Lucie Schmidt on How The Social Safety Net Has Changed Since Welfare Reform

Siers-Poisson [00:00:06] Hello, and thanks for joining us for the Poverty Research and Policy podcast from the Institute for Research on Poverty at the University of Wisconsin–Madison. I’m Judith Siers-Poisson. For this episode, Dr. Lucie Schmidt joins us to discuss her recent paper titled, “Did Welfare Reform End the Safety Net as We Knew It? The Record Since 1996.” That paper was co-authored with Lara Shore-Sheppard and Tara Watson. You can find a link to that paper in the program note for this episode. Lucie Schmidt is the Robert A. Woods Professor of Economics at Smith College and is a research associate at the National Bureau of Economic Research. She’s also an IRP affiliate. Lucie, thanks for joining us today.

Schmidt [00:00:47] Thanks for having me.

Siers-Poisson [00:00:48] I’d like to start with just a brief description of the safety net before 1996. What did it look like?

Schmidt [00:00:56] So the U.S. Social Safety Net is interesting in that it’s always been a patchwork of different programs that were created at different times with different intended beneficiaries. It was never really meant to be a coherent system. So, in the years right before welfare reform in 1996, I think the major programs that we think about are cash transfers through the Aid to Families with Dependent Children or AFDC program. What we would call in-kind benefits, so benefits that are not in the form of cash—think of things like food benefits through food stamps or public health insurance through Medicaid—and then a refundable tax credit, which I’ll explain a little bit more later for families with earnings, the Earned Income Tax Credit. So, all of these programs were important and had been growing in the years leading up to 1996.

Siers-Poisson [00:01:49] And what were common perceptions of what welfare was for and who should get it leading up to that critical year of 1996?

Schmidt [00:01:59] So, In the US, cash transfers through the AFDC program were primarily available to single parent families. That was the population that we tended to think of when we thought about welfare programs. And even the food benefits through food stamps and the Medicaid benefits, the public health insurance, were linked in many ways to that cash assistance and therefore served that same group of beneficiaries. And in general, the way that these programs were set up is that the highest level of benefits was received by the most economically vulnerable people, those who had no earnings levels.

Siers-Poisson [00:02:37] So the Clinton-era Personal Responsibility and Work Opportunity Reconciliation Act of 1996 was a really significant reform of the welfare system as it had been known. What were the stated goals that that legislation wanted to achieve?

Schmidt [00:02:55] So the PRORA, which is what people call it based on the abbreviation, it’s a bit of a tongue twister, had four major goals that it was setting out to address. One was to provide assistance to needy families, one was to reduce welfare dependency among these families, and then there were goals related to family formation. So, one about reducing the prevalence of out-of-wedlock childbearing, and one about encouraging and supporting two-parent families. So those were the four stated goals of the welfare reform system. In general, I think a lot of the impetus for the welfare form came from increasing concerns about work incentives that people had about the AFDC program.

Siers-Poisson [00:03:43] It seems like that’s been an ongoing issue, trying to find that balance and differing opinions of where that balance is between supporting families who are in need of assistance and making sure that there are not disincentives to work.

Schmidt [00:03:59] Absolutely. Since the AFDC program was started in the 1930s, it always had pretty strong work disincentives built into the system. So even from the initial date, cash benefits would fall roughly for a dollar with every dollar of earnings, which provides very stark disincentives for earnings. I think when the program was created, that was actually considered to be a feature and not a flaw. The program was in the early period serving mostly widows with children at a time when the social expectations were that women with kids would be staying home. So, I think that it was originally set up to disincentivize work or to not require work among recipients. But demographics changed a lot over the decades that followed. And by the 70s and 80s, most of the recipients on the program were either never-married women or divorced women, and they were often viewed by the public and by some policymakers as less deserving of support than the widows of the past. And sort of coupled with this are changing norms about labor force participation for women in general, for moms in particular. And as a result, I think all of these changes and trends really set the stage for the welfare reform bill. By dismantling the traditional cash welfare system, replacing AFDC with Temporary Assistance for Needy Families, or TANF, which explicitly included things like lifetime limits on welfare payments and work requirements.

Siers-Poisson [00:05:35] And I want to dig into the conclusions that you reach in this research, looking at the welfare safety net since 1996. First, I want to make sure we all know who you were looking at in particular, because many different people, many different groups of people, are eligible for some type of safety net support at some point in their lives. What population did you look at specifically in this research?

