Government administrators often write complex and interacting rules that make it harder to access public programs to improve health and social welfare. They impose compliance, learning, and psychological costs on the people that these programs are trying to help. Donald Moynihan and Pamela Herd have launched a renewed recognition of the barriers that program beneficiaries face and documented how many burdens are knowingly implemented to undermine government success.
Guests: Donald Moynihan and Pamela Herd, Georgetown University
Studies: Administrative Burden: Policymaking by Other Means
Matt Grossmann: How administrative burden undermines public programs, this week on the Science of Politics. For the Niskanen Center, I’m Matt Grossmann.
Government administrators often write complex and interacting rules on financial structures, eligibility, compliance, and documentation that make it harder to access public programs to improve health and social welfare. They impose costs on the people that these programs are trying to help for everything from direct compliance to understanding the rules to the mental burden of being caught in the red tape. This week I talked to Donald Moynihan and Pamela Herd of Georgetown University about their Russell Sage book, Administrative Burden: Policymaking by Other Means, and related research. They’ve helped launch a renewed recognition of the barriers that program beneficiaries face and a revolution of social science research designed to improve how programs are designed and implemented. But along the way, they’ve noticed that many burdens are knowingly implemented to undermine government success. I think you’ll enjoy our conversation.
So Pam, what is administrative burden and why do we care about it?
Pamela Herd: Administrative burden is the kind of onerous encounters that we have not infrequently with public services and benefits or with the government more broadly. It’s waiting for three hours at your local DMV to get your license renewed or having to fill out paperwork four different times seemingly providing the same information basically over and over again, the time involved, the money sometimes involved in meeting these kinds of requirements to access state benefits and services.
Broadly speaking, we think about this burden as having or entailing three different kinds of costs. So these include learning costs, which are sort of figuring out which benefits or services that you’re eligible for and need, compliance costs, which are that sort of time, paperwork, interviews, those sorts of administrative ordeals. And then the third category are psychological costs. These psychological costs can entail everything from just sort of stress and frustration. It can entail stigma. But broadly speaking, they’re the sort of psychological costs that might be a consequence of these encounters.
We think these burdens are important for a few different reasons. One reason they’re really important is because they actually impede access to needed benefits and services. So we know for example that one in five people eligible for food stamps don’t receive them in large part because of these burdens. Similar for Medicaid or public health insurance for low income individuals and families. And then other programs are much worse. So for example, the Supplemental Security Income program, which is a really kind of poverty based program for older adults and people with disabilities has take up rates about 60%. So 4 and 10 of people, really poor people eligible for these programs aren’t getting those benefits.
The second reason why we think it’s important is because we think it really affects our kind of understanding and trust in government. Can government deliver on what it’s promised to us? And these negative encounters, we think, and there’s some evidence to this, leave people feeling like government just isn’t up to the task.
Matt Grossmann: Don, which of these are the most severe and what are the kinds of examples that readily come to mind when you think about these costs?
Don Moynihan: Yeah, so I think there are examples we all experience in our everyday lives, but I remember when we started working on this project in Wisconsin, one that really struck me was the state was trying to figure out if people who were eligible for public health insurance were actually working and could get that public health insurance through their employer. And so with the best of intentions, they sent out a letter to these individuals saying, “Can you go and verify from your employer that you don’t have health insurance?” And as you might expect, something like 17% of those people in lost health insurance and didn’t get it recovered from their health employers. It was an example of where the policymakers weren’t necessarily trying to do anything ill intentions. They were basically trying to make sure that people were getting health insurance appropriately, but where individuals found that step of going to their employer and getting confirmation of health insurance status to be a known really difficult step in sort of a surprising way.
And so I think this sort of feeds back to what Pam mentioned about, and sometimes these burdens occur in ways that are surprising and are large in ways we might not anticipate. I think one other lesson from that example is that new burdens are especially onerous. We go through life with a set of expectations about how our interactions with administrative services work. And over time, if we are dependent upon certain services, if there’s a change in those, it can have these fairly dramatic effects. And so there are other examples in, for example, Arkansas when they introduce work requirements for Medicaid. A paper in the New England Journal of Medicine founded about 95% of those who lost their Medicaid coverage due to the work requirements were actually working or should have been exempted because of a disability, but where they really struggled was with the paperwork requirements. It was really the reporting part that they found difficult. It wasn’t that they weren’t actually completing the eligibility guidelines.
