Day Manoli Transcript — Oct. 7, 2025
Kevin Johnson/NPF (00:00:00):
Our first session is on policy and what makes good policy from taxes to social security, workforce development, healthcare education, the environment. Washington, as many of you know, is a wash in federal policy. But what is good policy and how should we be evaluating measures as they impact local communities? Our first speaker is here to help us answer those difficult questions. Day manoli is an associate professor in the McCourt School of Public Policy at Georgetown University, an interim faculty director of McCourt’s Data Science for Public Policy Program. Dave’s research focuses on empirical analysis to document and improve the impacts of government policy. His research interests include social security, retirement policy, income tax policy, and education policy, probably the waterfront of everything that you guys are dealing with in your communities. So please give a welcome to Professor Manoli.
Day Manoli/Georgetown University (00:01:15):
First, I just want to say thank you guys very much. Thank you very much to the organizers. Really excited to be here. So I’m a professor in the McCourt School of Public Policy. I’m an economist by training, so I just wanted to kind of start off by giving you guys a little bit of background about me and then we’ll kind of dive into some of topics that were mentioned. But I’m originally from Milwaukee, Wisconsin, so I grew up reading the Milwaukee Journal. It was very important to me because that’s when I got to actually know more about Milwaukee and reading about places near my neighborhood or around the city. So local journalism was very much close to my heart, and I totally understand that there are changes in local journalism. The Milwaukee Journal has since combined with the Milwaukee Sentinel become the Milwaukee Journal Sentinel, and now I’m much more likely to see New York Times articles in the Milwaukee Journal as opposed to actual local reporting.
(00:02:11):
So I understand things have definitely changed, but I went to graduate school at Berkeley and I did my PhD in economics at uc, Berkeley, and I think that that’s kind of when I really started to appreciate the importance of local context. I was an economics PhD student and I needed to do my tax return, and I realized that I didn’t know how to do my tax return. So I was looking for some help on how to do this, and I found a volunteer income tax assistance site at a community organization in San Francisco, the women’s building, and I started volunteering there and learned more about sort of local, the mission district in San Francisco. I was able to learn how to do my taxes, but I also got to know a little bit more about the community in San Francisco. And so I think that local knowledge really still informs a lot of my research about tax policy, about social safety net programs and thinking about economic opportunity and economic security more broadly.
(00:03:11):
So that’s a little bit about my background, but now let me get into a little bit more on the policy discussion and some of the topics that were kind of highlighted in the introduction. I’ll just mention at the out front or at the start, please jump in with any questions. I like to have much more of a conversational kind of interaction as opposed to me kind droning on. And yeah, it is better for everybody to hear other voices rather than just my own. So please jump in anytime. So we’re going to talk a little bit about good policy, then we’ll get into current policy and what we know from existing evidence, and I’ll try to use that to set up thoughts on how to think about policy changes and evaluate evidence going forward. So to start on, what is good policy? I just want to highlight that I’m not here to tell you what is good policy.
(00:04:05):
I think that there is a lot of debate about what political preferences are, what individual preferences are, what market failures exist. There’s a lot of discussion around what is a good policy and there are a lot of value judgements that come to that. I think that maybe from a very narrow perspective, I can speak to this question from an economist’s perspective. So in public finance, public finance is the field within economics that generally focuses on analyzing government policy and regulation. We generally focus on redistribution and efficiency as some of the core topics, social insurance, redistribution, these are kind of the two or the dichotomy within public finance. And when we think about redistribution, we could think about this as a very simple starting point that we just give cash to individuals if we’re trying to redistribute to them. I think that has a lot of the flavor of some of the policies.
(00:05:06):
You guys might be familiar with universal basic incomes, and especially coming out of the A RP or coming out of the recent Biden administration, there was just, let’s think about giving cash to individuals. Unconditional cash transfers. We achieve redistribution mechanically, right? That’s kind of by definition we’re giving cash to individuals, but this might create distortions, it might change people’s incentives that lead to inefficient behavior. So if we give individuals income or other benefits, this might create an incentive not to become employed or not to participate in the labor force. This might create some source of inefficiency. I’m just going to start here at the outset. That inefficiency is relative to what would happen if we didn’t have to change people’s incentives or change people’s marginal decisions. What would just happen if we gave them the cash and didn’t mess with how much they would make in the labor market or how much their marginal returns to effort would be?
(00:06:12):
Well, we can think about taking a step away from cash and say, well, okay, so given that cash might create some behavioral distortions or some potential inefficiencies, maybe we go the route of thinking of in-kind transfers. So this kind of starts to think about motivations along the lines of paternalism. Individual like voters might be okay giving individuals contingent cash transfers or particular goods. When we think about in-kind transfers, we have a whole variety of in-kind transfers in our society. So if we give individuals healthcare, that’s Medicaid. So we provide health insurance for lower income individuals. We don’t necessarily provide the cash, but we provide the healthcare. Similarly with food, we might feel okay giving individuals food as opposed to giving them cash. That’s basically what SNAP does. So we say, we’re going to give you income that you can spend specifically on food and maybe even on specific types of food.
(00:07:15):
We might not say that you can purchase alcohol with this, but you can purchase other types of food. We also say we are relatively paternalistic or we might value employment, so I might not give individuals unconditional cash transfer, but I might say, if you work, then I’ll give you this subsidy. So we have things like the earned income tax credit, we have job subsidies where we pay employers to try to get individuals hired, and then we also have child transfer. So if we think of benefits as being targeted towards children, maybe we provide direct transfers to children, things like public schooling, public daycares, or as we’ll see coming out of recent legislation, child savings accounts, these are sort of specifically tied to children as opposed to the adults that might be taking care of them. All of these might allow us to maybe get some redistribution happening more efficiently or in a way that we find more palatable than just by giving cash transfers.
(00:08:20):
But we also have to consider the administrative costs, and I think that’s what a lot of recent legislation has focused on. So I’ll come back to think about the recent legislation, but I want to start to put this in historical context and also in the context of, well, what do we know about changes in the social safety net and how those have impacted individuals. So I think probably the main comparison that I would highlight is thinking about 1990s welfare reform. The big beautiful bill this past 2025 really created transformative changes in our social safety net thinking about changes in Medicaid and SNAP that are coming up. These are very transformative reductions in funding for these programs. I think the maybe closest historical comparison would be thinking about welfare reform in the 1990s. So this was a time when we saw large transitions away from A FDC and towards what is the current TANF program.
