In 2002 Congress will revisit Temporary Assistance for Needy Families (TANF), often known as welfare reform. Many progressives, ourselves included, fought hard against the program that passed in 1996. We judged it too punitive and too far from the spirit of progressive reform, which would have focused less on reducing caseloads and more on both promoting employment and improving the well-being of low-income families with children. We worried that the low-wage labor market, which had been deteriorating for decades, provided little opportunity for families forced to leave public assistance. We feared that without work supports, such as child care and expanded earnings subsidies, the economic circumstances of some of our most vulnerable families would be severely diminished. We argued that the block grant funding approach of the new program revoked the important countercyclical feature of the entitlement program that TANF replaced.

So far, the evidence reveals that many of our fears have not been borne out, at least not to the extent we predicted. The labor market, particularly the low-wage sector, improved in ways we never foresaw. This in turn led to the first persistent real-wage gains experienced there in two-and-a-half decades. These wage gains, the new welfare policy, and other pro-work policies have attracted more low-income single mothers into paid employment. Some states, albeit too few, made considerable efforts to smooth the path to the labor market by providing the needed work supports to both the poor and near poor. And, thankfully, there has as yet been no recession.

Yet while many of our fears have not been realized, it remains the case that TANF has not focused enough on the goal of true reform: the improvement of the economic well-being of poor families with children. Caseloads have fallen sharply and employment rates have soared, but welfare reform has not fundamentally improved the living standards of many of the families it has affected. And if that's the case in this, the best economy in 30 years, what can we expect in a downturn?

At the heart of the TANF reauthorization debate is an assessment of the relative roles of welfare reform itself and the strong economy. This is crucial if we are to avoid over- or under-crediting the policy change. We also believe this debate should not stop at the gates of welfare reform, but should address the larger question of how to lift the living standards of all working families, particularly those who have only recently benefited from the boom. These families have long responded to the personal responsibility clauses enshrined in the law that created TANF, yet their good-faith efforts are inadequately reciprocated by public obligation. So the reauthorization debate represents a historical opportunity to frame a set of policies outside of the welfare system designed to end working poverty as we know it.

Certainly, the 107th Congress may not be very receptive to a progressive set of reforms, especially since many will be arguing that the program has been an unqualified success. Yet as has been written in these pages in recent weeks, this is a fine time to introduce a progressive agenda built around work. The core idea is that those who make a good-faith contribution to the nation's economy should never live in privation. They and their children should see their living standards rise over time, and if the market fails to deliver that result, then there is an explicit role for public policy to do so.

What Welfare Reform Did

Known as the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA), the controversial law included far more than welfare reform. Much of it was simply designed to cut benefits or spending in low-income programs. For example, the law sharply limited the situations in which legal immigrants can qualify for public benefits, narrowed the circumstances in which children qualify for disability benefits, and imposed an array of large and small reductions in the Food Stamp Program.

The centerpiece, though, was repeal of the Aid to Families with Dependent Children program, and the enactment of a system of block grants to states--Temporary Assistance for Needy Families. In the block grant structure, states qualify each year for a lump sum of federal money, with most states' allocations basically frozen at 1994 or 1995 federal funding levels through 2002. With their block grants, states were expected to run time-limited, work-oriented programs of cash assistance for poor families. The law eliminated federal entitlements: No family has a federal right to assistance, and states have no obligation whatsoever to provide families with welfare benefits. States cannot use federal funds to assist families for more than 60 months (subject to limited exceptions) but are free to provide assistance for shorter periods. States can use their block grant funds to help families prepare for, find, and keep jobs, but states are not required to do so and can use their block grant funds for a wide array of purposes. The law created a strong incentive to cut welfare caseloads--because there was no duty to assist families, states knew that they would get the same amount of funds whether their caseloads went up or down, and the surest way to avoid federal penalties was to bring down the state's caseload.