And so I think partly it’s about the nature of the burden itself when it occurs, if it’s new. It’s also about the situation that people find themselves in. So if they’re working with some evidence that if they’re experienced scarcity, that affects their cognitive processing and executive function and their ability to manage these sorts of tasks. Similarly, if you’re experiencing significant negative health events or have generally poor health or are struggling with mental health, that also diminishes your ability to manage things like forms, appointments in ways that can really be counterproductive if you’re trying to get public help for the health problems that you’re facing. So I think you need take a broad view of not just the design of the burdens that arise through public action, but also how they might affect different groups differently.
Matt Grossmann: So Pam, you do find that almost all of these burdens affect groups unequally and are especially hard on people lower in socioeconomic status or socially excluded groups. To what extent do you think those are by design that the policymakers kind of intended to make it easier on some than others, versus this is just what happens when you kind of increase regulation in support of other goals?
Pamela Herd: So this is of course the easy answer, but both. So it is definitely by design in some cases. And so we see this for example… One example from our book is around the earned income tax credit actually. There’s discussions basically among lawmakers where they’re basically saying, “Look, I get that it’s a program that is focused on people working, but it’s still a welfare program and it shouldn’t be that easy to access,” right? People can be very explicit about the idea that we kind of should have these kinds of burdens in programs. But it’s also the case that sometimes they are unintended. So Don gave the example about Medicaid in Wisconsin when they asked people to get letters from employers. They really didn’t intend to kick people off Medicaid who needed Medicaid. And they reversed course in the end actually once they realized.
Another really big example of this is actually California, the Supplemental Nutrition Assistance Program. So take up rates for food stamps vary substantially across the country. In places like Wisconsin actually, which are kind of purple, over 90% of people eligible for SNAP get benefits. In California, a state that we see as solidly blue as very progressive, the take-up rate for SNAP is about 70%. So the reasons for that are probably not, and there’s some evidence I think about this, not really ideological exactly, but they have in California a county level administrative SNAP program. So each county administers the program differently. You do tend to see Minnesota’s also like this, in states that choose to do county level administration, it seems to cause more burdens and barriers and you get lower take up rates. So it really varies some by design, and then in other cases there are sort of administrative issues that are driving that burden.
Matt Grossmann: Don, this is a kind of reputation that bureaucracies have that they create complicated rules for their own sake or that they’re used to filling out the form so they assume other people would as well. Is that a deserved reputation? And to what extent is it just about large organizations in general or government in particular?
Don Moynihan: So I think there are absolutely examples that we would be able to bring to mind of administrative burdens that we’ve encountered in the private sector. So if you try to cancel a subscription to the New York Times or if your gym membership or if you try to schedule a cable person to come out and help you. I think the lesson is sometimes these burdens in the private sector actually serve the bottom line of the company because they’re rationing the provision of services in some ways or it’s to their benefit that you don’t exit out of their services.
One difference with the public sector is there isn’t the same private sector bottom line. There isn’t the for-profit basis for the operation of the organization. And I do think there is some justification to the stereotype that public organizations are less able to get feedback about how well they’re serving their customers and they don’t have the same incentive structure and feedback loops to use that information to improve things. I think there is more attention to that than there has been in the past. Maybe we can talk more about the Biden administration and what they’re trying to do in this area. But certainly if you think about public org, one of the things that public organizations do is to impose negative externalities upon each other and also members of the public. They don’t always perceive those negative externalities, or by definition pay a cost for them.
Matt Grossmann: Pam, these burdens are often justified as necessary for reducing waste, fraud, or abuse of programs. So to what extent do they actually achieve that? There’s kind of a real trade-off here between less fraud and abuse and more burdens. Is it worth paying some cost in burdens to try to reduce even the appearance of fraud? I’m thinking of some of these public goods games that you see played all over the world where people are willing to pay some extra cost to punish those people that might be cheating off the system. Is that sort of something we should expect to have to pay?
Pamela Herd: Yeah. I mean, one thing that we kind of argue in the book is that if you’re going to impose burdens, you want to actually understand the costs and benefits. And so the problem at the moment actually is that federal programs do a much better job of tracking kind of fraud waste and abuse in programs than they do in tracking who ends up not being able to access services, right? And if basically you’re looking at fraud abuse and waste rates of say like 1 to 2% in a program, but 20% of people eligible for benefits aren’t getting that program, you really want to compare those benefits and costs basically associated with it. So I think that’s the first point, which is that’s sort of key. We’re not saying burdens are always bad. It’s just you really need to think through the costs associated with them.
