia in 1989. This is a voluntary labour market program, in which participants are assessed by a JET adviser, who can then refer them for education or training, fund a pre-vocational course, or refer them for other assistance in looking for work. In addition, lone mothers in receipt of an income support payment have access to additional financial assistance if they are enrolled in formal education.

This leaves perhaps the most significant change to the US system, the introduction of time-limits on receipt of benefits. There is some evidence to suggest that the impact of time-limits may not in fact be as large as is sometimes argued. For the majority of welfare recipients who reach a time-limit, the limit appears to be binding only in the case of non-compliance and for recipients who cannot persuade authorities that they are making good faith efforts to find employment.

Australia has for some time been encouraging lone mothers receiving income support payments to more actively participate in the labour market, education or training. Given that the Australian system already has many of the features introduced into the United States system, it is likely that the enhancement of work requirements will result in continued increases in the employment rates of lone mothers. A key constraint is likely to be lack of labour demand, particularly if the rate of economic growth slows. Raising the skill levels and job readiness of lone mothers via labour market programs such as the JET scheme, and the use of personal advisers to provide individualised case management, will be important to continued success in increasing the employment rates of lone mothers

References

• Acs, G. & Loprest, P. (2001), “Synthesis report on the findings from ASPE’s Leavers Grants”, US Department of Health and Human Services, Washington DC.
• Adams, G. & Rohacek, M. (2002a), “Child care and welfare reform”, in A. Weil and K. Finegold (eds), Welfare Reform: The Next Act, The Urban Institute Press, Washington DC.
• Adams, G. & Rohacek, M. (2002b), “Child care and welfare reform”, Welfare Reform and Beyond, Policy Brief No. 14, The Brookings Institution, Washington DC
• AIFS (2002), Annual Report 2001–2002, Australian Institute of Family Studies, Melbourne.
• Barrett, G. (2002), “The dynamics of participation in the sole parent pension”,Economic Record, vol. 78, no. 240, pp. 1-17.
• Bavier, R. (2001), “Welfare reform data from the Survey of Income and Program Participation”, Monthly Labor Review, vol. 124, no. 7, pp. 13-24.
• Blank, R. & Schoeni, R. (2000), “What has welfare reform accomplished? Impacts on welfare participation, employment, income, poverty, and family structure”, Working Paper No. 7627, National Bureau of Economic Research, Cambridge, Massachusetts.
• Blank, R. (2001), “Welfare and the Economy”, Welfare Reform and Beyond, Policy Brief No. 7, The Brookings Institution, Washington DC.
• Bloom, D., Farrell, M. & Fink, B. (with Adams-Ciardullo, D.) (2002), Welfare Time Limits: State Policies, Implementation, and Effects on Families, Report by Manpower Demonstration Research Corporation to the Department of Health and Community Services, Manpower Demonstration Research Corporation, New York.
• Bloom, D. & Winstead, D. (2002), Sanctions and Welfare Reform, Policy Brief No. 12, The Brookings Institution, Washington DC. 
  Burtless, G. (1995), “Employment prospects of welfare recipients”, in D.S.
• Nightengale and R.H. Haveman (eds), The Work Alternative: Welfare Reform and the Realities of the Job Market, Urban Institute, Washington DC.
• Burtless, G. (2000), “Can the labor market absorb three million welfare recipients?” in K. Kaye and D. Nightingale (eds), The Low-Wage Labor Market: Challenges and Opportunities for Economic Self-Sufficiency, Urban Institute Press, Washington DC.
• Burtless, G. (2002), “Can supply-side policies reduce unemployment? Lessons from North America”, Australian Economic Review, vol. 35, no. 1, pp. 3-28.
• Cancian, M., Haveman, R., Meyer, D. & Wolfe, B. (2002), “Before and after TANF: The economic wellbeing of women leaving welfare”, Discussion Paper No.1244-02, Institute for Research on Poverty, Madison, Wisconsin.