What was Congress trying to achieve? Different people had different goals, and the law reflects these differing views. For some the 1996 law was largely about cutting welfare caseloads or reducing spending; for some it was about promoting work; for some it was about broadening state flexibility, reducing federal authority, and curtailing individual rights; and for some it was about reducing out-of-wedlock births. For much of the public, though, the goal was that people who were able to work should do so. Many in the progressive community shared this goal but feared that the law's approach--freezing federal spending, ending individual rights, imposing time limits, creating strong incentives to cut caseloads--would mean that instead of helping parents enter and progress in the labor force, states would simply restrict assistance for families that needed help. And many progressives feared the consequences if public assistance was denied to families with the weakest labor market prospects in a low-wage labor market that was already failing many of its participants.

Throughout the 1996 debates, discussion of one goal was conspicuously lacking: There was much talk about the need to promote work and reduce welfare, but little discussion of the need to reduce poverty and promote the well-being of low-income families. Instead, both conservatives and, to a great extent, the Clinton administration, created a picture in which the principal problem was seen as too many families on welfare for too long. The obvious solution was to cut caseloads by getting families to leave welfare.

Once the law was in place, states had very broad discretion in designing their policies. As a practical matter, though, most states moved in the same direction. All but two states imposed time limits on assistance, with most electing a five-year limit (about 20 set shorter time limits). Most heightened the penalties for families that missed appointments or otherwise didn't meet work-related requirements; in two-thirds of states, cash assistance can now be cut off when a family is "sanctioned" for noncompliance. Work-related activities were required, and the opportunity to participate in education and training was sharply restricted. At the same time, most states also expanded supports for families that entered low-wage jobs, by extending eligibility for welfare when a parent was employed and by increasing the availability of child care and other supports.

What Hath TANF Wrought?

In trying to understand what has happened since 1996, there are really two stories to appreciate: One concerns what has happened to families, and the other concerns what the block grant structure has meant for state budgets and spending on low-income initiatives.

When we look at the impacts on families, the challenge is to disentangle the changes in welfare from those in the larger economy. Three trends are critical indicators: caseload decline, the increase in employment of poor single mothers, and the improvement in the low-wage labor market. The first two phenomena are partially related to welfare reform, while the third has occurred despite the increase in the supply of low-wage labor that welfare reform induced.

The impact most often cited is the decline in welfare caseloads. The number of caseloads began falling in 1994, but the decline greatly accelerated after the 1996 law was passed. In early 1994, 5.0 million families were receiving assistance. By the time the law passed, the number had fallen to 4.4 million; and by December 1999, the number was 2.4 million. In 1994, 14 percent of all children were receiving welfare benefits; by 1999 that figure had been cut by half.

But the fall in unemployment and the appearance of more and better employment opportunities in the low-wage labor market coincided with the introduction of welfare reform, raising the question of how much of the caseload decline is actually the result of welfare waivers and TANF ("waivers" were state experiments with welfare reform tried before the passage of TANF). After some debate, the current academic consensus seems to be that somewhere between 15 percent and 30 percent of the decline is attributable to welfare reform, though some estimates are higher. Of course, welfare caseloads can fall either because fewer families need help or because needy families are diverted from the rolls, and these have very different implications for evaluating the impact of reform. In fact, both dynamics are at play.

The share of poor single mothers in the paid-labor market is at an all-time high. The annual employment rates of women who received welfare benefits at some point in the year were relatively flat in the late 1980s and early 1990s. But around the mid-1990s, when welfare reform was being phased in, they took off, growing from 39 percent in 1994 to 57 percent in 1999. Of course, this was also the period when the labor market tightened up. So the challenge, again, is to figure out how much of this trend is specifically caused by welfare reform.