I think the second point I’d make is that I think this idea that we have to have burdens to keep fraud and abuse low is wrong. So the program actually that has probably the least fraud waste and abuse would be the Social Security Retirement program. It’s also our least burdensome social welfare program. The government keeps close track of our earnings. They do that in conjunction with employers so you don’t really have to do anything except sign up whenever you want to kind of partake in those benefits. And there’s very low rates of fraud and abuse in the program. So these two things don’t have to go together basically. And in fact, sometimes the processes that you need in place to deliver benefits with very low levels of burden are also processes that can ensure that you don’t have that kind of fraud and abuse.
Matt Grossmann: What about the second point? Is there sometimes when it might be worth paying extra burdens just to reduce the appearance of fraud, in other words, to get people to go along with the existence of the program and see these as protections?
Pamela Herd: Yeah, I mean I think the question is, going back to the benefits and costs, whether or not they actually equal out, right? So yeah, I mean I guess in a hypothetical world, a little bit of burden to ensure the popularity of a program and maintain it could be a reasonable trade-off. I’m just not sure that we have a lot of evidence though that is working in practice, that the relative cost and benefit trade-off is working in practice.
Don Moynihan: There was a nice paper by Susan Miller and Leo Kaiser that looked into degree of confidence in vignette experiments where people were described a program with different levels of burdens and found that people tended to have higher confidence when there was more steps that people had to take, but it was driven by more conservative voters.
And so I think empirically this may be true. You might see an increase in confidence if you held people’s hands and said, Here’s what happens. And then we do this and then we do this.” I think there’s a second question then about, is that how we actually communicate this information to the public, right? Do we provide these point by point explanation of steps? Or is the communication about these issues much more crude and maybe actually in fact unrelated to the actual empirical risk that programs would face? So I think in any given program, one thing we try to talk about the book is to use some of the logic of cost benefit analysis, like have a realistic discussion with evidence about what are the trade-offs between how many people will be negatively affected versus the benefits that you’re hoping to get from imposing more burdens.
Matt Grossmann: So you find it at least some examples where high burdens may be a product of low administrative capacity, but we also can sometimes hear the argument that giving administrative agencies more capacity may incentivize them to create more rules and regulations. So is there kind of a trade-off with building big agencies or is bigger mean more help to potential beneficiaries?
Don Moynihan: So I think if you start from a state of nature and there are no public services, then there are no burdens, right? So when we talk about some of these discussions in comparative terms with colleagues studying in much poorer countries, unless developed welfare states, some of the discussions are not relevant to them because they simply are not providing the types of services where they’re trying to determine eligibility.
As you start providing those services, especially ones where you’re trying to sort through which citizens are eligible, which are not, then you have to think about those mechanisms of sorting them out. So I think partly it’s about the growth of the state and when it takes on new services that generally tends to come with bureaucracy. So first of all, that relationship is not automatic. So for example, we now provide much more of our welfare state through the tax system, which doesn’t come with a big bureaucracy beyond the initial investment we make in the IRS. I think it’s also about the design of bureaucracies and not just their size. And so this is something I mentioned earlier about I think the Biden administration doing a relatively good job in trying to communicate to agencies, “Hey, you have to think about the effects of what you’re doing on citizens and redesign administrative processes around the knowledge of those effects.”
So I think the ideal is a high capacity administration that is very attuned to when it’s imposing a necessary burdens on the public as opposed to a very large administration, which is sort of blindly imposing hassles on the rest of us that we don’t really have an opportunity to protest against.
Matt Grossmann: Pam, you focus on burdens for individual recipients, but I can imagine a small business lobbyist saying, “You all finally get it, but everything you’re saying here applies to business regulation.” There’s lots of regulations that aren’t clearly tied to the policy goals. There’s unnecessary costs that are imposed beyond what’s needed to actually carry out the stated purpose. And a lot of it’s just because the bureaucrats are creating rules for their own sake. So to what extent do those arguments apply the same way to business regulation as they do to these kinds of instances?
Pamela Herd: Yeah, I mean, I think we kind of talk about this in the book in terms of pointing out that in fact I think Congress has been very receptive to concerns by both small and large businesses actually about burdensome regulations. And there in fact has been legislation to try to address that. I think the irony that we kind of point out is that there’s the same folks, particularly in the Republican party who are most concerned about reducing burdens for businesses are pretty enthusiastic oftentimes about burdens on citizens in certain kinds of public programs.