(00:09:23):
So A FDC was generally operating in the 1980s to 1990s here. That was the assistance for families with dependent children. This was kind of creating, I think what was often referred to as the welfare state that would provide cash assistance to individuals that were out of the labor force if they had children. As individuals increased their earnings, the benefits were reduced that created sort of this incentive not to increase earnings to avoid the benefit cliff and maybe stay out of the labor force. Well, in welfare reform in the 1990s, what we saw was a reduction in A FDC grants and a shift more towards an employment-based social safety net. So that’s when we saw very large trends, large expansions of the earned income tax credit. And so when we think about single women with children here, so we have single women with one, two, and three or more children in these orange, blue, and black lines.
(00:10:22):
We saw very transformative increases in labor force participation for those individuals affected by welfare reform and the EITC expansions for single women without children. These were individuals perhaps not affected by the program changes. We didn’t really see any impacts on their labor force participation. To put this in context, this is one of the largest labor force transitions of a demographic group in the last 30 years. So this is a really transformative change where we saw large reductions in welfare and significant increases in labor force participation. Okay. Okay. So that’s kind of thinking of evidence from the 1990s. I don’t have too much in nostalgia for the 1990s, although I think that that seems to be pretty significant in pop culture space, but it is what, 35 or 30? It’s a long time ago. So what other evidence can we speak to that might be more recent?
(00:11:20):
Well, I also want to highlight another recent analysis of cut and benefits, and this was coming from Tennessee. So in Tennessee in 2005, beginning in July, 2005, there was a dis-enrollment of roughly 100,000 individuals from Medicaid. So there was a large change in eligibility. So the State Tennessee Medicaid program dis-enrolled these individuals. There was a law passed that changed eligibility requirements on January 1st, took effect in July. We saw large reductions in the coverage of Medicaid for individuals in Tennessee. So you can see this line here is looking at the fraction of individuals with Medicaid coverage in Tennessee. That’s the solid line relative to other southern states or other neighboring states. That’s the stash line here. You can see about 20% of individuals in Tennessee were on Medicaid prior to this change, and then a very significant drop down to about 11 or 12% within this six months of implementing the policy change.
(00:12:29):
This creates a pretty useful opportunity for research to kind of think about what happens when individuals are taken off of Medicaid, what happens to their insurance coverage and what happens to their employment? Well, when we think about the outcome of UN-insurance, this is looking at the fraction of individuals who are uninsured. Again, the solid line here is Tennessee, the other southern states is the ash line here. At the same time that we saw this large decrease in Medicaid coverage, we saw a larger increase in the fraction of individuals that were without insurance. Okay, so when we think about this question of if you lose Medicaid, are you more likely to find other insurance options? We didn’t see that so much in the case of Tennessee in 2005. I want to highlight 2005. This is maybe a different policy environment than we currently have. This is pre a CA, right?
(00:13:31):
So maybe things would be slightly different, but at least from what we can see from historical evidence, we didn’t see large increases in insurance coverage in Tennessee during this experience. What happened with employment? So one idea might be that maybe some individuals lost health insurance, maybe they found a job that provided health insurance benefits. My students kind of tell me the job market is pretty tough these days. I think that’s understandable, probably an understatement. It’s not exactly easy to go out and find a job that has health insurance benefits or has exactly what individuals are looking for. And so when we see this period in 2005 or mid 2005 where we saw a large reduction in Medicaid coverage, we don’t really see large increases in full-time employment for individuals. So it doesn’t seem that individuals were able to find employer sponsored health insurance.
(00:14:32):
At the same time, we see some increases in part-time employment. Oftentimes part-time employment might not offer employer sponsored health insurance, but it offers cash or it offers income. So individuals might’ve been able to make up for some of the out-of-pocket costs that they were experiencing by not having Medicaid coverage, by taking on a part-time job during that time of the loss of coverage. All of this is to say that, okay, well, we can see that we can learn something potentially from recent experiences, but again, this is about 20 years old. So we can think about other recent evidence.
(00:15:13):
One piece of evidence here is coming from the Oregon Health Insurance experiment. So the health insurance experiment was literally a randomized experiment. So Oregon, the health department randomly selected some individuals and offered them health insurance. So their program was oversubscribed. They could not provide health insurance to all of the individuals that applied, but they randomly selected who they could offer to. They offered, I think, roughly 10,000 slots to individuals for public health insurance coverage. The question I think that many researchers examined was, well, what was the impact of getting access to Medicaid? One of the first outcomes that I think emerged pretty clearly was the impact of having Medicaid coverage on out of pocket medical expenditures. And so when we’re looking at this graph here, we’re looking at the distribution of out-of-pocket medical expenditures. And so you can see here, so this is roughly 60% of individuals had about a hundred dollars or lower of out-of-pocket medical expenses.
(00:16:26):
So thinking about paying for insurance drugs that were not, sorry, prescribed drugs that were not covered by Medicaid, for example, these individuals might’ve had to pay for that out of pocket. That’s looking at the control group for the treatment group. That’s the red dash line here. We saw fewer out of pocket medical expenses. Individuals were able to use the Medicaid coverage to cover prescription drug costs. That was probably the most immediate impact of receiving health insurance. So individuals out-of-pocket costs decrease. Some individuals have very significant out-of-pocket costs reaching up to about $1,500, and we saw reductions. Some individuals had some reductions in the treatment group when they had access to Medicaid, but those are also kind of situations where individuals might have catastrophic situations and might have very significant costs overall. But for the most part, a lot of the reduction in out-of-pocket costs came in this lower dollar of $500 or less.
(00:17:35):
In terms of having out-of-pocket expenses, that might seem relatively small in terms of an absolute amount. When you’re thinking about individuals on Medicaid, it’s really important to understand the income context. And so you’re thinking about individuals that have 10, $10,000 of income annually. $500 out of pocket is very significant. So the percentage of income, this was a pretty large reduction in the percent of income that was dedicated to out-of-pocket or unanticipated medical expenses. I’m going a little bit over these examples somewhat quickly, but let me pause and see. So any questions so far or any thoughts, comments so far?
Simmerdeep Kaur | Walla Walla Union-Bulletin (00:18:14):
Hi, my name is Simmerdeep. I’m a federal policy reporter with the Walla Walla Union Bulletin. I was wondering if you have any insights about if there were any consequences of the Medicaid disenrollment on hospitals in Tennessee? Was there an increased strain on the emergency departments?