One useful approach is to compare the increase in work among those most likely to be affected by the policy, low-income single mothers of young children, to those less likely to be affected--namely, low-income married mothers with young kids. Such tabulations by the U.S Department of Health and Human Services reveal that large increases in employment rates in the latter 1990s were concentrated among single mothers, not married ones. Between 1994 and 1999, low-income married moms' employment rates were unchanged at 39 percent. Those of low-income single mothers grew from that same level (39 percent) to 55 percent. Another convincing analysis (by Signe-Mary McKernan and her colleagues, and available at the Urban Institute's Web site) goes further than this, comparing employment among these same types of women in waiver versus nonwaiver states. These authors attribute employment rate increases of 6 to 8 percentage points over the latter 1990s to welfare reform. By their estimates, given the magnitude of the full increase in employment rates by single mothers, reform could explain as much as half of the increase.

Although this is compelling evidence of the impact of welfare reform, other developments significantly enhanced the incentive to work in the paid-labor market. During this same period in the latter 1990s, the overall unemployment rate fell from about 6 percent to 4 percent, a 30-year low. This surge in labor demand was particularly important at the low end of the labor market. Here the decline of 2 percentage points translated into declines of more than 10 percentage points for some disadvantaged groups of workers, such as young, less-skilled, non-college-educated African Americans, whose rate of unemployment fell from about 30 percent to 20 percent over this period. Obviously, 20 percent is still alarmingly high, but the steep decline in unemployment rates is a dramatic indicator of the benefits that accrued to minorities from the tight job market.

Along with more employment opportunities, both pretax wages and after-tax earnings of low-wage workers have increased dramatically since the latter 1990s. On the wage side, the primary factors are the tight labor market and the 1996-1997 minimum wage hike (with the former being more important). These changes led to significant real-wage gains among low-wage workers for the first time in two decades. For example, in the 16 years between 1979 and 1995, the real wages of women earning at the 20th percentile of income fell by 8 percent; from 1995 to 1999, it grew 8 percent, almost fully reversing the decade-and-a-half decline. Earlier in the decade, the Earned Income Tax Credit (EITC) was significantly expanded, leading to higher after-tax earnings for low-wage workers in low-income families.

Life after Welfare

To get a more precise picture of reform's impact on those who have left the welfare rolls, we need to turn to the so-called leaver studies. These projects, mostly conducted by individual states, use surveys or administrative data to track former recipients. Here are some key findings.

¥ At any given point in time, the majority of former welfare recipients are employed, but for many the connection to the labor market is quite tenuous. Both national and state studies find that about 60 percent of leavers are currently working, with 70 percent or more having worked at some point over the course of the year. But only about 40 percent worked consistently throughout the year. Why? Women leaving welfare tend to face more employment barriers--low-skills, lack of transportation, sick or disabled children--than other populations. When they do work, however, most tend to work full time.

¥ Recipients who leave welfare for work earn very low wages. The studies are unanimous on this point; they document wages in the range of $6 to $8 per hour. It is too soon to tell whether the minority with solid labor force attachment will be upwardly mobile, but the analysis done so far is not encouraging. In a compilation of nine state and local leaver studies, most involving families that left welfare in late 1996, nominal median earnings were $2,526 in the first quarter after leaving welfare and $2,821 in the fourth quarter. Another study, which followed the families that left the Wisconsin welfare rolls in late 1995, found that for those with earned income, median earnings in constant 1998 dollars were $8,608 in the first year, $9,627 in the second, and $10,924 in the third. Thus, three years after leaving assistance, median family earnings were still insufficient to reach the poverty line for a family of three ($14,150 per year).

¥ A significant share of leavers are not working and have very high poverty rates. Nationwide, probably about 40 percent of leavers aren't working. This group is far more likely to have limited education and work history and to run into other obstacles that make employment more difficult. Typically, the most common reason they give for not working is illness or disability. Some states have made extensive use of "sanctions"--cutting off assistance to those who don't meet attendance or other requirements. Studies of the sanctioned group consistently find that they have less education and less prior work history. A recent Michigan study found that they were also more likely to have been victims of domestic violence and more likely to be mentally ill. A three-city study found that sanctioned leavers had an 89 percent poverty rate after leaving welfare.