Schmidt [00:05:58] So the bulk of the research is really focusing on a group of non-elderly, non-disabled, low-income families with children. That’s the group for which the safety net, I think, has changed the most over this time period. We do talk briefly in the paper about individuals without children. That tends to be a group that’s left behind and left out of most safety net programs. And then there are other programs that are targeting the elderly or people with disabilities. Those haven’t changed as much, so we’re really not discussing them in the paper.

Siers-Poisson [00:06:34] Let’s take a look at the major programs and how they were revised with welfare reform. And I’d like to start with cash or what you also call near-cash benefits. You did mention the move from Aid to Families with Dependent Children, AFDC, to Temporary Assistance for Needy Families, TANF. And what were the components of that transition? Who was eligible and how was it administered?

Schmidt [00:07:02] That’s a great question. So, AFDC really targeted families with children and primarily single-parent families with children. TANF also targets families with children. It has more flexibility for married parents, but it’s still aimed at that same core group of low-income families with children. I think the big changes which we already talked about were the work requirements and the time limits. But another big change was really how the program was funded. So, the AFDC program was provided as a matching grant from the federal government. So, the federal would contribute a share of every dollar that was spent by the state on the program. So what that meant is that if state spending went up, you know, the economy turns down, you have more people in need of benefits, the federal contribution would also rise accordingly. What PRORA did in part was restructure the funding stream to be what we would call a block grant so that states would receive instead of this matching amount, they would get a fixed amount from the federal government that would not change in response to changes in economic conditions or changes in the number of recipients. Another large change is that states gained a lot of flexibility to spend TANF funds on other programming besides cash transfers. So, the old system primarily spent money on cash assistance direct to families. TANF allowed states to spend money on other programing as long as it was loosely associated with those four goals of the welfare reform legislation that I mentioned earlier. And as a result, the share of TANF dollars that are spent on cash to families versus other spending types has fallen pretty dramatically. It used to be that about two thirds to three quarters of all TANF dollars were spent directly on cash assistance to families. Now it’s about a third or less actually goes in the form of cash to these needy families.

Siers-Poisson [00:09:11] And if I’m remembering correctly, TANF has become really one of the smaller safety net programs now, whereas AFDC, I think, was actually a much larger portion of the safety net.

Schmidt [00:09:24] That’s right. So immediately after welfare reform, the caseloads for TANF decreased pretty dramatically and they’ve continued to decrease. So, it used to be the case that for every 100 families below the poverty line with kids, about 68 of those families were receiving benefits from TANF in 1996. Now it’s less than 23 and there’s even more variation across states. In general, TANF has become a really small part of the safety net. Most low-income families with kids don’t receive cash transfers through TANF.

Siers-Poisson [00:10:03] Lucie, you mentioned the poverty line and I’m wondering, was it ever the intention that either AFDC or TANF would actually lift families over the poverty line or was it more just to minimize the impact of living in poverty?

Schmidt [00:10:20] So, I think it was never the intention that these benefits would bring families out of poverty. The benefits that were available to these families were not sufficient either before welfare reform or now to bring families out of property since welfare reform. However, the benefits have fallen. It’s becoming less generous. There’s a larger gap in many states between the benefit levels and the poverty line.

Siers-Poisson [00:10:46] Let’s move on to food assistance or, as it’s called now, SNAP. What does it mean that SNAP is a near-cash program?

Schmidt [00:10:55] Right, so SNAP provides these monthly benefits that families receive on an electronic benefit card, kind of like a debit card, that they can take to the grocery store and it can only be used to purchase food to prepare at home. So, when we say SNAP is near cash, really what we’re thinking of there is the idea that because the benefit amounts for SNAP are lower than typical grocery expenditures for most families, it will cover some of the grocery bill, but not all of the groceries bill. As a result, it kind of frees up income that can be spent on other purposes. And at least in theory, it doesn’t necessarily change the amount that people are spending on food because it is lower than their general food budget.

Siers-Poisson [00:11:40] So who is eligible for SNAP and what restrictions were put in place at that time for some recipients? And I do want to acknowledge that as we’re having this conversation in late May, there are significant changes being considered in Congress and the White House that relate to many of the topics we’re talking about today.

Schmidt [00:12:00] Like I mentioned earlier, TANF was really focused and is focused on families that have kids. SNAP is an interesting program because it’s available more broadly to most individuals with income below a particular level. So, there’s no restriction to families with kids. There are some additional restrictions on benefit receipt for certain groups. work requirements for what SNAP calls the ABAWDS, the Able-Bodied Adults Without Dependents, essentially working age individuals who are not disabled who don’t have kids. So, they face some additional restrictions. But SNAP is one of those programs that is more broadly available than some of the programs that are focused more narrowly on families with children.

Siers-Poisson [00:12:45] And has the way that benefit levels are determined changed over time?