So I think that’s more where we’re coming from. Yeah, absolutely, they have an argument there in the same way that I would as an individual. I just think on average they have more political leverage to actually get those kinds of burdens reduced for themselves in a way that, for example, people in targeted programs have less leverage. The flip side though is I would say a program like more universal programs like Social Security and Medicare, you know have organizations like AARP because you have a much broader swath of the population that has some political power and leverage that gives more space for those folks to be active and push to improve the delivery of those services and benefits.
Matt Grossmann: But should liberals take more seriously the conservative arguments that would be analogous on this issue? So maybe their psychological costs, cost to the people’s images of government, cost to willingness to work on behalf of public goals from private sector organizations going through this same kind of paperwork.
Pamela Herd: Yeah, no, I mean I hear what you’re saying. Yeah, it’s funny. The part of the reason though that we wrote the book was because we were actually concerned that progressives and liberals actually weren’t taking these burdens seriously enough for individuals trying to access public programs. So the point of the book actually was to get liberals to pay more attention to it, but get them to actually pay attention to it in programs like SNAP and Medicaid.
I mean, I was in graduate school in the 1990s during welfare reform. Sort of so much changed in public welfare programs during that period, and there was just no attention at all or very little concern at all about these kinds of burdens embedded in all these programs that they wanted to expand and people to access. The thinking at the time I think was more ordeals mechanism thinking, which was that if people really needed these benefits, they would overcome these obstacles and they were a lot more concerned about doing effective targeting and making sure that programs were targeted in ways that produced the kinds of outcomes they wanted with the assumption being that if people really need them, they’ll muddle through.
But I think one of the things that changed was effectively behavioral economics and kind of realizations that came from that literature about people don’t always behave quite as rationally as we think they do.
Matt Grossmann: Don, there’s a whole industry of usually nonprofits that are designed to try to help people with these administrative burdens in some ways that shows that the policymakers knew they were imposing big burdens. They’re often willing to support kind of a nonprofit sector to help people deal with those burdens. How should we think about these kind of quasi official roles that these folks take on? Do they end up being excuses for policymakers to make more complicated systems because there’s a group of people that they believe stand ready to help recipients?
Don Moynihan: Yeah, it’s a great question. When we started writing on administrative burdens, we thought we were writing about citizens state relations, but really it’s very hard to write about any aspect of American policy in any depth without taking into account third parties. And these can be nonprofits or private organizations that played a sort of outsized role in helping in a variety of ways to make services more accessible. And so the most obvious role that some of these organizations play is the helping role. So there’s evidence, for example, that if you are a member of a veteran’s organization in a state where those organizations are very active and engaged, you’re more likely to get disability benefits because you can rely on that solidarity based set of connections to help you negotiate the hassles and the paperwork that’s involved with that.
And in some cases, we pay these third parties. So if you think about navigators with the Affordable Care Act who are being rewarded to help people negotiate the hassles. Or if you think about the American tax system, and now we’re not talking about nonprofits here. We’re very much talking about for-profits like H&R Block or TurboTax. They do a tremendously good job at helping people get access to the earned income tax credit. I think take up of that benefit would be much lower without those tax preparation professionals.
Another role that these third parties play is that there’s sometimes the target of burdens. So sometimes if you think about abortion providers as an example, these are groups that governments have very strong interests in regulating closely and will impose a set of burdens upon them in a way that’s unusual for medical providers with the intent of limiting their ability to provide services to members of the public.
A third role is that of lobbyists. And so we live in America and people have the opportunity to make their claims about how services should be provided. I think in some cases that’s very good. You have nonprofits who are closely connected to people receiving services. Think about the immigration domain where they can articulate in a way maybe that the actual individuals themselves are perhaps afraid to about hassles that are imposed in immigration processes. And so look for ways to identify and reduce burdens.
Veterans Affairs is a great example where again you have really active organizations who are in constant conversations with veterans administration about how to make things work better. But sometimes those lobbyists also operate out of self-interest. And again, the tax preparation industry that I just praised a minute ago, which does a great job at helping people get access to the ITC, has also been a strong opponent of making tax reporting more broadly simpler for the rest of us to access because that would cut into their bottom line.
I do think this sort of feeds into a broader pattern of fragmentation where one reason why people experience burdens in America is we have a federalized system. You may be jumping between organizations that have different definitions of standards and it all seems bewildering. But sometimes the role of these private third parties is to help knit that fragmentation together and help you to negotiate those hassles.