¥ After leaving assistance, many families lose their benefits under the Food Stamp Program and Medicaid, even though they are still poor; most working leavers aren't receiving child care subsidy assistance. Participation in the Food Stamp Program and Medicaid drops dramatically after families leave welfare. In a national sample of families that had left welfare in the previous six months, only about half were receiving food stamps or Medicaid. Fewer than one-third of working leavers are receiving child care subsidies. Reasons for this low participation are many: administrative mistakes, lack of information, families wanting to leave stigmatized systems that treat them badly. The net result, though, is that the support systems intended for low-earning working families aren't reaching many in need.

¥ Working as well as nonworking leavers often face financial deprivation after leaving welfare. Nevertheless, the majority indicate that life is better after having left. In a national study conducted by the Urban Institute, 33 percent of leavers reported that they had had to cut the size of meals or skip them entirely because there wasn't enough food; 39 percent reported that there had been a time in the last year when they had been unable to afford to pay rent, mortgage or utilities; and 7 percent reported having had to move in with others because of inability to pay these housing costs. State studies report similar findings. At the same time, state studies consistently find that roughly half of those surveyed report that life is better since having left welfare--and that if they could choose to go back on welfare, they would not want to do so.

In sum, these studies suggest that two of the goals stated above--reducing caseloads and increasing employment--have certainly occurred. And while the strong economy has contributed, welfare reform has also played a clear role in both. But what about the most important, if least discussed, goal--improving the well-being of low-income families with children?

Here the story is at best a mixed one. The data tell us that most leavers are working but that most are not employed on an ongoing, sustained basis. The families that are working typically have higher cash incomes than they had before, but they face higher expenses (transportation, clothing, day care). At the same time, the necessary work supports--food stamps, Medicaid, and child care subsidies--are often not reaching them. And a large minority of families--inside and outside of the welfare system--remain jobless despite the economy's unprecedented strength. Based on the evidence so far, the story is surely more positive than many anticipated in 1996. But welfare reform cannot be called an unqualified success.

How Did States Spend the Money?

In 1996 one of the most bitterly contested features of the new structure was the shift in financing from an open-ended matching program (in which federal funds matched state dollars for all allowable costs) to a block grant, whereby states would receive the same amount of federal funds each year whether their welfare caseloads went up or down. At the time, we like many others expressed great concern about how a block grant structure would function during a recession, when the need for assistance could be expected to escalate. We still don't know how the block grant structure will function in a recession; but providing states with constant funding while caseloads are shrinking has had a different and largely unanticipated effect: It has meant that a historically large amount of money has been made available to states not just for running their welfare programs but for financing an array of other initiatives for low-income (and in some cases, not-so-low-income) families.

Nationwide, the federal government provides states with about $16.5 billion in block grant funds each year, and states are required to spend an additional $10 billion to $11 billion in state funds in order to qualify for their block grants. Since 1994, however, the annual spending on cash assistance for poor families through the welfare system fell from $23 billion to less than $14 billion. As a result, states can now spend almost half (and in some states, far more than half) of their block grants for purposes other than cash assistance.

Where is this money going? It's impossible to get a precise picture, but it is clear that in some instances the money is funding very worthy programs. In 1999 about $3 billion in TANF funds was directed toward child care for low-income families. And some states have used their TANF funds to help establish or expand refundable earned-income tax credits, increase job-training expenditures, expand access to community colleges for low-income families, fund programs that promote individual-development accounts, expand services for victims of domestic violence, and pay for an array of other constructive ventures. At the same time, there have been examples of what has become known as "supplantation"--substituting federal funds for existing state spending and redirecting the freed-up state dollars to purposes that have little or nothing to do with helping low-income families.

