Obamacare substantially increased American health coverage, but now some states and the Trump administration are acting to curtail benefits. Do Obamacare exchanges and Medicaid help stimulate new voters or even help Democrats win? Jamila Michener finds that Medicaid mobilizes voters, but only if it is well-administered and effective. States, counties, and even neighborhoods matter to how beneficiaries react. Vlad Kogan finds that, while Medicaid and Obamacare generally helped Democrats, a last-minute sticker price hike in 2016 moved votes to Trump. Democrats were able to win votes with new policy, but Republicans also win votes by undermining it.

Studies: Fragmented Democracy” and “Obamacare Implementation and the 2016 Election” 

Interviews: Jamila Michener, Cornell University; Vlad Kogan, Ohio State University 

Subscribe to the Science of Politics podcast on Soundcloud, iTunes, or wherever you get your podcasts.


Grossmann: This week on The Science of Politics: How Medicaid and Obamacare drive voting. For the Niskanen Center, I’m Matt Grossman.

Obamacare substantially increased American health coverage, but now some states and the Trump administration are acting to curtail or condition benefits. Do Obamacare exchanges and Medicaid help stimulate new voters, even helping democrats win? And if the policy design moves toward disarray, will it hurt beneficiaries, or even help republicans?

Today, I talk to Jamila Michener of Cornell University about her Cambridge book, Fragmented Democracy. She finds that Medicaid mobilizes voters, but only if it is well administered and effective. Otherwise, it can backfire with stigma and discrimination. States, counties, and even neighborhoods matter to how beneficiaries react.

I also talk to Vlad Kogan of Ohio State University about his paper with Thomas Wood: “Obamacare Implementation and the 2016 Election.” He finds that while Medicaid and Obamacare generally help democrats, a last-minute sticker price hike moved votes to republicans. Democrats were able to win votes with the new policy, but republicans could also win votes by undermining it.

Let’s start with Michener’s book, an in-depth at how Medicaid shapes politics. It matters how it’s designed.

Michener: The big takeaway from the book, Fragmented Democracy, is really about the kind of democratic consequences of Medicaid, this large government program that is aimed at providing healthcare for folks who are in the most need. And while we wouldn’t really think about a healthcare program, necessarily, as having much to do with democracy, what I find in the book is that it does, and that experience is what this program can affect the way that people engage government and politics, and whether they participate; and that the structure of the program, especially over a place, the kind of geographic structure of the program, and how that connects to the way that it’s designed, and the benefits that people are able to receive is really what drives that relationship between Medicaid and political participation.

Grossmann: Kogan and Woods’ paper is a close analysis of Obamacare’s effect on the 2016 election. They find Obamacare helped democrats, but not if the price went up.

Kogan: The main question that motivated us is what impact did Obamacare have on the 2016 election? And so we wanted to know, what was the impact of people getting health insurance, and what was the impact of the prices they pay for health insurance. And I think our finding is yes, healthcare and Obamacare mattered. People gained insurance were more likely to vote for democrats, and people who had to pay a lot for their insurance were less likely to vote for democrats. So those who gained rewarded the democrats, and those who maybe gained, but gained really unaffordable insurance, tended to kind of punish the democrats and vote for republicans instead.

Grossmann: Michener is raising a connection that others haven’t made, and clarifying when policy feedback happens.

Michener: On the broadest level, it’s just not a connection that a lot of people are thinking about. And even as the book has been published, and I’ve had opportunities to engage with health policy people and health advocates; even though they’re in the realm of health, and they’re dealing with these issues day-to-day, they often say, “It never occurred to me that this might have something to do with political participation.” So I think on the broadest level, that just making the connections is a contribution.

Michener:But for political scientists, it’s a little bit more nuanced than that, because there’s a whole group of political scientists who study these processes that we call policy feedback, and the point is to think precisely about this: How do public policies feed back into our politics by affecting the way that people behave politically, or the way that they think politically?

Michener: And then we can think about that in terms of how it affects regular people like Medicaid beneficiaries and all sorts of other political actors. But there are political scientists who have found these feedback effects for different programs; not necessarily Medicaid. The sort of biggest studies are around cash welfare programs, around Social Security, around the GI Bill. And those studies find that policies can affect people’s political participation.