Day Manoli/Georgetown University (00:18:30):
Great. So thinking about this example, so the question was kind of thinking about, well, what did hospitals or health insurance provider or healthcare providers experience when they saw this decrease in Medicaid coverage? So overall, there were significant increases in emergency room visits, so we didn’t see too many other changes in hospital admissions or emergency room usage or other healthcare utilization. And so I think that this speaks to health insurance coverage. When it’s exercised, it’s exercised with coverage for people that might be going to the emergency room. Those health events might just be occurring at random times, whether you have health insurance or whether you don’t, when you lose Medicaid coverage, that might not be effective. The people going to the emergency room, that just is pretty constant. But what does change, I think, are the sort of preventative treatments. And so maybe taking, if you think about medical adherence, taking your prescription drugs for things like depression, blood pressure, those kinds of pieces, that’s what we’ve seen from subsequent research. But we didn’t see any changes in emergency room visits, particularly in the Tennessee example.
Whitney McKnight | The Edge (00:19:50):
Hi, this is Whitney McKnight from the Edge in Kentucky. This is maybe a meta question and you may get to it later, but I’m watching these studies and I’m seeing the data makes me wonder if legislators are looking at data before they make their decisions legislatively, and if they are actually looking at data, what data are they looking at?
Day Manoli/Georgetown University (00:20:13):
Great question. So I often talk to in my class about evidence-based policymaking, and one of the most frequent questions I get when I use that phrase is What’s the alternative? And I think that the alternative is I think people going with preconceived notions or what I thought was evidence, but I never questioned. And so I think that when I have, in the experiences that I have had, I think more often than not, legislatures or policymakers might often be operating under a preconceived notion that might be informed by their own particular experiences or some reading of the evidence. But I think that when I have had opportunities to speak to them and present studies, my own research, others, it has sort of I think been somewhat eye-opening to see their reactions, which is also eyeopening. And so it does seem to be that communicating the evidence from research or from evaluation studies to policymakers, there seems to be some barrier, some market failure there. Yeah,
Whitney McKnight | The Edge (00:21:23):
Actually there were congressional research staff.
Day Manoli/Georgetown University (00:21:26):
Yes,
Whitney McKnight | The Edge (00:21:26):
I don’t think they exist anymore. Or if they do, they’re much smaller in number. Okay, here we go. Is that coming from, I mean, I have an unfair advantage. I used to be a Washington reporter, so I know that a lot of the lobbying groups would be who gave you the research and that’s what they use. So is it possible that this research that we’re basing legislation on now is coming from insurance companies or something like that?
Day Manoli/Georgetown University (00:21:51):
So I do think that there is a lot of research, and I’ll talk about this towards in the second half or maybe the last couple slides in particular, I think there is a lot of research that can drown out solid evidence. And so when you think about nonprofits or politically funded nonprofits in particular, I think that you might get a biased view, I would say, or an agenda driven view of evidence as opposed to just looking at objectively what happened. And so I think that I kind of put that as a market failure, that as a professor, as someone working on producing research, producing evidence, I have to do a better job of communicating that to policy makers, making sure that the assumptions that I’m making, the conclusions that I’m drawing, that that is accessible to individuals and that it’s not drowned out by other noise.
(00:22:50):
So yeah, I do think that there’s sort of more of that out there than there used to be. But trying to get this to congressional staff, trying to get this to committee staff, I think is really critical so that they can see what is the evidence on work requirements, what do we know, what do we not know? Totally agree. I totally understand. And I guess I view myself or my role and maybe the role of us here that we can bring, we lead a horse to water. It’s up to the horse to drink. So that’s the goal that I would say here.
Lauren Gibbons | Bridge Michigan (00:23:30):
Other questions, comments? I have a question over here. I just figured I’d take the mic. I’m Lauren Gibbons. I’m a reporter with Bridge Michigan. We’re a nonprofit newsroom in Michigan. And when I’ve covered Medicaid specifically, the biggest thing that comes up is a lot of concerns from rural hospitals. And as that uninsured number goes up, as we’ve seen in examples from other states, they’re very concerned about closures. But as we were looking for data for how many hospitals could be affected, it was kind of dicey. There were a lot of different estimates. So I’m curious if in your research you’ve covered the rural hospitals issue and what you’d recommend for the best data sources.
Day Manoli/Georgetown University (00:24:12):
Yeah. Great. Okay. So definitely I’ll get into thinking about data sources and data-driven journalism. I think, I’m sure you guys are more familiar with this than I am, but one caveat that I want to say is that especially when you start thinking about underserved communities like rural groups or rural locations, native American reservations, particular groups, oftentimes the large aggregate data sources might not cover enough of those areas. And so I think that’s where some of the qualitative evidence really elevating surveys, focus groups in those areas can just incredibly compelling. I’ve done a lot of work on thinking about tax filing assistance and volunteer income tax programs. Think about the non-filing population. That’s a lot. It’s a relatively hard to find population, but when you kind of focus on those rural communities or when you focus on tribal populations or minority communities that are not English speaking, really elevating what is going on there I think is very valuable.
(00:25:18):
The data, I would say that if you are looking at an underserved community, you’re setting yourself up for that. You’re slicing the data really small, so you’re going to need a large data set to start with. And so then I would say starting with government administrative data sets can be extremely valuable. So partnering with social service agencies or human service agencies like your snap, tanf, Medicaid offices, those are the individuals that they touch in those areas and they will have a large enough population to be able to work with. So that’s kind of where I’ve done a lot of work thinking about working with state and local governments, but I would definitely recommend those as data sources. Yeah.
Micaela Watts | The Institute for Public Service Reporting (00:25:59):
Hi, I am Micaela Watts. I’m from Memphis, Tennessee, and I can tell you firsthand what the impact has been of denying Medicaid expansion as well as a a expansion. And I can tell you that the data has not kept up with the devastation that we see, especially among our rural communities. All types of preventative healthcare services have been decimated, especially prenatal care among minority women. The bottom line is we are a GOP heavy state, and these lawmakers absolutely hate anything that ever touched Obama. They hate anything that carries a price tag and they act against the interest of just about everybody in Tennessee. So my question is, I mean, we can look at this data all day and we could look at these contexts and examples, but there’s a serious ideological interference. Where does that come in this discussion?
Day Manoli/Georgetown University (00:26:54):
Okay, so two things. One, coming back to the data question, I think that that’s right, that I can show you data on individual outcomes. I think that the sort of full picture is much larger than any set of outcomes is going to be capturing. And so I think that the, I’ll talk about the importance of storytelling later, but I really think that capturing that full effect is really, I need to know the stories of what is going on to really capture everything that folks are experiencing, the political angle. I think that when I’ve gone down the route of just understanding how policy is working and what are the impacts of individuals, I think there’s a way to just tell that in a factual manner. And then getting to the normative side of what are the sort of solution strategies, what are the sort of pieces to speak to those shortcomings or those impacts of policy? I think that’s sort of untethered to ideology. It’s really like, okay, so when you understand what is going on, then you can see where in the pipeline are things breaking down and what can be addressed.