So what, on balance, has implementation of the TANF block grant meant? While we can't get an exact accounting, it seems clear that as welfare spending has gone down, state spending for low income families--particularly, low-income working families--has been higher than it would have been without the block grant. In other words, while many people viewed TANF as a cut in federal spending in 1996, the effect has been the opposite as state initiatives to help low-income working families have benefited from the available funds. At the same time, in some states, there is deep frustration, based on the sense that a significant opportunity has been missed because too little of the freed-up funds has gone to new investments to support low-income families [see Michael Massing, "Ending Poverty as We Know It," TAP, June 19-July 3, 2000].

Visit TAP Online's Special Segment on Children and Families

Personal Responsibility and Public Obligation

TANF reauthorization may represent the best opportunity in the next congressional session to articulate progressive values and visions about welfare policy. To be sure, the balance of power at the federal level will limit the potential for progressive change. But we believe that what's occurred thus far can be framed in a way that builds on welfare reform's more positive aspects. Moreover, to work solely within the confines of a welfare program, however progressive, is to miss a historic opportunity to expand our vision of reform to include improvement of the social welfare of all families, not simply those leaving the welfare rolls.

The starting premise is that if the polity wants to promote work and improve the well-being of families with children, personal responsibility must be accompanied by public obligation. As the evidence has shown, most former welfare families have embraced the "PR"--the personal responsibility--in PRWORA (indeed, many had always embraced it). Millions more low-income working families never joined the welfare rolls. These families are playing by the rules; yet even in the best economy in decades, they need more help. For others personal responsibility by itself will never lead to sustained employment without help from government. The best use of reauthorization would be to build on and expand the positive aspects of welfare reform--policies designed to ensure the economic well-being of working families--and to restructure those components of the current law that work against this goal. We also need to remember that a recession may be lurking out there somewhere, and that the legislation, as it stands, is unprepared for this looming possibility.

The reauthorization agenda should include at least these six points.

1. Change the law's central focus from reducing caseloads to reducing poverty. In 1996 Congress emphasized the need to cut welfare caseloads and states responded impressively. But can states respond equally well to a national goal of reducing, and ultimately eliminating, child and family poverty? Block grant funds alone are not enough to accomplish this goal. Yet states should be required to explain how they will use block grant funds and other state resources to fight poverty, and they should be measured by their success in doing so. This does not mean a shift away from trying to promote employment: The most straightforward way to reduce family poverty today is to help parents enter the labor force, maintain employment, and gain access to the supports that are supposed to be available to working families.

2. Increase funding and make states more accountable for how they use their funds. If states are being asked to broaden their focus from reducing welfare to addressing poverty, then they will need more resources, even if their welfare caseloads have shrunk. There should be continuing efforts to enhance the funding in states that receive the least resources in relation to their population of low-income families. At the same time, states should describe how they plan to use block grant funds, report on how they actually used them, and make a commitment that funds will not be diverted to refinance other parts of the state budget.

3. Expand and improve the supports for low-income working families. The availability of child care and health care assistance needs to be increased, and there should be a major effort to simplify and improve the accessibility of food stamps, Medicaid, and other benefits that could improve the well-being of families headed by parents in low-wage jobs.

4. Revisit the federal time limit. In most states, the families remaining on the welfare rolls haven't reached their time limits yet, but they soon will. There is no good reason to cut off help to families in which the parents cannot find steady work or earn enough to support a family. At minimum, time limit rules should be revised to allow more flexibility. States should be allowed to stop the clock for working families while continuing to provide assistance. States should also be able to provide extensions to working families and families in which a parent cannot find or maintain work. And states should be encouraged to operate publicly funded jobs programs instead of terminating assistance for families who are able to do some work but unable to maintain unsubsidized employment.

5. Eliminate the federal bias against education and training efforts. At the very least, it should be up to each state to decide what role education and training will play in its welfare efforts. Federal law could go further and require states to explain how they will use block grant funds to expand access to education and training programs for low-income families and how they will effectively coordinate their welfare reform efforts with broader state strategies for work force development.