Michener: So that part of this isn’t new. I think the specific focus on Medicaid is relatively recent. It’s really myself and only a handful of other people have paid attention to this program. And the focus on geographic variation, how Medicaid is a program that’s very different in different places, and that that structures its political effects is really the contribution, the unique contribution in the book that you really won’t find in many other places.

Grossmann: Kogan said he’s overturning the conventional wisdom that people dislike Obamacare, but finding evidence more consistent with vote-buying.

Kogan: One is that Obamacare was a total failure, people hate it, it’s really unaffordable, people are paying more. And I think that’s not consistent with our evidence. Right? If people hated it, they wouldn’t have rewarded the party that was behind the policy. So I think the voting behavior is inconsistent with the idea that voters; at least the votes who were personally affected, really dislike this policy. So I think that’s kind of one big takeaway for me.

I think the other big takeaway is, as you mentioned, Mitt Romney in 2012 had the undercover video where he said, “Half the country won’t vote for us because democrats give them welfare benefits.” I think our evidence is somewhat in line with at least that idea that you can earn the loyalty of voters by making them better off, including through the provision of subsidized health insurance; but I’ve looked at my other work on things like food stamps, and I think it’s pretty consistent with that policy as well, that poor people who are really disadvantaged, if you enact policies that help them, that’s an effective way for you to win more votes, which is, I think, consistent with that conventional wisdom. I think I wouldn’t necessarily interpret it in a negative way, in the same way that I think Mitt Romney did, but I think at least it’s consistent with that story.

Grossmann:Michener’s study came about from her interviews with low-income voters.

Michener: I kind of stumbled onto Medicaid as an issue as far as my focused was concerned. When I was in graduate school and I was thinking through writing my dissertation, I had one member of my dissertation committee who was not a political scientist, who was a sociologist, in fact; who said, “You know that you care about these particular groups of people and how they’re faring in the political system. But instead of kind of making up a research question that you think is relevant to them, why not go talk to the folks whose you’re interested in understanding and explicating, and see what really matters to them? And focus your research on that.” And I thought, “Hmm, that’s an idea.” And I started doing interviews with low-income people all across Chicago, which is where I was at the time, because I was in graduate school there

And I was just talking to them about what matters in their lives, and how they think the government figures into their lives. And many people brought up healthcare. Many people brought up Medicaid. It wasn’t something that I was thinking about a ton before that, but I sort of went to communities and went to people, and they told me that it mattered for them, and so I reoriented my research at that point, and really started to focus a lot more on health policy and Medicaid, specifically.

Grossmann: She also learned about extensive geographic variation from those interviewees.

Michener: Talking to people is always, in my view, really informative; and brings to the floor things that we might not have otherwise understood. So one of the big things that I focus on in the book, sort of geographic variation in Medicaid benefits and policy design, and how that matters for people, is not something that I came to the project thinking about. I sort of came to the project thinking either I’m going to find that Medicaid is actually a boost, something like Social Security or GI benefits, because it’s healthcare, and it’s something that people are going to value, and they won’t feel like it’s stigmatized.

Or, I thought, I might find the opposite; which is people sort of treat Medicaid like cash assistance, like what we think of as traditional welfare. And to the extent that it’s stigmatized, it might be bad for political participation; it might demobilize people. But I thought I’d find one or the other of these stories, and I thought that the interesting question was, which story would I find.

But when I started to interview people, I knew enough to know Medicaid’s different in different places, so I guess I’ll interview people in different places. And when I started to talk to people, what they were talking to me a lot about were those places. They were saying, “In Michigan, it’s like this; but when I lived in Wisconsin, it was different.” Or, they were saying, “In this neighborhood, it’s like this; but when I lived in the suburbs once, it was different.” Place just kept coming up again and again, whether it was states or counties or particular neighborhoods. And it wasn’t what I expected. It wasn’t what I brought with me when I went into the project, it was what people told me.