(00:28:06):
I understand, I don’t mean to be naive about the ideological bent or reading of the evidence, but in my experience, I think that when you get into the weeds of what is actually going on, the sort of pathway of like, well, let’s test what strategies work that has been somewhat untethered to ideology. And you can say, let’s test whether the EITC works or providing employment incentives work. Let’s test whether if we provide transportation subsidies, we get individuals to jobs. So if we understand the context of what is going on, why we are seeing individuals out of the labor force, and we can build in ideas of, yeah, let’s test that. And that I think has been people are like, oh yeah, that’s almost just this curiosity of that seems like a good thing to test. Let’s try it. But I don’t mean to be naive. I do think that there are political angles that put a lot of constraints potentially around these. Yeah. Lemme take one more and then I’ll continue.
Sean Keenan | Atlanta Civic Circle (00:29:19):
Thank you. My name’s Sean Keenan. I cover housing for a little nonprofit called Atlanta Civic Circle. I cover other subjects for other publications. I heard an interesting misunderstanding of an idiom the other day. You talked about leading a horse to water the data, the evidence that supports good policy. This person said you could lead a horse to water. You can also drown the horse. And in Atlanta and the state of Georgia, our governments kind of have a bad habit of studying things to death when there seem to be common sense solutions. They will instead create a study committee, and then that study committee will take a year or two years and then say with the housing policies, we have pretty common sense legislation that would confront our scourge of corporate landlords dominating our single family housing markets, but we’re just going to keep studying it. And so I wonder if there’s a sweet spot to presenting lawmakers with sound data that is digestible but doesn’t drown them.
Day Manoli/Georgetown University (00:30:29):
I think that’s right, and I would group that in some of the discussion earlier here about when presenting this to legislators or policymakers thinking about actionable evidence as opposed to, okay, here’s just evidence. Okay, well, what’s something concrete and simple or transparent that we can do based on this evidence? So very concretely in some of the work on thinking of some of the changes to SNAP that are being debated, we’ve seen a lot of SNAP has employment and training programs. Participation rates seem to be extremely low in the single digits of percent in terms of the percent of SNAP beneficiaries who are participating in employment and training. I think a natural question there is why in thinking about some of the caseworkers assessments like childcare and transportation assistance are some of the key barriers that beneficiaries have identified. Some legislatures and policy makers have said, okay, let’s try to provide them with rideshare vouchers or let’s try to provide them with metro cards and test that.
(00:31:43):
And that has led to actually, we have some evidence now of increases in employment when you provide transportation assistance. Another one I think in the housing context is well suppose that you provide what’s called shallow rent subsidies. So individuals that fall behind on rent, they might just be coming up a couple hundred dollars short each month, but if that accumulates, that can become very daunting. And so maybe you can just provide a couple hundred dollars or like $300 a month for four months, and maybe that’s enough to help them get up back on their feet and meet their rent payments. So I think that has been a project that we have tested in San Diego County, and I think other cities are also looking into this now based on some of that evidence. But I think some of this evidence, you can get into the weeds pretty quickly and you can kind of get lost in the weeds, I think. But thinking about keeping it digestible and transparent in terms of potential solutions, I think is really critical.
(00:32:46):
I guess some of that is on the part of people producing evidence, but maybe that last mile of communicating that to policymakers in a transparent actionable way is really critical. And coming back to this other point about political ideology, I think a lot of problems or a lot of inefficiencies in social safety, net programs, tax policies, these are relatively low hanging fruit that I think it’s almost like most people would kind of call like, oh yeah, this makes sense. Let’s just do this. It’s not really. Maybe there’s a larger sphere where things are sort of more political, but that the sort of bottom 5% of low hanging fruit easy solutions, I think most people would agree on. Yeah. Okay, great. Well, thank you guys for the questions, and so I’ll definitely pause again, but please feel free to jump in. The other one that I want to speak to, the other example that I want to speak to is thinking about Medicaid work requirements.
(00:33:45):
This has been in the news a lot recently, but the basic idea is to say that if you are receiving Medicaid and you are able to work, then you should be working. And so to enforce this, we’re going to have assessments to determine work capability and then impose work requirements for individuals. And so this was, Arkansas was one of the earliest states to test this in 2018. So beginning in 2018, for the first half of 2018, they imposed a work requirement on Medicaid recipients, and then a judge struck down the program saying that it was not, it wasn’t within the law to impose a work requirement on Medicaid. The program would have to be changed through legislation. So then the work requirement disappeared. So again, that creates a pretty clean, natural experiment to think about what happened to those individuals that were subject to the work requirement in Arkansas relative to other neighboring states before and after the policy change.
(00:34:49):
So here we’re setting our benchmark to 2016 in Arkansas when there was no work requirement. So we set that to zero here. Then we’re looking at what was the percentage change in Medicaid enrollment or coverage once the work requirement took effect. And we saw a decrease in Medicaid coverage of about 13%. So about 13% of individuals, one subject to the work requirement, were no longer eligible for Medicaid. They didn’t meet the work requirement, so they were not eligible for the program. What happened to these 13% of individuals? Well, we saw an increase of about 7% in the uninsured rate in Arkansas, again, relative to other states. So some of these individuals, or about half of them that were no longer on Medicaid were no longer covered by any insurance.
(00:35:47):
Other individuals, we saw an increase in employment, so the remaining 50, so about half of individuals went uninsured, the other half of individuals went on to get employer sponsored health insurance. So individuals were able to find some employer coverage, but not everybody. So roughly half of individuals went without health insurance coverage. Well, was that employer sponsored covered or employer sponsored health insurance coming from increases in employment? Well, if we look at employment, we didn’t really see any changes. So this is again, looking at changes in employment, individuals working more than 20 hours per week in Arkansas after the policy change relative to other states. And you can see we don’t really find much of an effect. So zero here is kind of in this middle of the estimates. So we didn’t really find any clear evidence of a distinct increase in the share that was employed.
(00:36:50):
We don’t find significant increases in hours worked either. So a lot of individuals that were on Medicaid, it seems that if they were taken off of Medicaid, if they had employer sponsored health insurance, then they could cover themselves through employer sponsored health insurance. But once individuals that were subject to the work requirements, if they were not already employed, they were not able to get health insurance. So the uninsured rate of that population went up. Again, it seems like somewhat intuitive that I think this is kind of like the intuition that emerges from caseworkers that why don’t requirements seem to work well? Most individuals that are able to work are already working. Remaining individuals are kind of selected on not being able to work. These are the individuals that might have other health constraints, other barriers to work. And so when you impose a work requirement on that remaining population, that’s the population that’s going to drop out of the program.