Matt Grossmann: Yeah, I don’t know if either of you saw Dylan Matthews post recently about being a assistant for a tax charity, but one of the things he pointed out is that there’s five different definitions of a family unit that they have to walk people through, and so they have to ask these intrusive credit questions kind of repeatedly just on a basic point. But I think part of the issue there is that these are five different policy regimes all running through the past tax code. So to what extent is that the complexity driven not by any individual set of regulators trying to set the standards, but by the fact that we just are trying to achieve quite a bit?
Pamela Herd: Yeah, I mean this sort of goes back to a basic question of like, is this policy design or is it administration? That’s the problem, right? I do think if you look at the ITC and the CTC, there are aspects particularly to the ITC about the design of that policy that just make it really hard to administer. It’s confusing. Low income people got really confused about who’s eligible, which child is with which person, is the grandparent eligible. Family forms are really complicated now. And so trying to intersect that with the way that policy is designed clearly create… I mean, it creates a lot of error.
I do think on average, we know though that generically, even the people who get the ITC who aren’t really supposed to, they’re still largely low income. It really does appear to be that people just really struggle to navigate, in that case, the design. So there is this real tension about how complicated the design is and the degree to which you can reduce burdens with policies that are just very, very complicated. I do think it’s a key challenge on this issue around with some policies.
Don Moynihan: I would add, I think the lesson for policymaker from this domain is not just to put yourself in the shoes of people who are on receiving end of policies, but to understand that they’re on receiving end of multiple overlapping and not always consistent policies. And so there’s an opportunity there, I think, for greater coherence in how the American administrative state operates.
Matt Grossmann: Pam, you mentioned being inspired to write this book to inform liberals, but of course there are polarizing trends especially in American state governments. And these issues are sometimes talked about in partisan terms where Republicans want to reduce the number of recipients for social welfare programs. And obviously, in areas like abortion that you also cover have a very explicit rationale for reducing the ability to actually access that service. So to what extent are we seeing kind of polarization around administrative burdens? Do Democrats have any kind of coherent response to reducing burdens and are these going up more in red states?
Pamela Herd: So it’s interesting because I do think this is now an issue that progressives are actually really taking more seriously. So for example, in the debates around the expanded child tax credit, the Romney proposal included the idea that it should be administered by the Social Security Administration, which I think was in fact pushed by some think tanks in DC and folks in that space whose point was they’ll be more effective basically at ensuring people access the CTC. And there really were problems. I mean, there are millions of people who we most wanted to access the CTC who didn’t in the end. And not because the IRS is a terrible organization, but just because they really are not a social welfare agency, right? That wasn’t what they were designed to do.
So I do think there is more attention to it now on the part of liberals and progressives in terms of understanding this actually really does matter and they’re kind of organizing around that. I’m not sure it’s really any different. I mean, I think part of our point in the book was conservatives actually had been way more effective at using this as a political strategy, particularly around abortion and voting rights where they were pretty explicit about their goals and strategies to achieve these outcomes using burdens. I’m not sure that much has changed in that way. I think there’s still real evidence that they continue to see that as a kind of viable strategy.
Is it getting more burdensome in red states? I think it just depends on which policy domain. I mean, I think that programs that are highly federalized, there’s only so much they can do. And the Biden administration, for example, around programs like Medicaid and SNAP has been pushing states really hard to administer those programs in ways so that they’re not super burdensome. And states can always kind of push back. But in general, the more federal control over programs, the less burdensome they tend to be. So it just depends on the program.
Matt Grossmann: So you both have also been actively researching health policy and were following the pandemic. Is there anything that we can say we’ve learned from the pandemic experience about the administration of programs? Were there positive examples of cutting through red tape because we needed to deliver services quickly? Or did all of the same problems reemerge?
Don Moynihan: I think there were some positive lessons. I think the headline would be in case of emergency, keep it simple or build on what works. So for example, when we hit an economic emergency or a health related emergency, expanding SNAP access is a pretty proven way to get money to people who need it and will spend it and will generate economic activity. And so that is attack that we can sort of turn on when major events happen. I think the unemployment checks worked pretty well, right? So we managed to get… Or sorry, the pandemic relief checks, not the unemployment checks, which I’ll talk about in a second. But the pandemic relief checks worked well. We were able to identify a subgroup of the population and give them a specific amount of money. And that’s because the program was very simple and because we had the administrative data on hand to execute it well.