Schmidt [00:12:51] It’s been pretty consistent. So, SNAP kind of has the feature of an old-school welfare system where the benefits are highest for people that have no income and then benefits are reduced with income. It’s about 30 cents on the dollar. So, as you earn an additional dollar, your SNAP benefits would go down by about 30 cents. And the maximum benefit that families would receive is determined annually based on the estimated cost of groceries that would provide sort of a healthy but low-cost diet to beneficiaries. And one thing that’s interesting about SNAP is that in general, the benefits are constant across the US, across the continental US. There are higher benefits in Alaska and Hawaii.

Siers-Poisson [00:13:36] So let’s shift over to safety net benefits that are administered through the tax system. And we talked earlier about there being this constant tension between providing support for families who need it and wanting to make sure that people who are able to work are incentivized to work. And I think the Earned Income Tax Credit, or the EITC, is a really interesting example of that. When did the EITC first become a way of supporting families, working families, that need some assistance?

Schmidt [00:14:12] That’s a great question. So, the EITC begins in 1975, and it’s really started in part as a response to some of the concerns about the work disincentives for the other safety net programs that we’ve already talked about. Because it has these work requirements associated with it, it has often received much more bipartisan support than other programs that are aimed at low-income individuals. And as a result, it’s been expanded several times through the years as a way of trying to, Clinton talked about making work pay and that was associated with an EITC expansion that was large enough so that families that were working full time at the minimum wage would earn enough with the Earned Income Tax Credit to pull them out of poverty.

Siers-Poisson [00:15:01] I think one thing that stands out in this is that families who do not have any income and families who are in probably the deepest poverty, the most challenging situations—it’s literally in the name: Earned Income Tax Credit. So, is there a group of people who are living in pretty severe poverty who are just not eligible?

Schmidt [00:15:24] Yes, the short answer, yes. I think, you know, one of the nice features about the EITC is that it does incentivize work. You’re not eligible for any benefits if you have no earnings. And actually, the credit increases with your earnings for people that have the lowest levels of earnings. We call that the phase-in range, where the credit gets larger with the amount that you earn. But it does mean that it’s kind of dependent on people’s ability to find a job that can limit how helpful it is in a recession, for instance. And there are definitely groups that have no earnings that are sort of falling, I think, through the cracks in terms of that piece of the social safety net.

Siers-Poisson [00:16:10] We’ve talked a little earlier in the conversation about some differences between states. Is there any flexibility at the state level to supplement the federal Earned Income Tax Credit?

Schmidt [00:16:23] Yes, so the EITC itself is federal. It works through the federal tax system administered by the IRS. But a large number of states at this point have actually started their own earned income tax credit. So currently, 31 states plus the District of Columbia have their own state-level EITC. And those are most commonly structured as a percent of the federal EITC. So often, if the federal EITC is expanded, the state EITCs will increase as well.

Siers-Poisson [00:16:54] Also administered through the tax system is the Child Tax Credit. How is that different from the Earned Income Tax Credit? For instance, who’s eligible and for what level of benefit?

Schmidt [00:17:05] So, the Child Tax Credit was introduced in 1997 as a way of providing more support through the tax system to help families with kids. It is also not available to families that don’t have earnings. So, in that way, it’s similar to the EITC. You have to have some earned income in your family to be eligible. But it also hits at a different place in the income distribution than the EITC does. So, it doesn’t phase out currently until your income hits $200,000 a year at the family level. So, it helps families that are much higher up in the income distribution than any of the other programs that we’ve talked about here. In 2024, it provides about $2,000 per child, but not all of that is refundable, which is a really important feature of tax credits when we’re thinking about this low-income population.

Siers-Poisson [00:18:01] So can you explain, for me and others, what that means, non-refundable versus partially refundable tax credit?

Schmidt [00:18:11] Yeah, absolutely. It’s a little tricky, but the idea here is a refundable tax credit is a credit that you can get as a refund, even if you don’t owe any tax. So, when we think about tax credits, what they tend to do is you subtract a tax credit from the amount you owe the government, the amount of tax liability that you have. So, most tax credits can bring your tax liability down to zero, but not below that. A refundable credit, means that if your credits exceed the amount of your tax liability, you can get the excess refunded to you as a tax credit. So, for low-income families, because of our progressive tax code, most of those families don’t owe taxes, they don’t have a federal tax liability. So, a non-refundable tax credit will give them no benefit whatsoever. By making them refundable, we’re allowing them to sort of benefit from that extra amount. And the way that it’ll work in practice is they’ll get that as a tax refund once a year after they file their taxes, usually in March or April.