(00:37:52):
And then they already had some barriers to work initially, and so then not surprisingly, we don’t really see an increase in employment for that group. So essentially the work requirement does not bind on individuals, individuals that don’t have any barriers to work. Okay. The last piece that I kind of want to think about in terms of where we are with evidence is the following implementation really matters a lot for all of these programs. All of this evidence that I’ve shown you is really about the program impact. So when you take individuals off of Medicaid, what happens to employment? What happens to emergency room visits? What happens to health insurance coverage? That requires a lot of implementation. There are frontline workers that are going through application processing pieces that are going through income verification pieces, that are going through recertification pieces. All of that is really, really critical.
(00:38:56):
When we think about these pieces with the big beautiful bill, there are tremendous changes in work requirements and recertifications going on. In many other cases, when we have had snap TANF Medicaid changes, there’s been a tranche of funding that has gone on for administrative costs here. There is no cost for the changes that are being implemented. This is not that states are receiving additional funds to cover program changes or application system changes or getting new income verification systems up and running. So the implementation here is going to be first order. It’s true at the state level. It’s also going to be true at the county level. Many SNAP programs are state, some are state administered, some are county administered. So have, so we refer to this as fiscal federalism, but at all levels of government, there are significant changes going on for Medicaid and for snap.
(00:39:55):
So we need to know how are states actually going to implement work requirements? How are they going to implement recertifications? What does that mean? Well, how do you have a case worker recertify somebody every six months? A lot of times that involves somebody getting on the phone calling their employer or calling the individual and asking, were you working at this firm? Do you have any documentation to verify that you were working at this firm? Reaching out to HR departments, calling the Walmart HR department to say, is this person employed here? Is this person employed at Target? That requires a significant amount of time. If those individuals are not able to process that in time, what happens? I think that that is an open question about implementation. When we think about SNAP payment error rates, if your SNAP payment error rate goes up beyond certain thresholds, if you cross into a high income or a high error rate state, you will have a larger amount of your federal funding cut.
(00:41:00):
There’s a lot of pressure to reduce snap payment error rates. Well, how do you go about doing that? You’re going to have to think about, do I have sufficient documentation to verify this person’s income? Is my benefit system up to speed and calculating benefits correctly, or is my benefit system out of date and actually incorrectly computing people’s benefits because it doesn’t have all of the policy changes? What verification do I have that this child is having meals with this adult? How do I verify all of those pieces? Child savings accounts are beginning next year, right? So the IRS and treasury will have to administer child savings accounts where for every child born in twenty twenty five, twenty twenty six, twenty twenty seven, these are all going to have accounts open for them and seated with $1,000, you guys, you might’ve heard of MAGA accounts, but these are basically child savings vehicles where individuals can deposit up to $5,000 per child tax free or tax deductible.
(00:42:08):
How are those accounts going to be set up? How can individuals access those? Those are all pretty first order implementation questions. When we think about recent history of implementation, I think that there’s potentially a lot to be concerned about. And again, this is just across many different administrations, so I don’t think this is necessarily tied to ideology, but we have many examples where implementation mattered and we weren’t able to deliver healthcare.gov. This was a transformative piece of legislation with the A CA, right? Coming online, a simple website to get individuals access to health insurance plans rolled out, and it was a disaster.
(00:42:58):
The website crashed, people weren’t able to access it, and when it did work, they were not able to find the information that they needed to actually enroll in plans fafsa. So I have a niece and nephew that are applying to college this year, but people applying to college in the last couple of years when they try to apply for financial aid, it’s been a disaster. I understand that the education department has had a lot of changes and had a lot of black eyes, but when you think about being able to submit an application for financial aid, you would expect that the application system would be able to populate your IRS income given that it’s part of the government that IRS data retrieval tool did not work properly. The FAFs of submission system crashed system. Many individuals, thousands of students were not able to successfully submit applications to the FAFSA or their FAFSA applications, their applications for Pell grants.
(00:43:58):
Many individuals ultimately did not end up applying. Were not even sure yet on what the enrollment impacts were. So this is just the last two years. If we think about other tax benefits like the earned income tax credit, well, the IRS has been doing correspondence audits for the last 30 years. These are audits of the earned income tax credit, the child tax credit where individuals are sent letters asking them, can you provide verification that this is actually your child? Can you provide verification that this is actually your income? In my own experience, actually when I had, my son is now eight, but when my son was two, I was actually thinking, well, how would I verify that this kid lives with me? And it’s not really easy. I mean, I had to go talk to a tax lawyer and it turns out that a note from the pediatrician is sufficient.
(00:44:52):
I had no idea. I don’t even know if the pediatrician knows that my child would satisfy the residency requirement, but why does the IRS use that as documentation? I’m not sure. But what I found out even going beyond this was that most audits are never responded to about 70 to 75% of audits. The taxpayer never responds. Maybe that could be a non-compliance. Maybe individuals are like, oh, you caught me. I’m just not going to say anything. Maybe some of it’s confusion. Maybe some people had no idea what income information to provide or what child documentation to provide. Either way, what we see long-term is that when individuals are audited for the EITC, most of them don’t reply. And then for literally the next seven years or the amount that we have data, we don’t see them claiming the EITC. When you think about somebody that is going through SNAP or going through Medicaid and having income verification, if they don’t know what documentation to provide, it’s going to be pretty difficult and they might not apply this year.
(00:45:59):
They might go without coverage for multiple years. So all of these pieces, implementation matters a lot. Okay. I think that the term user experience was never used in economics before or the analysis of government policy before, but when we think about programs like IRS direct file, when we think about how to actually go through implementing a program so that users have an easy to use interface and be intentional about that, that can have a pretty significant impact. So I want to kind of highlight that. We know from past experience that we have to take implementation pretty seriously. So before I get into thinking about thoughts on how to evaluate policies, let me just pause for a second and see thoughts and reactions so far to thinking about implementation or some of the other pieces we’ve talked about. Yeah,
Jack McGee | Springfield Daily Citizen (00:46:55):
My name is Jack McGee. I’m with the government reporter with the Springfield Daily Citizen in Springfield, Missouri. Going back, you don’t have to go back, but on the first slide regarding Medicaid in Arkansas, and the final data point on there showed the employer sponsored health insurance, which obviously increased when the law initially went to effect and then slightly decreased whenever the judge struck it down, but it was still measurably higher than what it was before that law took effect. How do you, not to try and get to the weeds, but how would you interpret something like that?