So the eligibility was basically based on dependence and whether you were below a certain income level, IRS had that information. And based on that, it could easily target the people that it needed to. It was still an achievement. But I think the lesson again for policy makers is you can do these big things if you don’t add a ton of eligibility criteria to the design of the program. Obviously, there are programs that did not work as well. Things like, I think, Rental Assistance really struggled partly because that was a new program, partly because there was also a sort of double burden involved in that you had to get the participation and administrative engagement of both the rentor and their landlord. And so you’re sort of layering additional amounts of complexity onto the process. I think clearly also the UI system revealed existing cracks in both the design and administrative capacity of state governments who had either neglected that area or in some cases had deliberately designed it to be inaccessible and just really collapsed when huge volumes of people tried to use that system.
Matt Grossmann: So we’re entering a period of prolonged budgets, debates in DC, and one of the main things that house Republicans are pushing is for additional work requirements for social welfare programs. So I guess let’s say that we agreed in principle that there should be programs that people are only eligible for if they work during the year. Why is it that something that sounds like it should be simple actually become a big source of complexity and burden?
Pamela Herd: Yeah. So anytime you add more conditionality to a program, anytime you add an additional eligibility criteria basically, you’re going to make the program more complicated. So there’s just sort of that basic reality. Now, why work requirements in particular? So I think the first issue is that a lot of programs, both the most relevant ones for this discussion, both Medicaid and SNAP, include a bunches of populations who would not be subject to a work requirement. They’re either too old, you’re too young, or you’re disabled. And so you automatically have a group of people who ideally you want excluded from these requirements. But it’s not actually so easy in practice to actually exclude them. So basically you risk shoving people off the program who shouldn’t even be subject to these work requirements in the first place. So that’s one part of it.
The second part of it then is, basically for the populations in these programs, one, most of them work anyway. The majority do at least. Two, the challenge with the actual documentation and showing that you’re meeting these work requirements imposes much higher hurdles than we kind of realize. We have actual evidence about this. So they tried to do this in Arkansas with the Medicaid program. Basically what they found is they basically kicked most people off the program who were actually working. Now you can say, “Well, maybe they could have designed it better,” et cetera, et cetera. And that may be true, but work is just really complicated for these populations. You’re looking at people working in hourly jobs, their schedules are all over the place, they’re in and out of jobs, they get laid off. And so it’s just ends up being kind of particularly complicated to document.
So I think the bigger question is not so much whether or not people will get kicked off, who should still be enrolled. I think the bigger question is actually how many, how big of an effect, was a huge effect, huge negative effect in Arkansas. I think the question is like, is there a way to reduce that negative impact? And I’m suspicious based on what we saw in Arkansas and what we know from other programs that you’re not just going to have just tons of people losing benefits who should still be on the program.
Don Moynihan: Yeah. And I’d add to that, I think going back to this coherence point, the people who have to do work reporting requirements for SNAP are probably also being asked to do it now for Medicaid if the policy’s changed there. And so you have the same population who are just consistently having to prove that they’ve met a certain set of requirements. It’s really the difference between the eligibility for those programs. If we agree that work requirements are important, if the state can capture that information seamlessly, that is one thing, versus the work reporting requirements excluding many people who are eligible, that is quite a different thing. And that ultimately, I think, undermines the purposes of the policy. The argument against… In particular, some of the legal arguments against work requirements in Medicaid is that it runs contrary to the statutory goals of Medicaid, which is to help poor people get health insurance. It’s not encourage labor force participation.
Beyond the purposes, you also think about what are the costs. So we know with some evidence for those who are purged from the Arkansas Medicaid work requirements, that about half of them reported experiencing serious problems paying off medical debt, 64% delayed taking medications because of the cost that they encountered. And so there are real human costs that come with imposing some of these work requirements.
I think partly what makes work requirements in particular stand out is that they are popular. If you run a survey and you ask members of the public about work requirements, for anything, you will get a solid majority because people do have this innate notion that people should participate and engage and not freeload. But I think when you start to then see some of the human costs that are accompanied with them, people might be more cautious about imposing them.
Matt Grossmann: So I have a rulemaking simulation that I use with my students responding to comments, and they inevitably make it much more complicated from the first rule to the last rule. So we’ve been talking about it as if this is something that the agencies impose, but especially with your comments about public opinion, it may also be just something we decide to impose that is, is this just a product of us having a lot of contradictory goals and our first impulse to fix something like a burden might be another set of rules.
Pamela Herd: Yeah, I mean there’s clearly some truth to that. And again, going back to the debates in the ’90s about designs around new social welfare programs even in the 1990s, there was a lot of that effectively like, “Oh, but we need to make sure that this group gets it and this group doesn’t” and “Oh, wait a minute. We need to narrow it even further because what about this incentive and that incentive?” And you just end up with these incredibly complicated programs that a lot of people ultimately can’t access.