• Campbell, N., Maniha, J. & Rolston, H. (2002), Job Retention and Advancement in Welfare Reform, Policy Brief No. 18, The Brookings Institution, Washington DC.
• Chalmers, J. (1999), Sole Parent Exit Study: Final Report, Report to the Department of Family and Community Services, Canberra.
• Corcoran, M., Danziger, S.K., Kalil, A. & Seefeldt, K.S. (2000), “How welfare reform is affecting women’s work”, Annual Review of Sociology, vol. 26, pp. 241-269.
• Danziger, S., Heflin, C., Corcoran, M., Oltmans, E. & Wang, H. (2002), “Does it pay to move from welfare to work?”, Unpublished mimeo.
• Dawkins, P. (2002), “The ‘five economists’ plan: The original idea and further developments”, Australian Journal of Labour Economics, vol. 5, no. 2, pp. 203-230.
• DeAngelis, T. (2001a), “Young children benefit from experimental welfare programs”, Monitor on Psychology, October, pp. 73-75.
• DeAngelis, T. (2001b), “Welfare reform and women: Five years later”, Monitor on Psychology, October, pp. 70-72.
• Department of Family And Community Services (2001), Annual Report 2000–2001, Commonwealth of Australia, Canberra.
• Duncan, A. (2002), “Promoting employment through welfare reform: Lessons from the past, prospects for the future”, Paper for the 2002 R I Downing Lecture, Faculty of Economics and Commerce, University of Melbourne, Melbourne.
• Ellwood, D. (2002), “The US vision of work-based reform: Promise, prospect and pitfalls”, in T. Eardley and B. Bradbury (eds), Competing Visions: Refereed Proceedings of the National Social Policy Conference 2001, SPRC Report 1/02, Social Policy Research Centre, University of New South Wales, Sydney.
• Evans, M. & Kelley, J. (2002), “Changes in public attitudes to maternal employment: Australia, 1984 to 2001”, People and Place, vol. 10, no. 1, pp. 42-56.
• Finegold, K. & Staveteig, S. (2002), “Race, ethnicity and welfare reform”, in A.Weil and K. Finegold (eds), Welfare Reform: The Next Act, The Urban Institute Press, Washington DC.
• Gregory, R. (1999), “Children and the changing labour market: Joblessness in families with dependent children”, Centre for Economic Policy Research Discussion Paper No. 406, Centre for Economic Policy Research, Australian National University, Canberra.
• Gregory, R. (2002), “Can this be the promised land? Work and welfare for the modern woman”, Notes to accompany the National Institutes Public Lecture, 5 June 2002, Parliament House, Canberra.
• Gueron, J. & Hamilton, G. (2002), “The role of education and training in welfare reform”, Welfare Reform and Beyond, Policy Brief No. 20, Brookings Institution, Washington DC.
• Haskins, R., Sawhill, I. & Weaver, K. (2001), “Welfare reform reauthorization: An overview of problems and issues”, Welfare Reform and Beyond, Policy Brief No. 2, Brookings Institution, Washington DC.
• Howe, B. & Pidwell, R. (2002), “After AFDC: Lone mothers and their children”,Just Policy, vol. 26, May, pp. 48-55.
• Ingles, D. (1998), “Overcoming anomalies in the interaction of tax and social security”, Australian Economic Review, vol. 31, no. 3, pp. 271-280.
• Kamerman, S. & Kahn, A. (2001), “Child and family policies in an era of social policy retrenchment and restructuring”, in K. Vleminckx and T. Smeeding (eds), Child Wellbeing, Child Poverty and Child Policy in Modern Nations: What do we Know?, The Policy Press, Bristol.
• Keating, M. & Lambert, S. (1998), “Improving incentives: Changing the interface of tax and social security”, Australian Economic Review, vol. 31, no. 3, pp. 281-289.
• Lichter, D. & Jayakody, R. (2002), “Welfare reform: How do we measure success?”, Annual Review of Sociology, vol. 28, pp. 117-141.