Day Manoli/Georgetown University (00:47:34):
Yeah, I think of this as when given the option between Medicaid or the public coverage versus employer coverage, people have opted for Medicaid coverage. It’s more generous. It’s easier to deal with if they are facing some requirements or some additional recertifications rather than going through that administrative burden, they’ve switched over to employer sponsored health insurance where you don’t have to go through a recertification as long as your employed, your employer knows that and provides that coverage. Whereas if you’re on Medicaid and you have to go through this recertification every six months, that administrative burden by itself becomes enough for you to switch over to the employer sponsored health insurance. But other things being equal, if people have the choice between Medicaid versus employer sponsored health insurance, it seems that they go with Medicaid given the sort of simplicity or generosity of the public health insurance option. Yeah, yeah.
Kirstin Garriss | Independent (00:48:37):
Hey there. Kirstin, Garriss, creator, journalist based here in DC with the new Substack. I am curious with the one big beautiful bill act, obviously a lot of the program implementation matters. A lot of that will happen next year. Thinking about the work requirements for Medicaid, you mentioned one example that takes a lot of time to verify if one person is actually fulfilling these requirements. I imagine. Are there any case studies looking at how much that costs for government offices to realistically do this efficiently? And I mean, I guess I can’t imagine. I’m trying to figure out how much does that cost to do that when you have to do it across the country?
Day Manoli/Georgetown University (00:49:13):
Yeah, okay, great. So I think putting numbers to the costs of doing this, so even though this is a work that is coming up next year, states are already taking on this work, and I think some of the estimates here, when you start thinking about the staff time to actually go through the income verification, like the hours to contact somebody and get the documentation, it’s just prohibitively costly. And so there’s been a lot of push to just change the landscape of the game by getting more access to more data. So CMS and is looking into building like an IRS data retrieval tool, similar to the FAFSA data retrieval tool where individuals can provide their state agencies with access to their tax return to verify income. It doesn’t quite meet the need because we file tax returns annually. If we need to do recertification every six months, that’s not really great. So then we have to start thinking about, well, what data do we have that could meet a six month requirement? We have quarterly unemployment insurance records, but that again, is just basically W2 or unemployment insurance covered jobs. If you’re driving for Lyft, if you’re doing some gig economy work or if you are self-employed, how do we verify that? And the short answer is that the costs seem to be prohibitively expensive in terms of staff time.
(00:50:47):
I think that there’s a lot of push to really think about, well, what other technological solutions could we really implement? But again, getting to the implementation matter piece, if it’s difficult to verify when you’ve set up a whole system on difficult to verify information, then maybe we need to change the underlying rules for eligibility. And so I think you’ll start seeing more of that policy discussion emerging towards the latter part of next year when states are right now, it seems to be this sprint to try to stand up what we can and then make recommendations on this would really help us a lot if there was this sort of change in regulation or change in the legislation. Yeah.
Mini Racker | The Nevada Independent (00:51:34):
Hey, I’m Mini Racker with the Nevada Independent here in dc. I just wanted to push on that a little bit more. If I wanted to report a story on this policy change is going to cause this much administrative cost, what data would I look at? How would I pursue that story?
Day Manoli/Georgetown University (00:51:50):
Yeah, great. One, I would start with your State Department of Human Services or the State Social Service Agency. The reason why I would start there is because all of the programmatic costs are going to be contained at the human services agencies. So thinking about how many hours of work does a caseworker need to spend on recertifying Medicaid case or recertifying, a SNAP case, the benefit amounts that information you can calculate using publicly available information on what are average SNAP benefit amounts, what are average Medicaid costs, but thinking about the costs of recertification, that’s really going to be the staff time that’s involved. And so working with the social service agencies to get estimates of that will be really crucial here. The other pieces that there are some, so one, working with the state and local level data, but two, there are some organizations sort of multi-state organizations like A-P-H-S-A, the American Public Human Services Association that cover information across states.
(00:53:11):
So to think about why is California a high payment error rate state, but Kansas is not, what are some of the differences in cases? What drives costs in California versus in Kansas? That’s where the multi-state organizations can be pretty helpful. So I would start with the individual state thinking about Nevada. I actually have talked to folks in Nevada and with other states, so happy to help facilitate some context, and then I would think about the cross state comparison to put it in a better context. Yeah, great. Okay. So I think maybe that’s a nice segue to think about how do we understand or think about evaluating policy changes? And so I think of this as where our fields kind of meet, I think of this as storytelling. So if you think about an economic model, you’re really just specifying the plot of the story or the characters of the story.
(00:54:12):
Who are the active players in the story? So when we think about recertification, that’s going to be thinking about caseworkers. The caseworkers are an active entity in doing recertification. The beneficiaries are also active. If I am a caseworker and I have to ask the beneficiary for some additional information that’s involving both of us. If you think about the IRS auditing taxpayers, that involves an IRS examiner as well as an individual taxpayer responding to that examiner. If we think about well calling the HR department at Target to verify employment, the employers are going to be pretty active agents here. When you think about recertification and work requirements, are we going to see, are employers seeing large droves of individuals now banging down their doors looking for jobs? We want to know, well, what are the employers seeing? So really understanding who are the active agents in a given story or in an economic model. An economic environment is pretty, I think like the first starting point.
(00:55:24):
Okay, so you think about motivations of characters, right? So what are the incentives that are at play in the environment? If somebody loses benefits, well, what are they going to do? And I would say, what is the path of least resistance as a starting point? Usually the path of least resistance is doing nothing. If you lose health insurance coverage and you do nothing, you see increases in the on insurance rate. So if people are uninsured, but they’re still getting sick, well, they’re going to still continue doing what they were doing. They have to go to the emergency room, they’re still going to go to the emergency room, but now they’re moving from a situation where they were covered under Medicaid to maybe a situation where they were not covered under Medicaid. Do you have more uninsured individuals showing up? You’ll have the same number of individuals, but maybe a larger percentage of them are uninsured.
(00:56:21):
Similarly with snap, if you were relying on receiving food assistance and now you don’t receive food assistance, well, presumably people still have the same demand for food, but so now they’re potentially spending some money that they would’ve spent on something else. Maybe housing on food. If they’re not spending money on housing, are they at higher risk of becoming homeless? Are they at higher risk of staying at a homeless shelter? So you can look at stays at homeless shelters. There is administrative data from HUD management, the Homelessness Management Information System. HMIS has data on each locations the number of stays at homeless shelters, but we also know that that’s a pretty far end kind of measure of housing instability, staying with friends, couch surfing, staying in cars. All of these are sort of intermediate measures. We don’t see an administrative data, but thinking about what is the richer story of what is going on, we need to know those kinds of outcomes as well. Okay, so thinking about the local context, the local environment is going to be really, really critical here.