6. Build in economic stabilizers. The strong economy and, in particular, the tight labor market have been central to TANF's functioning thus far. In the absence of full employment, work-based welfare reform will be severely challenged unless it ceases to depend solely on private-sector employment opportunities. The current law has a "contingency fund" intended to make additional resources available to states during an economic downturn; but as a practical matter, the fund is structured in a way that will be of little or no use to all or most states. A true contingency plan for a recession (or even for a weaker labor market) would involve providing states access to additional funding during an economic downturn and encouraging them to provide both assistance and, when needed, publicly funded employment if private-sector employment is unavailable.

Beyond TANF

The larger goal of public policy should be a transformation of the low-wage labor market and the economic prospects of the working poor. Historically, working families have been forced to turn to welfare for two reasons: first, because other systems--such as child support, unemployment compensation, the disability benefits system, and parental leave--failed or were not available; and second, because the compensation available to working parents in the low-wage labor market was insufficient to meet their families' basic needs. The next stage of welfare reform provides an opportunity to foster a broader discussion of policies that can promote work and support families. We have in mind a four-legged policy stool.

1. Maintain full employment. Throughout this piece, we've noted the importance of tight labor markets. Maintaining full employment is still the most effective social policy in terms of lifting the economic prospects of poor working families. Much of what we strive for pales by comparison. Yet too often progressives have treated full employment as a happy accident--as something that's outside our purview.

The reversal of two decades of declining earnings among low-wage workers was largely a function of achieving full employment in the latter 1990s. This was no accident of fate. It stemmed largely from the actions of the Federal Reserve; the Feds' liberalized interest rate policy allowed the unemployment rate to fall below the level that most economists had argued would trigger runaway inflation. True, other factors were at play, including a strong dollar holding down import prices and a surge in productivity growth, but other factors are always at play. It now looks as if the old speed-limit rules were not only wrong; they consigned millions of poor working families to falling real wages and incomes during the 1980s and early 1990s. We must use the reauthorization debate to enshrine full employment as our most sacred social policy.

2. Raise pretax wages. The minimum wage needs to be set high enough to create a reasonable floor on the low-wage labor market. Given the limited bargaining power of low-wage workers (only about 5 percent belong to unions), the minimum wage is a crucial labor market institution protecting the lowest-paid workers--who are disproportionately minorities and females--from exploitation. Of course, every time an increase is proposed, opponents trot out the argument that the increase will hurt its intended beneficiaries by pricing them out of the labor market. But reams of research now show that this is a largely bogus argument: Moderate increases in the minimum wage have almost no identifiable impact on employment. Again, the latter 1990s are an excellent example.

A much more challenging question is how high we should push the minimum wage. Though the moderate increases of the 1990s failed to generate job losses, if the minimum were raised high enough, negative effects would appear. Our research suggests that the minimum could at least be set back to its 1979 peak of $6.75 in today's dollars without generating job losses [see the minimum wage section of the Economic Policy Institute's Web site (http://www.epinet.org/)]. Reopening the indexing debate would also be worthwhile.

3. Raise after-tax earnings. Here the best policy by far is the EITC. It is politically popular, its antipoverty effects are well established, and it is very finely targeted at poor working families. Some problems with the EITC have recently been identified, such as the marriage penalty and the high marginal tax rate faced by those who are on the downslope of the program. But excellent plans are afoot to repair these, expand the policy, and lift the incomes of the working poor and near-poor higher still [see Max B. Sawicky's article "It Takes a Tax Credit to Raise a Child," on page 23 of this special report]. Also, 15 states have introduced add-on earned-income tax credits, thus allowing families that receive the federal credit to collect an additional, smaller benefit from the state. (In only 10 of these cases are the credits refundable, so the benefit is unlikely to reach most of the working poor in the other five states).