The flip side is, I do think though that we also… I mean, the flip side is we have a lot of evidence that people prefer to have more simple interactions with government themselves. So I might be willing to impose X requirement on someone else, but then if I’m the person who experiences those burdens, I’m a lot less excited about them. So I think this is to the generic 1980’s point that would come out of social welfare research during that period, which is more universal programs have the capacity to achieve a range of better outcomes. So more popular and they do also tend to be a bit simpler as well.
Don Moynihan: I think there is some evidence from psychology that humans have a tendency to add rather than subtract when facing complex problems. They’ll take the status quo the starting point and then ask to solve for a certain solution. Their first step will be like, “What else can I do?” rather than, “How can I take away to make things better?” And I think in particular, progressives are probably very inclined to see addition of requirements and eligibility and targeting and carving out as a way to expand the reach, add more nuance, provide sensitivity to different groups when it comes to the implementation of policies and administration. So I do think there is something to that.
But to Pam’s point, when you asked the public about, and we did a survey recently when we asked public about what they would like to see happen with the end of the public health emergency and asked them about different policy choices states could make to reduce hassles like using prefilled forms for example, you get really solid bipartisan majorities who would say, “Yes, make it simpler, make it work better.” And so I think that’s the challenge for policy makers is to have that in their head as they’re doing their job.
Matt Grossmann: So you all have been exemplars at integrating public administration concerns with the traditional concerns of social sciences like political science and sociology. What’s kind of the current state of that connection as you see it? Is there still kind of a separated public administration that’s geared more towards practitioners and deals with things that the other parts of the social sciences ignore? Or is there more of an integration in thinking about implementation and interfacing with government programs in the social sciences and public administration being more open to these broader political and social forces?
Don Moynihan: So I think there’s a long history of public administration in America. And if listeners are not familiar with it, if you go back a hundred years in the social science in America, public administration was at the forefront and people who were running the American Economic Association or the American Sociological or Political Science Association tended to be public administrators. And they also tended to be deeply involved in the practice of government. Many of them went in and out of government. I think at some point in the ’60s and ’70s, you saw the separation emerge where public administration started to… There’s some question of whether it was pushed or whether it jumped out of political science and either set up standalone institutions or become part of policy schools.
If I engage in some self-criticism here about my field of public administration, I think we are very good at thinking about how publicness matters to the ways in which public employees work and how that then has consequences for how the administrative state operates. I think we’re bad at placing that in the context of what are the policy suggestions to fix things sometimes. I think we’re maybe a little bit myopic and focused on our own topics. Similarly, I would say there is a lot of really great ideas from public administration that it is hard to get attention to amongst those who are in the core disciplines. And so we’re working on administrative burden, which has had, I think, some success in that area. But I could give you the example of public service motivation, which I think has been a phenomenally smart insight into why humans want to serve in government, which hasn’t gotten a ton of traction, at least not until relatively recently even though it’s been an idea hanging around in our field since the 1980s.
So I feel like I’m putting myself on the couch here a little bit. What I would say is the field is very open to outsiders who want to come in and engage with discussions of the nitty-gritty of policy implementation. And if you talk to people who actually work in government, that’s incredibly important. I think we need some of the perspective that comes from economics, from political science, sociology and elsewhere to help inform our work. Similarly, I think we have something to offer when it comes to informing those disciplines and maybe drawing on a hundred years of insights from the way in which public organizations work that you can’t just sort of start with some armchair theorizing you have to have some deeper understanding of the public setting. What do you think, Pam?
Pamela Herd: Yeah, I mean, very briefly I would just simply say I do think kind of limitation, particularly in sociology or the literature and sociology around public policy is that it hasn’t actually paid much attention to public organizations. It’s focused much… With the exception of welfare agencies around TANF and welfare reform, this might even be true in political science although, but certainly in sociology, we have not paid attention as much to effectively implementation. And I think sociology has a lot to add in terms of how we think about inequality. And so I think that’s one of the things we’re trying to do when we published on racialized administrative burdens with Victor Ray, who’s an organizational sociologist, in terms of trying to bridge those kind of two domains. So yeah, I mean, I think PA actually has a lot to offer, at least in sociology, how we kind of understand the ways in which policy produces inequality.