And even as I drilled further in, and I started to think about, “Well, which aspects of policy design might matter?” There are probably 500 ways that Medicaid is different across states and localities. How do I know which of those ways might matter for affecting people’s political participation?” Well, I knew because people told me, and their stories were both consistent in the sense that different people were telling me very similar things, and they were nuanced in the sense that people had really well thought out connections that they were making between how they understood the world politically, and what they had experienced with Medicaid. And I realized many of those insights, most of those insights, I wouldn’t have gotten unless I sat down and asked people about what was going on in their lives.

Grossmann: Kogan was motivated by the old finding that voters usually care about recent performance.

Kogan: You know, what got me interested and my co-author interested originally is … As I said, there is this literature on pocketbook voting, and on performance voting, generally, that says in an election year, voters tend to re-elect the incumbent party when things are going well, and it’s really, really recent performance that matters more. They don’t look over the whole term, they don’t look over the whole four years, they really look over the past year, or even the past six months.

And so I was familiar with this research in 2016, as I was just following the election. And although I think most people don’t remember anymore, in August of 2016, so about three months before the election, there was national headlines because a lot of insurers were leaving the Obamacare exchanges. They were losing money, they were shutting down. And so I think a significant number of counties, there was just one company offering just one plan. So there was not a lot of choice. And so that gained a lot of attention, and it got me thinking, “Well, if performance matters, and performance right before the election matters, this is a really salient performance metric,” right? It’s really telling voters that maybe Obamacare isn’t working.

And so the day after the election, after Trump won it, I talked to Tom, my co-author, and I said, “Wow. So was this Obamacare? We got to find out.” Right? Did Trump win because of Obamacare? So that’s what kind of prompted our foray into this project. We wanted to really understand the 2016 election and understand, were it not for Obamacare, would it have turned out differently.

Grossmann: That led him to focus on prices in the Obamacare exchanges.

Kogan: I think some of the most interesting, most convincing evidence is really not on the Medicaid side, but on the exchange side and on the prices side. I think we have really good evidence that voters were responding to premium prices of their health insurance, including in the ways that we think was really not rational. They were responding to the sticker shock, to the sticker price, not to the actual price they were going to personally pay.

Grossmann: In health policy, they both say the decentralized and complicated structure matters. Michener says Medicaid empowers the states with big impacts.

Michener: Medicaid is an intergovernmental program. It’s a program that is the responsibility of both the federal government and the states, and in a variety of ways, also localities. And it provides predominately low-income people with health insurance. Right? But the folks who receive health insurance through Medicaid are not just low-income, but they’re often also people who have a variety of medical challenges, who are disabled, and who have other challenging life circumstances.

So 10 percent of Medicaid beneficiaries are low-income elderly people. Many beneficiaries in many states, approaching 50 percent are really children, low-income children, whose families, for a variety of reasons, can’t afford to provide health insurance to their children. And often, other low-income beneficiaries are working, or sometimes non-working adults with or without families, who just don’t make enough income in order to provide themselves with health insurance either through their workplaces, or otherwise.

So we’re talking about folks who are kind of the most economically disadvantaged. Medicaid is also racially disproportionate, so that the number of blacks and Latinos, for example, that rely on Medicaid is really disproportionate in comparison to their existence in the population. And that’s largely because poverty is racially disproportionate. But the crucial thing about Medicaid that I focus on in the book is that it’s really different in different places. Not all social policies work this way. Some policies, like Social Security or Medicare, policies that we think of as more universal, where everyone in a particular category just gets the policy, those kinds of policies are largely the same in different places, because they’re centrally administered. Right? The national government says, “We’re basically going to do all the work, here.” States might do some administration, but there’s a lot of commonality across place.

Medicaid is part of a group of policies, and this is how most anti-poverty policies are structured, that are very different in different places because they’re not centrally administered. They’re administered by states and localities. And those states and localities are purposely, by design, given a lot of choices, and they’re given choices about how to design the policy, what to cover, what not to cover, who’s eligible, who’s not eligible, and how to administer the program. So Medicaid, a lot of the people that I interviewed says, “Medicaid isn’t one program. Medicaid is 50 different programs,” because it’s significantly different in different places that it’s not necessarily the same program.