Siers-Poisson [00:19:19] We saw a very significant increase in the Child Tax Credit during the pandemic. How did we see that benefiting families?

Schmidt [00:19:29] So, the CTC played a major role in some of the pandemic stimulus relief. So, in 2021, the federal government substantially expanded the credit. So, it went from being about $2,000 per child to between $3,000 or $3,600, depending on the age of the child. The pandemic expansion made the entire credit available to families that didn’t have earnings for the first time. So, it helped that vulnerable group with zero earnings that we were talking about earlier. And then rather than pay it all as a lump sum, it directly paid some of it to families monthly and most families didn’t have to apply for it. They were receiving it directly without any application procedures just based on information that the IRS had. So, these changes really significantly cut child poverty in the latter half of 2021. The expansion was really only in place for six months and then it expired, but over that period child poverty rates decreased pretty dramatically.

Siers-Poisson [00:20:34] A huge part of safety net coverage and spending is healthcare programs, in particular, Medicaid. Can you give us a brief history of what healthcare coverage for disadvantaged people looked like pre-1996.

Schmidt [00:20:50] Sure. So public health insurance really dates back to the War on Poverty in 1965. That’s when Medicaid was created. At the time, it was really limited to recipients of cash assistance programs like AFDC. Beginning in the 1980s, there were a series of expansions to make other vulnerable groups eligible at higher income categories, people like pregnant women, infants, and young children. Welfare reform in 1996 delinked Medicaid from AFTC or from cash transfers, but really retained in general these categorical requirements. The big change with that came in 2010 when the Affordable Care Act was passed. So, the Affordable Care Act had Medicaid as kind of a centerpiece of its efforts to reduce on insurance among Americans. It was meant to make Medicaid available to anyone who had income below 138 percent of the federal poverty line. So that would have provided this benefit to people outside of the categories that had traditionally been eligible for Medicaid. But in 2012, a Supreme Court decision made that Medicaid expansion optional to the states. And initially, only about half of the states took up the Medicaid expansion, and it largely fell along the lines that you might expect. The states that took up Medicaid expansion tended to be blue states, red states tended to not take up the Medicaid expansion.

Siers-Poisson [00:22:18] Today, how would you position Medicaid within the larger social safety net?

Schmidt [00:22:25] So, it’s definitely the largest and the most expensive of the safety net programs at this particular point in time, right? So, in 1996, the amount we were spending on Medicaid for these populations was roughly similar to what we were on the earned income tax credit and AFDC. And since then, I think expenditures for this particular population have tripled. So that’s partly due to some of these expansions that I was talking about. It’s partly just due to the fact that health care costs are increasing much more rapidly than other types of spending. I think there are a couple of things that are particularly interesting about Medicaid. I think one is that with the Medicaid expansion, it does become one of those programs that affects not just people with children, but people without kids who might not have access to other safety net programs. So, if you are in a state that took up the Medicaid expansion, all you need is for your income to be below that threshold to be eligible. So, it does serve kind of those groups that are otherwise excluded from the safety net. So, I think that’s one piece. I think the other piece that’s really interesting about the Medicaid expansion, and this is not something that you asked, but something I’ve been thinking about a lot is that it’s gotten more popular since it was passed, since the Affordable Care Act was passed. So, I said initially half the states did not take up the Medicaid expansion. But a number of states have been sort of gradually adding the Medicaid expansion as we sort of move forward in time. So now there’s only 10 states that have not expanded Medicaid. The additional states that have taken up the Medicaid expansion are not simply blue states. They tend to be more diverse politically. The most recent state that took it up was North Carolina in December of 2023. So, I think it is really interesting that among both policymakers and individuals in many of these states, it’s become more popular over time.

Siers-Poisson [00:24:27] I want to look at the conclusions you come to in your paper. One is that the safety net has actually expanded in generosity since 1996, which I think might be a surprise to some people.

Schmidt [00:24:40] Absolutely, I think there was a real concern in 1996 that welfare reform would kill the safety net. It was politically unpopular, the welfare reform bill, among many of Clinton’s supporters. He had high-level officials in his administration resign in protest as a result of the welfare reform legislation. So, I think people were concerned that the safety net would disappear, that this would have large implications for the well-being of low-income families. As it turned out, the concerns about, the cash piece of that, the monthly cash transfers, I think were well founded, right? TANF is not an important part of the safety net anymore, but the expansions in these federal programs, SNAP, Medicaid, and the tax credits have really more than offset those post-1996 losses and cash transfers.

Siers-Poisson [00:25:32] You also say that that move to more federal programs means less variability between states. And I found this really interesting: Also, an ability to respond more quickly in times of crisis.