(00:57:38):
And what I mean by local, this could be variation across different types of neighborhoods. So even within Washington DC, when you see individuals losing SNAP benefits, you could have a very different impact in southeast Washington DC than thinking about individuals near say, Navy Yard or thinking about northwest Washington DC or Northern Virginia. Local context will really matter here. Access to homeless shelters, access to hospitals, other sort of nonprofits, all of those pieces will be pretty important. What are the underlying assumptions here? So when you have a story, you’re really telling a causal effect. You’re kind of saying like, well, A is causing B. I have some assumptions behind that. And so you just have to be upfront about what are the assumptions? Okay, every statistic or these are just numbers. These are correlations that I see in the data. Causality is something that I bring to the table.
(00:58:38):
So as a researcher, as someone consuming this evidence, I am bringing that causal story to this interpretation of data. I just have to be upfront about what are the assumptions that I’m making in order to tell that causal story. Once I have those assumptions explicit, I can try to provide supporting evidence. Maybe that’s convincing, maybe it’s not. But then at least if we’re on the same table about what are the assumptions, then we’re onto the next step of is this evidence credible or not credible for making those assumptions? You and I might see the same data, the same set of assumptions, the same supporting evidence, and just come down on different sides. You find it compelling. I don’t. Okay, but we’re on the same page in terms of what is the necessary assumption in order for this to be credible, I think researchers, evaluators, we’re here to help translate this evidence or these assumptions to what is credible, what is not credible, what is known versus what is unknown.
(00:59:46):
I think you guys have already asked about data sources, thinking about what studies are credible, what research is useful or not. You should feel free to just fire away with myself, with anyone else, with any researchers. I think tap into these sources. I think that they’re there to help and really hopefully provide this kind of gut check on what is a credible assumption or what is credible evidence. I think you’ll kind of see from the way I’ve structured this talk, I think pictures are really important. So graphs really tell a lot. Pictures, I think tell a lot, whether this is local context versus a statistical graph, I think that the pictures are really meaningful. To give you an example, when you think about implementation matters, if I asked you guys to think about the IRS’s implementation, I think many of you would be familiar with the picture of the cafeteria in the IRS that was filled with boxes of unprocessed paper returns. That picture I think really captured the IS in a nutshell.
(01:00:55):
I mean, it came up in discussions of 80 billion of funding for the IRS, that one picture. So pictures have a long lasting impact. That also kind of highlights the importance of making the pictures clear. If you have a story that you are telling, make sure that the evidence, the graphs, the pictures are very clearly and compellingly communicating that evidence. I’ve tried to pick out a couple examples here to really drive that home, but I think that the picture drives home the main message or the takeaway that resonates with people, the data, the numbers, I’m not necessarily going to remember numbers, but it sort of fills in the details of the story. So adding the quantitative evidence, I think helps round out the picture. But the picture is really like your lead, I think, in terms of both communicating evidence and really giving people a takeaway.
(01:01:56):
The last piece that I’ll mention here is really thinking about counterfactuals. What would’ve happened in the absence of the policy change? And I think that’s a critical question to ask in order to evaluate the outcomes of any analysis, the outcomes of any story, to understand if these people were not taken off of Medicaid, what would’ve happened if these people were not taken off of snap? What would’ve happened? And when you have that counterfactual in mind, that comparison that allows you to clearly communicate the overall story of the evaluation, any statistical evaluation of policy analysis should clearly address all of these.
(01:02:48):
I think of this, again, as researchers or economists doing statistical analysis. This is all just storytelling. If you read a paper and you are wondering, well, okay, so what are the assumptions for these individuals? Am I assuming that they were able to access the IRS website to submit income information? Am I making the assumption that the caseworkers knew how to verify income? What are those assumptions? All of that should be clear in any statistical analysis or in any other story of what the policy impacts are. Okay. So again, I think of professors, researchers, like people doing the data analysis as sources that are very much going about the same processes of storytelling. So I think of this as we’re in the same boat. Happy to put heads together, refer you to additional sources. We talked a little bit about state, state agencies, federal agencies or consultants, data sources, or other papers for other evidence. Feel free to reach out anytime. So I’ll stop here and happy to take any additional questions, but I think you should definitely feel free to get in touch anytime. I am happy to talk about research, tax policy, workforce development, et cetera. This is kind of what I do. So thank you guys very much, and yeah, happy to open it up for additional questions.
Mary Steurer | North Dakota Monitor (01:04:27):
Hi, my name’s Mary. I’m from the North Dakota Monitor. Thank you for being here. I know you said at the beginning that you’re not here to tell us what good policy is, but you’re an expert, so I’m wondering if you have an all time favorite policy where there was a goal and the lawmakers really nailed it in terms of accomplishing whatever they were trying to do. I guess that would include both really wonky policy or even, I don’t know, civil rights policy. Is there anything that really stands out to you as your favorite or that you’re a fan of? I guess to put it a different way, thanks.
Day Manoli/Georgetown University (01:05:12):
Sure. Okay. So I think two examples come to mind. One that I did not talk about in this talk, and so I think one is a negative example on something that I think was just not clearly thought out, and the other I’ll come back to in the context of this talk. So the first example on something that was I think well-intended that was really badly executed was the tax treatment of reparative payments. So there was a case where individuals had claimed to a historical property because their ancestors were slaves on that property. It was initially under a historic, it was a historical foundation that owned the property, but they brought a case saying that they were rightfully owed the property after a pretty lengthy court battle. This was before the pandemic, so it was a couple of years before the pandemic. They ended up receiving ownership of the property, but the tax treatment was not specified in the ownership of the property.
(01:06:22):
So essentially they were handed a massive property tax bill. They were not able to finance that property tax bill, so they had to sell it to the historical foundation. I think that it’s sort of a very simple kind of idea that this was meant as a reparative transfer, the execution of it, the court just didn’t specify that this should not have been a taxable transfer at that time. And so it was a pretty unfortunate kind of circumstance, and I think it’s literally this property now is, I think that there is still very much a lot of contentiousness around it. That’s one example. The other example that I would say is very close to home is based on welfare reform. And so what I want to kind of highlight here is a couple things. So one of, as an economist, especially working on employment policy in the United States, if you were to ask most employment economists or most public finance economists about the earned income tax credit, I think that they would say the earned income tax credit is a big win for policy.