And so if you’re a low-income mother who’s giving birth, there are some states where it’s really likely that you’ll be able to have healthcare coverage, and in other states, where it will be much harder. And some states where, when you do have your baby, you’ll be able to stay in the hospital for several days after, and others where you’ll need to be gone within 24 to 48 hours. And so that matters for you, it matters for your baby. And those kinds of differences carry across the life’s course, down to hospice care. If you are a non-elderly adult who has a terminal illness and you’re dying, and you’re very low-income, there are some states where you cannot get hospice care, and you won’t be able to receive the kind of care that will allow you to die a pain-free death. And there are other states where you will.

And so in almost every way that we can imagine, the kind of care that you can receive through Medicaid is going to depend on who you are, categorically and where you live.

Grossmann: In addition to Medicaid, Kogan says, “Obamacare relied on complicated exchanges making price increases salient despite their lack of real impact.”

Kogan: Obamacare, I think, really has kind of two ways of providing healthcare, subsidized healthcare to poor people. One of those is Medicaid and it expanded the number of people and the type of people eligible, and the idea is if you’re really poor, if your income is below 138% of poverty, the government will basically give you free healthcare. If you’re poor but not quite that poor, so between 100% and 400% of poverty, then the government will ask you to pay a little bit, but then will pay the rest. They will subsidize you and help you in buying private coverage.

Those are two different coverage pathways, both for low income people, but depending on how low income, and so that second pathway, the exchanges, are really an online marketplace that was set up to create standardized health insurance policies, to create more competition in the health insurance market, to set some minimum benefits to make sure that people were buying policies that would actually cover them, and essentially under the law if you’re poor enough, again which is four times poverty or less, the government says you will pay what you can afford and then we’ll pay the rest. As long as you buy a standard policy, I will pay the rest. If you want to buy kind of a Cadillac policy, then you’ve got to pay the difference, but we’ll make sure that you can get insurance that you can afford. You have to pay a little bit towards it, but we’ll pay most of the cost. If you are a person that’s buying health insurance through these online marketplaces, poor enough you can get a subsidy and in 2016 the vast, vast majority of people, like 85% of the people buying on the exchanges, weren’t paying the full price. They were being subsidized by the government.

Every year the prices, the premiums you paid, changed depending on the experience of insurance companies. If a lot of sick people sign up, it costs them more, so they’ve got to raise prices the next year, but the nice thing about this policy is if prices go up the subsidies go up. Because 85% of the people were low income, they had already maxed out their out of pocket costs, so even if insurance premiums went up the next year it really wasn’t going to cost them anything. The government was going to pick up the tab, so I think that’s where a lot of the interesting findings in our paper come from because we showed that people who were really maxed out and wouldn’t have had to pay anything extra were still responding to these price increases that were being publicized in the press, that were being publicized by the Trump campaign in a way that really is from an individual point of view not rational, right?

Yes, prices were going up. Yes, maybe the exchanges weren’t working exactly the way they should have been working, but from the point of view of an individual person with health insurance none of those things really matter because your cost, your insurance is probably going to stay the same. Most of that cost is going to be borne by the federal government, and yet people responded as if they were personally going to be paying the difference.

Grossmann: Mitchner focuses on turnout, finding that more generous Medicaid and good administration promotes participation.

Michener: The largest source of variation really has to do with the generosity of the program than the benefits offered, right, so when Medicaid beneficiaries are in places where benefits are being rescinded, right, where eligibility levels are being pulled back on or the scope of benefits, for example, are being limited, so states that say, “We’re having budgetary problems. We’re going to stop covering dental benefits.” This happened just last summer in Kentucky or, “We’re no longer going to cover podiatry,” so that if something happens with your feet that’s really on you. Maybe some states will do this, they’ll say, “How about we won’t cover vision?,” and so you sort of have to figure out on your own how you’re going to pay for your glasses or your eyecare, those kinds of decisions about which benefits are going to be covered and which aren’t, which states have a lot of control over.

There’s a basic set of benefits that all states have to cover, a kind of floor below which no one can fall, but that floor is pretty low and lots of things above the floor are optional. States’ decisions about which of those optional things they’re going to provide and which they’re not are hugely important for Medicaid beneficiaries. It directly affects the ability, their ability, to have their needs met and it’s noticeable to them. It’s salient in their lives.