• Loprest, P. (2002), “Making the transition from welfare to work: Successes but continuing concerns”, in A. Weil and K. Finegold (eds), Welfare Reform: The Next Act, The Urban Institute Press, Washington DC.
• Magnum, G., Sum, A. & Fogg, N. (2000), “Poverty ain’t what it used to be”, Challenge, vol. 43, no. 2, pp. 97-130.
• Meyers, M., Han, W., Waldfogel, J. & Garfinkel, I. (2001), “Child care in the wake of welfare reform: The impact of government subsidies on the economic wellbeing of single-mother families”, Social Service Review, pp. 29-59.
• Millar, J. & Rowlingson, K. (eds) (2001), Lone Parents, Employment and Social Policy: Cross-national Comparisons, Policy Press, Bristol.
• Moffitt, R. (2001), “The Temporary Assistance for Needy Families Program”, Paper prepared for the NBER Conference Means-Tested Programs in the United States, May 11-12, Cambridge Massachusetts.
• Moffitt, R. (2002), “From welfare to work: What the evidence shows”, Welfare Reform and Beyond, Policy Brief No. 13, The Brookings Institution, Washington DC.
• Moore, K., Glei, D., Driscoll, A., Zaslow, M. & Redd, Z. (2002), “Poverty and welfare patterns: Implications for children”, Journal of Social Policy, vol. 31, no. 2, pp. 207-227.
• Morris, P. & Duncan, G. (2001), “Which welfare reforms are best for children”, Welfare Reform and Beyond, Policy Brief No. 6, The Brookings Institution, Washington DC.
• Morris, P., Knox, V. & Gennetian, L. (2002), Welfare Policies Matter for Children and Youth: Lessons for TANF Reauthorisation, Manpower Demonstration and Research Corporation, New York.
• Morris, P., Huston, A., Duncan, G., Crosby, D. & Bos, J. (2001), How Welfare and Work Policies Affect Children: A Synthesis of Research, Manpower Demonstration Research Corporation, New York.
• Ogborn, K. (1984), “Training scheme for widow pensioners 1968-1974”, Social Security Journal, December, pp. 1-10.
• Perry, J. (2002), “One language, three accents: Welfare reform in the United States, the United Kingdom and Australia”, Family Matters, no. 56, pp. 40-47.
• Pollack, H., Danziger, S., Seefeldt, K. & Jayakody, R. (2002), “Substance use among welfare recipients: Trends and policy responses”, Social Service Review, pp. 256-274.
• Press, F. & Hayes, A. (2000), OECD Thematic Review of Early Childhood Education and Care Policy: Australian Background Report, Commonwealth of Australia, Canberra.
• Rangarajan, A. & Novak, T. (1999), The Struggle to Sustain Employment: The Effectiveness of the Postemployment Service Demonstration, Mathematica, Princeton.
• Saunders, P. (ed.) (2000), Reforming the Australian Welfare State, Australian Institute of Family Studies, Melbourne.
• Stanton, D. (1973), “Determining the poverty line”, Social Security Quarterly, Spring, pp. 18-32.
• Waldfogel, J., Danziger, S.K., Danziger, S. & Seefeldt, K. (2001), “Welfare reform and lone mothers’ employment in the US”, in J. Millar and K. Rowlingson (eds), Lone Parents, Employment and Social Policy: Cross-national Comparisons, Policy Press, Bristol.
• Weil, A. & Holan, J. (2001), “Health insurance, welfare and work”, Welfare Reform and Beyond Policy, Brief No. 11, Brookings Institute, Washington DC.
• Whiteford, P. & Angenent, G. (2001), “The Australian system of social protection: An overview”, Occasional Paper Number 6, Department of Family and Community Services, Canberra.
• Whiteford, P., Stanton, D. & Gray, M. (2001), “Families and income security: Changing patterns of social security and related policy issues”, Family Matters, no. 60, pp. 24-35.
• Zedlewski, S. (2002), “Family incomes: Rising, falling or holding steady?”, in A.Weil and K. Finegold (eds), Welfare Reform: The Next Act, The Urban Institute Press, Washington D.C.