Don Moynihan: Let me add, if you can have those spaces, because we recently were working on a symposium with Russell Sage Foundation, which plug will come out sometime in the fall of 2023 with people from econ, from political science, sociology. When you can build those spaces where everyone has a common term of reference and are just bringing different insights and skills to the table on the same problems, it’s incredibly fun and insightful and engaging, but it takes a lot of work to create those spaces. And it takes, I think, a certain set of academics who are interested in looking beyond the narrow incentives within their field. So it’s not easy, but it’s great fun when it works.
Matt Grossmann: As you know, the next Republican administration is preparing to politicize more of the bureaucracy and does so with and have more political appointees lower across agencies. And they say that one reason that they are doing it is because the administrative state is too powerful and they want to kind of reduce its role, but obviously it has implications for the potential for administrative burdens to be increased on purpose. So how should we expect that to play out if it does come to pass?
Don Moynihan: Yeah, that’s a great question. I’m actually writing something about this now, so you’re catching me as I’m thinking about it. I think this is quite bad for the American administrative state and I think it’s quite bad for American democracy. And in particular, there have been debates around this topic for a long time. The Heritage Foundation just released their mandate for leadership, which they released every four years since Reagan administration’s basically telling the Republican candidate, “Here’s how you should govern.” And what’s different about it relative to prior examples is now that they include Schedule F as a recommendation, which is this executive order that would allow a president to wave a magic wand and converts potentially tens of thousands of career employees into political appointees, which means the president could fire them and they lose job protections. I think that would really undermine the quality of government. Political scientists who study expertise like Gilmer and Patty argue the importance of having sort of stability and a long term ability on the part of career employees to shape the government as a requisite for a functional government. So you would remove that.
The best evidence that we have from political scientists like David Lewis is that more political politicization leads to worse outcomes in terms of competence and capacity. So I think once you start devaluing expertise, you are going to worsen the quality of government in fundamental ways. But the democracy part is the more important part. I mean, if you look at what happened during the first Trump administration, the impeachment process related to Ukraine basically involved some career employees saying, “This is illegal. We shouldn’t be doing this.” And some political appointees saying, “No, this is fine.” And it turned out it was illegal and some of those career officials ended up resigning or got denied promotions. But in the world where Schedule F existed, those people would’ve been fired or they would never have voiced their concerns in the first place. So values like transparency and accountability… The government works for the legislature, it doesn’t just work for the executive branch as well. And if you sort of make them more responsive to the president, the chances are they’re going to be less responsive to legislature.
So I am not a fan. I think there are many better ways to make the government more attentive to reducing burdens both on businesses and on individuals. I think there are lots of things the federal government should be doing. I think there’s room for a broader policy discussion on how do we build the American administrative state that’s framed in more positive terms. Top of my list would be it has to be easier to hire smart people in government than it is currently. I see this from colleagues, students. It’s just incredibly difficult to get people into government quickly especially if they don’t follow the traditional, “You’re going to be here for the rest of your career” model. I think there may be cases where it should be easier to fire people, but the biggest challenge is we have a ton of people who are about to retire in the federal government. We have real skill gaps. We need to be able to hire better and more quickly.
Matt Grossmann: Anything else we didn’t get to that you wanted to include or anything you’re working on now that you want to talk?
Don Moynihan: Well, let me give a plug for the idea that we should be paying a lot more attention to rule of technology as an enabling factor and the use of administrative data as an enabling factor to making government work better in American life. I think this feeds into a broader discussion about where we think the American administrative state is going and whether it can do big things again, but those are two of the tools that we haven’t been using as aggressively as we could to enable government to actually provide value in ways that citizens appreciate.
Pamela Herd: Yeah, and just to give a super concrete example of this, the Department of Education could benefit a lot from having more access to IRS data to ensure people are eligible for varying student loan relief programs. Were actually receiving benefits from those programs for example. This is a common issue across the federal government that they really struggle for real statute reasons and I think also organizational cultural reasons. Both really struggle to share data in ways that can really significantly reduce burdens in really high impact ways. So I do think I agree with that. That’s going to be a key issue moving forward if we’re really going to improve.
Matt Grossmann: There’s a lot more to learn. The science of politics is available biweekly from the Niskanen Center. I’m your host, Matt Grossmann. If you like this discussion, here are the episodes you should check out next, all linked on our website; How Bureaucrats Make Good Policy, Why Lawyers Rule American Politics, Can Democrats Design Social Programs that Survive, How Obamacare and Medicaid drive voting, How Presidential Appointments Reveal Policy Goals and Elite Interests. Thanks to Donald Moynihan and Pamela Herd for joining me. Please check out Administrative Burden: Policymaking by Other Means, and then listen in next time.