(01:07:31):
It’s very successful in creating incentives to get people that were not in the labor force into the labor force. More recently, I think we’re sort of revisiting this and sort of saying welfare reform was happening at the same time. A lot of the same individuals that found jobs were also kicked off of welfare, and so was it the stick or the carrot that really mattered when you start thinking about this, okay, it’s kind of hard to disentangle, but let me just kind of focus on the implementation of the EITC. If I really believe that the EITC is driving individuals to get a job, it’s a little bit peculiar that, well, 20% of individuals that are eligible for the EITC don’t claim it. Most individuals, when you ask them about their earned income tax credit, don’t seem to be aware of it. How is it that I can tell you with a straight face that this earned income tax credit that is pretty complicated, buried in your tax refund is generating huge increases in employment, but then also at the same time, a lot of people aren’t claiming it.
(01:08:37):
It’s super confusing. So we’re revisiting that evidence and saying, oh, you know what? We can design a much better EITC. I think you’re starting to see some of that discussion and maybe happening more with states even. So if I’m working with the SNAP employment and training folks in Maryland, for example, we’re thinking about can I provide a stipend to get individuals into a training program if I provide them with a quarterly employment bonus that says, Hey, you’re making $2,000 this quarter, I’m going to give you an additional $500 if you stay with this firm. So it leads to this idea of, well, you can improve on policy, I think if you believe that these are effective or if you think that there is evidence. And so I would say that what’s really cool about that is we’re kind of learning from past evidence, questioning it, and then thinking about ways for incremental improvement. For better or worse, it’s not sort of night and day changes quickly, but it’s sort of like incremental progress. Yeah.
Taylor Vance | Mississippi Today (01:09:44):
Hi, back here.
Day Manoli/Georgetown University (01:09:46):
Yes,
Taylor Vance | Mississippi Today (01:09:47):
My name’s Taylor Vance and I’m a reporter with Mississippi today, and you mentioned tax policy. Mississippi this year just decided to do a pretty radical experiment and eliminate its state income tax, which is a third of our total state budget. Mississippi is also one of the most federally dependent states in the nation, and of course we’re seeing a lot of instability on the federal level right now. So I guess with that being said, it’s also a little bit of an interesting problem because when you report on tax policy, reader’s eyes kind of clays over, and if you ask people, well, do you support paying less taxes? Of course they’re going to say yes. But if you also ask them, well, do you like driving on safe roads and bridges? They’ll also say yes. So I guess I would just ask you what advice do you have for trying to contextualize that, make it interesting, but continue to evaluate this policy going forward in the years to come?
Day Manoli/Georgetown University (01:10:57):
Sure. Okay. So a couple pieces on that. So other states have also experimented with abolishing state income taxes. So thinking about Kansas for example, and I think elevating some of those stories is pretty important. There’s been a lot of research on thinking about what were the impacts of eliminating the state income tax in Kansas? What were the impacts on school funding in particular was very much impacted, and so we’re seeing that aftermath lasting almost four or five years after the pandemic even. They’re still grappling with some of these changes and going back on some of them. So one I think is sort of using historical evidence to guide some of the policy discussion. The other thing I think is maybe on the sort of research production side or the evidence production side, recognizing that people don’t seem to have the tax benefit kind of linkage.
(01:11:54):
Maybe this is a failure of civics education. When I talked to my students about filling out tax returns, many of them have just done their only experiences doing this on TurboTax. How to connect that, the roads that you’re driving on all of that infrastructure, how that’s connected to tax revenue might not be clear, but that seems like, oh, actually, in terms of evidence, that’s something that we need to now address, right? To say like, oh, okay, well if I frame this as are you in favor of taxes that sort of provide employment incentives or provide breaks for small businesses if I make it easier to file taxes for a small business owner? Those are kinds of things again, that I think coming back to this concern about political ideology, it seems that most people just kind of generally agree on, but you have to have that sort of communication about what is the tax benefit link? And researchers policy makers need to understand, oh, that link wasn’t there, or it was there, but it faded out. And so I think we’re just kind of learning about how people think about fiscal sustainability, how people think about government debt, how people think about taxes and redistribution, and maybe realizing we didn’t fully understand some of the context in which people were making these decisions. Yeah.
Simmerdeep Kaur | Walla Walla Union-Bulletin (01:13:17):
Hi, this is simmer again from the Union Bulletin. Thank you for the presentation. I was wondering, according to your expertise, are there any reforms in the big beautiful bill to Medicaid that are going to positively impact its enrollee or just the overall community?
Day Manoli/Georgetown University (01:13:32):
So I think that it’s going to, yeah, again, this is just sort of me editorializing, but I think that across states, I think you’re going to have a lot of separation in terms of some states kind of trying to shore up their programs with both SNAP and Medicaid, trying to introduce additional taxes to gain additional revenue to make up for the lost revenue on the federal side, and then you’ll have some states that I think shrink their programs and maybe even shut them down. Okay. So what could be positive about some of this? Well, I think that there is something to be said for improving government data systems, government or public administration systems. This is somewhat something that maybe should have happened a while ago, but now there’s more impetus to do this. So everything from designing a very intuitive user-friendly SNAP application or Medicaid application that feeds in data to a caseworker that can then say, I can now query my quarterly wage data to verify that this is actually satisfying the income documentation.
(01:14:39):
That seems like something we should have had all along, but now we have to do this because of tracking all the work requirements, tracking the recertifications, even if it’s for a smaller subset of people, maybe it’s a more efficient program administration. So I would say adopting new technology is maybe one positive piece that comes out of this. At the same time, I think that we’ll probably start to see a little bit later, I think towards the end of next year, beginning of 2027, recommendations for what would be better policy or what would be policy improvements rather than going down the road of the big beautiful bill and the cuts and funding that are coming up. I think now people are trying to meet the requirements or the sort of a payment error reduction, the work requirement implementation. They’re sprinting to get that done. I think in a couple months after that, we’ll start to see, here’s what I would’ve liked policy to look like, and I think that will be a pretty exciting time to see what some of those insights, what insights are emerging. And then you might have some more separation in states adopting some of those in their individual state funded programs as opposed to the federal landscape. Or maybe there could be some insights that are, we have 50 different labs of democracy here. Maybe some of those insights will be elevated to future federal legislation. But very concretely on snap payment error reductions. I know multiple states are considering how would I snap audits to be conducted if I could do them myself, and maybe I can inform future federal legislation.
Kevin Johnson/NPF (01:16:23):
This is the worst part of my role here. I have shut it down knowing there are hands and arms in the air. But in order to make our next stop at the reception for the next session, we’re going to have to wrap up, but please join me in thanking Professor for such an instructive session.