At the same time, the EITC shouldn't be the only policy approach used to raise the incomes of low-earning families. Ninety-nine percent of those who receive it get it only once a year, as a lump sum; and while many families like the "forced savings" aspect of the EITC, it means that the credit does not function as a way to meet ongoing, month-to-month costs.

4. Support work. Regardless of a person's income, work-related expenses can be very significant. Research shows that for low-income families the costs of working can easily absorb 30 percent of family income. Given the pay scales in the low-wage labor market, the instability of employment there (which is itself related to the lack of work supports), and the lack of fringe benefits such as paid sick days or maternity leave, low-wage workers cannot meet their basic needs without help. Perhaps the most obvious barrier to steady employment for these families is child care. Transportation assistance has also proved useful, particularly in solving spatial mismatch problems in areas where the laborers can't get to the jobs. Health insurance is another obvious problem for the working poor.

Economist Randy Albelda has made the important point that along with making sure moms are job ready, we ought to make sure that jobs are "mom ready." That is, single parents--or for that matter, low-income families with two working parents--can too easily be thrown into poverty when they hit a bump in the road, whether it's a sick child, a flat tire, or the loss of a temporary job. One good idea is paid family and medical leave, perhaps financed through the unemployment insurance trust fund.

In a similar vein, advocates for the working poor have long realized that the nation's unemployment insurance program--a system principally designed with full-time working males in mind--has failed to keep pace with the evolution of either the labor market or other social policies, such as TANF. The increase in part-time and temporary work along with job instability among low-income parents, together with the loss of public assistance as an entitlement, has meant that displaced low-wage workers have nowhere to turn for support when facing a gap in paid work. Good ideas for reforming the system abound (see the Annie E. Casey Foundation's "Making Wages Work" Web site--http://www.makingwageswork.org/--for an excellent review of them).

The last few years have witnessed some truly amazing changes in social and economic policy. A Democratic president signed a largely Republican bill that ended the entitlement of welfare and emphasized work more than any past reform effort had done. At the same time, economic conditions in the low-wage labor market improved far more than we could ever have expected.

The overlap of these two events has led to greatly reduced welfare caseloads and a lot more single mothers working in the paid labor market. That much is known. But as TANF comes up for reauthorization in 2002, a much more fundamental question must be addressed: Has welfare reform improved the well-being of poor families with children?

Our review of the evidence suggests that for some it has but for many it has not. There's more work but not much more disposable income, especially after one takes into account the expenses associated with work. For the families who haven't been able to break into the labor market, the tattered safety net is providing less help than ever. Furthermore, the TANF program, which has been greatly supported by the strong economy, is not prepared for the next recession, which seems to be edging closer by the quarter.

Yet it would be a mistake to write welfare reform off as a failure. The evidence does not support such a judgment. Some states put their surplus welfare dollars to good use, paving the way into the labor market with earnings subsidies and work supports. These programs in tandem with the tight labor market provided employment opportunities that seem to have made a real difference in the lives of former welfare recipients, many of whom report that they are happier with their lives now that they are off public assistance. It appears that these families have a sense of hope for the future that was absent in the past.

The reauthorization debate should be used to keep that hope alive and to extend it to all low-income families. To do so means fixing the program from within by strengthening the positive aspects of the welfare-to-work approach and giving states the resources and responsibility to make work pay and to sustain the gains made so far even through the next recession. It means softening some of the punitive aspects of the program that have been shown to cause unnecessary harm to some of our most vulnerable families.

But it also means thinking beyond TANF, toward a progressive agenda tied to work that emphasizes the public obligation that should come with personal responsibility. We have framed these policies in terms of earnings and work supports, but it is early enough in the debate to begin this dialogue in earnest; new ideas may surface. Our hope is that a coalition will form--a coalition composed of those affected directly by the policy, their advocates, and their political representatives. Armed with a balanced view of TANF's impact that is based on the evidence thus far, and with a policy agenda targeted both at and beyond TANF, such a coalition can make a real and lasting difference in the lives of working Americans. ¤