If you need dental help and you can’t get it, you know that, right? There are people I talk to in the book who talk about having to take out various teeth because of health issues that they had, but realizing that getting dentures, in order to have teeth back in your mouth, was not covered by Medicaid. Medicaid would cover the medical procedure that removes your teeth, but they wouldn’t cover the dental procedure that would allow you to get dentures, and so you’re walking around without teeth in your mouth. That’s going to affect somebody’s life in a pretty profound way and they’re going to know that the reason why they don’t have those teeth is because this government program doesn’t cover it.

Those kinds of very obvious choices about which benefits to provide and which benefits not to provide, those are policy design choices that states make based on budgets and based on lots of other calculations, but they really matter for how people experience the program and how they therefore behave politically in response to those experiences.

That’s one set of things that matters, but also what’s going on kind of administratively on the administrative level matters too. One of the things that I find when I look at larger quantitative data is that the ratio of the number of workers that states have in sort of their welfare bureaucracies to the number of people living in poverty that they have really matters for shaping the relationship between Medicaid and political participation. The logic is that when states are really strapped as far as bureaucratic resources Medicaid beneficiaries go into offices and they experience those offices differently. Their ability to get guidance, to get help, to have bureaucratic burdens lightened is much more limited when states are strapped and they’re employing relatively few people given the demand.

Grossmann: Neighborhoods matter as well, even within the same jurisdictions.

Michener: People interpret Medicaid in relationship to the institutions where they’re receiving their Medicaid benefits, so places like local healthcare clinics, public hospitals, and local doctor’s offices, but those places are also places that are deeply marked by place based inequalities, right? Urban inequality, rural inequality, those kinds of inequalities are very clear when you go to a health clinic, right? If you go to a health clinic in a struggling neighborhood in Detroit, it’s going to be really different than what it’s like if you go to a health clinic in a middle class or upper class neighborhood in Ann Arbor. Now, both of those clinics are in Michigan, but the experience that you have will be completely different. How long you wait, the quality of the doctors, the way that people talk to you, even what you’re seeing as you walk into the clinic. Is there someone shooting up drugs in front of the clinic as you walk in? If people experience those kinds of things in the course of receiving their Medicaid benefits, they start to associate those experiences and connect them with Medicaid, and so the context in which you receive benefits shapes the way you think about those benefits and the way that you experience and make sense of the program and the government that is providing you with the program and the benefits.

Grossmann: Counties matter, especially for addressing problems.

Michener: If you have a problem with Medicaid, and one of my chapters focuses on this, and you want to file an appeal or have a fair hearing because your benefits have been cut or you’re not being given something that you should have been given, often counties are adjudicating those procedures, those administrative fair hearings, and so that’s a big part of the process that they can shape as well.

In the book I look at whether counties matter for affecting the frequency of fair hearings and the sort of likelihood that people will initiate those kinds of hearings, and I find that they do. Not in any easy or straightforward ways, right? It’s really complicated because then you’re getting layers of geographic variations, and so you have states that are really varying in many ways, and then counties within them, and then neighborhoods within them, and so at each level you have to think about the level itself and then you have to think about the kind of cross level dynamics. It ends up being pretty complex, but all of it is significant and all of it really … I think the larger argument I make in the book includes a set of things that we should be thinking about as scholars when we think about policy feedback, especially for policies that vary a lot across place.

Grossmann: Kogan and Wood focused on the more recent Obamacare changes, looking at both aggregate voting and individual level vote choice.

Kogan: In the aggregate level analysis we’re essentially comparing counties over time and we’re asking, in places where more people gained health insurance either through the Medicaid expansion or through the exchanges, did Clinton underperform Obama in the previous election or two or did she over-perform? Then in places where premiums are going up the most between 2016 and 2017 did Clinton underperform Obama or did she over-perform? That’s kind of the nature of the analysis, controlling for basically everything else that you can possibly might affect the election. In some ways I think that’s a pretty credible design in terms of speaking by causality, but I think the big limitation was that if we do find differences, which we did find, we can’t be certain that those are differences driven by the voting behavior of the people who were themselves affected. Is it people who have Medicaid or is it their friends, or is it the healthcare providers who are now making more money because they’re not providing uncompensated care? Looking at the county level, we can’t. It doesn’t untangle all those mechanisms, so that’s why looking at individuals is so important.