Schmidt [00:25:45] So, I think when welfare reform was passed, there was this real concern that by giving more power to the states to determine welfare policy, that we would see this race to the bottom, right? That states and in particular states that have historically not been particularly generous to low-income individuals would cut their safety net benefits and we would see sort of widening gaps between the generous states and the less generous states. That did happen with respect to cash, but because the federal programs grew by so much, that really reduced the variability across states in terms of the total package that’s available to low-income individuals, and in fact in all states the total package of cash plus newer cash benefits increased. So, I think that piece of it is really, really important. I think the other piece of this, the fact that we were expanding these federal programs in the years following welfare reform, I think did mean that the federal government was well positioned to use those programs to help stabilize the economy during the pandemic recession. So, the safety net response to COVID included things like big increases in SNAP. We spent a lot of time trying to put policies in place that could increase SNAP participation. And so, we were well positioned to use SNAP as part of the pandemic response. Similarly, the Child Tax Credit was a big tool. Medicaid, there were continuous coverage policies that meant that people didn’t lose Medicaid coverage and that helped them maintain that during the pandemic. So, we think that these federal expansions of the safety net in those programs, really facilitated the very rapid and dramatic policy response we saw during the COVID recession.

Siers-Poisson [00:27:39] We’ve certainly covered this already, but just to kind of tie it up, it seems like there was a real shift in focus to supporting the working poor and near poor families more than people who were completely without income and rewarding modest earnings more than supporting people with no earnings at all.

Schmidt [00:28:01] Absolutely. So, I think that the combination of work requirements in TANF, the reduction in importance of TANF, the importance of the tax credits, all of those meant that the big expansions and the benefits that we saw really hit people that were sort of working poor or working near poor, they were much more impactful for those particular groups. I think we show in the paper that the largest increase in benefits really went to people with incomes between $15,000 and $45,000. We show in the paper that all groups with kids saw increases in safety net benefits that were available to them except for people with zero earnings, and that group is worse off as a result.

Siers-Poisson [00:28:48] So going back to the title of your paper, bottom line, did welfare end as we knew it in 1996?

Schmidt [00:28:56] Our take is that the welfare reform legislation essentially kept President Clinton’s campaign promise. It did end welfare as we know it, but it didn’t kill welfare, it just changed it. So since then, the safety net has both expanded and changed character, moving away from these monthly cash transfers towards these in-kind supports and towards refundable tax credits. And with the same distributional implications that we talked about, sort of this shift away from those with no earnings towards those who had modest levels of earnings. I think a couple of other groups that were sort of affected differentially, there were more supports after welfare reform to married parents, fewer supports potentially to single parents, especially to the extent that they had no earnings. I think one group that’s fully been left behind are people without kids. So, they’re eligible for some Medicaid if their state took up the expansion, some SNAP, but with increased requirements, a very small EITC, but that group really does not receive much help from the social safety net.

Siers-Poisson [00:30:02] As we wrap up, is there any further research you’d like to do or see done on this topic?

Schmidt [00:30:08] So, there’s a lot of work that’s been done and that is being done on the effects of different social safety net programs on well-being of low-income families, of kids, of parents. I think much of that literature shows that all of these safety net programs that I’ve talked about have positive impacts on outcomes that we might care about, like health, education, self-sufficiency. I think many of those research papers focus on one program at a time, and I’ve been really increasingly interested in how the programs work together as a package. So, I think one of the things I’m interested in continuing to work on are papers that look at the effect of the total safety net package on a number of different outcomes. I think we will need research to examine the effects of any changes that occur to the social safety net in the current era. And I also, I think one thing that I’ve been thinking about a lot with respect to future research is the role of data. So, most of the data that we used in this paper is collected by government agencies. I worry that some of the cuts that we’ve seen to the government over the past few months are going to weaken the infrastructure that’s needed for collecting data and therefore threaten our ability to examine the effects of the safety net on outcomes and on participation moving forward.

Siers-Poisson [00:31:36] Well, Lucie, thank you so much for taking the time to discuss your research with us.

Schmidt [00:31:40] Thanks for having me, it was very fun.

Siers-Poisson [00:31:43] Thanks so much to Dr. Lucie Schmidt for joining us to discuss her recent paper titled, “Did Welfare Reform End the Safety Net as We Knew It? The Record Since 1996.” You can find a link to that paper in the program note for this episode. Please note that views expressed by our speakers don’t necessarily represent the opinions or policies of the Institute for Research on Poverty or of any other sponsor, including the University of Wisconsin–Madison. Music for the episode is by Poi Dog Pondering. Thanks for listening.