That’s what we can do with survey data, but again the survey data has its own limitations. Again, very hard to measure Medicaid eligibility and it turns out, something I really didn’t think about but I think is a huge problem, that people have no idea what kind of health insurance they have. For our purposes and really for any study trying to understand the politics of healthcare, we’ve got to have really good measures of peoples’ healthcare coverage, health insurance coverage, and it turns out a lot of people don’t know. They answer these questions with a lot of error.

Grossmann: They were able to show that people changed their mind in 2016 when the premium increases were made public.

Kogan: For the exchanges I think the evidence is consistent across both sets of analysis. Places where people were more likely to gain health insurance through the exchange, and as long as that health insurance was reasonably priced, they were more likely to vote for Clinton and we replicate that in our individual level analysis. Then in places where premiums were going up a lot that Clinton premium kind of shrank, and then in places where it was going up really a lot it became a Trump premium. Again, we replicate that with the aggregate and individual level data, and the individual level survey data I think is really nice because we’re using a panel survey that asks people in October. Before October we’d ask them about their health insurance. In October they asked them, “Who are you going to vote for?,” and then in November they asked them, “Who did you actually vote for?,” and the insurance premiums were announced right in between, so we can actually see how people changed their minds depending on where they lived, depending on how much premiums are going up in that area, and depending on whether they were personally affected by the premiums or not, whether they were buying through the exchanges or not.

Kogan: I think for my money that’s the most compelling part of our analysis. It’s that we can, I think, really cleanly isolate that causal effect by comparing the same voters before and after the premium information is public.

Grossmann: The election’s timing mattered and so did political sabotage.

Kogan: The election coincided right with the time when people had to make decisions, so it was right when the news media were covering things. I think the second issue was premiums were going up for a lot of reasons, but they were going up because there was a lot of policies enacted really at the state level that made the health insurance market not work as well as it could. Especially in Republican states, Republican governors, and Republican state legislators, really enacted policies that aptly sabotaged the individual market and actively drove up the premiums. They, for example, let healthy people stay out of the exchanges and keep their old plans, which made it more costly for everybody else to buy health insurance. All of that, I think, made premiums go up much more so than they would have in the absence of this kind of sabotage.

I think the interesting part of our finding is who did voters blame, so we had a lot of policy decisions, we have this complicated implementation process where part of it is done at the federal level, part of it is done at the state level, part of it is done by Republicans who don’t want this policy to succeed, and then voters just see the end result. They have to make sense of this entire mess and figure out if I’m happy who is responsible, if I’m unhappy who do I blame, and I think in this case it’s this interesting dynamic where Republicans did all they could to sabotage Obamacare. It kind of worked. Premium prices went up, and then voters didn’t punish them. They punished the Democrats because they said, “It’s Obamacare. Obama must be responsible.

Grossmann: Kogan says he agrees with Mitchner’s finding that sub-national variation in policy matters a lot.

Kogan: I think both her findings and ours, I think, speak to the fact that state level barriers really matter. Again, it’s not a surprise to economists that study take up of health insurance programs. I think that’s a well known finding and I think we’re both kind of taking that finding and saying what’s the implication of that for politics. I think that’s kind of the big takeaway, that you can really affect how people experience these programs. For Medicaid you can really affect their personal experience. For people on the exchanges you can really affect the prices, even though they’re not going to be paying the prices in a way that has fallout, right, in a way that has downstream consequences. Downstream consequences for whether people actually get health insurance, whether people keep health insurance, but also downstream consequences for whether people vote and how they vote.

Grossmann: Mitchner also thinks Medicaid is likely to have effects on vote choice, but she says it’s likely to be a lot more complicated.

Michener: The effects on vote choice are also an important outcome. As you’ve said, I’ve done some work and I’m glad that others are doing work, really thinking about that, so not just turnout. Who is participating or whether people are participating, but also who they’re voting for when they show up and what the implications of that are for who our elected officials are going to be. I think that that’s relevant and it’s related.

I think that brings in a few additional dynamics. One of the dynamics is partisanship, right, and dynamics around partisanship. I think that complicates things a little bit because people can be experiencing a policy in a particular way and the way that they understand, that they make the political connections between those experiences and how they should be behaving politically, whether they should be voting, who they should be voting for, those kinds of decisions are going to also be mitigated by partisanship. When we start thinking about vote choice, that’s going to be a lot bigger of an issue, and so it’s good to focus on that because it brings in a really crucial aspect of our political system that I don’t focus on as much in the book.

Michener: I also think though that it gets a lot more complicated, so in order for either your experiences with the policy or the design of a policy to affect your vote choice. You have to be able to know or have a sense of who is responsible for that policy and even know to connect that policy to the government as an entity. This is one of the reasons why, for example, the marketplaces and thinking about that aspect of Obamacare, it can be a lot more complicated and it’s really just different than the sheer focus on Medicaid. People generally have less information about the marketplaces. There are kind of more submerged policy that people may or may not recognize is directly connected to government and even if they recognize that it’s directly connected to government, they may not correctly infer which political party or which political official at what level is responsible.

Grossmann: Her lessons for Medicaid may travel to other policies, but perhaps not all.

Michener: The most basic lessons of the book that when we’re dealing with policies that one, are very different across place where states and localities have a lot of discretion and two, tend to be targeted to more marginalized populations, and populations that otherwise maybe folks aren’t paying as much attention to and or they don’t have as much political power. I actually think Medicaid is the best case scenario on that metric because Medicaid beneficiaries are such a motley diverse group, right? You do have, for example, because many, even middle class people end up relying on Medicaid as their longterm care option, so there are plenty of people who were middle class their whole lives and simply spent down their options and now are the resources and now Medicaid is paying for their nursing home care. And those people have families and other folks around them who are middle class. So there’s a significant middle class constituency that really favors Medicaid and that knows people, whether it’s your mother or your grandmother or your aunt who are Medicaid beneficiaries.

So in that sense, Medicaid has an advantage. It’s more salient and more broad of a program than many other antipoverty programs. I would actually expect that these place based dynamics are going to be even more acute in programs that are antipoverty programs. But that vary a lot across states and so I expect that, that the level of variation and the extent to which place matters is going to be heightened for many other antipoverty program. Some of the programs that are more universal, right, where it’s free college for everyone, but states are going to have a huge role in structuring what that looks like and how it’s implemented. We should expect that state level variation to matter, especially with regard to the how such policies will feed back into politics, but I think perhaps not quite as much when the target populations are either more universal, a larger swath of people and or when the policies themselves are not targeted towards or focused on kind of the most marginal members of the polity.

Because those folks are going to have some political clout or be able to carry some political weight anyway and that will shape the nature of the programs, and to some extent maybe limit the differences between the programs across place.

Grossmann: Both agree on the key lesson that it’s hard to design policies that create positive political feedback.

Michener: One of the things that’s really struck me is how tricky it is to design health care policy that is going to build power and not erode it. So I think that the major kind of thing that I would tell anyone who asked me was that as you’re thinking about your policies, there are a set of policies right now that you know are sort of gaining a lot of popular momentum. And you know, I’m a big fan of many of them, not as big as others, but one of the things that I would tell anyone who was kind of thinking through these policies and how to evaluate them and how to think about which ones were most and least promising is not just to think about popular opinion or think about the kind of specific dynamics, the fiscal constraints and all of those things really matter.

But to think about the potential politics of these policies, what kind of political groundwork will any health policy that you’re considering lay in a way that affects the political power of important constituencies in our democracy going forward. And so you know, things like policies for example, that proliferate more state discretion without any kind, not that state discretion as bad. Sometimes it’s great, right? Sometimes it works well, sometimes it doesn’t. But the key is to think about, how do we think about the kinds of inequalities that might be created across states? How do we minimize those? How do we build in policy backstops to prevent those kinds of inequalities from getting out of control or having to negative effect. I also think it’s important to think about how any health policies that we’re proposing right now affect the landscape for policy advocates who are doing the work of trying to see these policies through.

One of the things that really stuck out to me when I talked to policy advocates with respect to their perspective on the ACA, was that most health policy advocates like the ACA and thought that it was good and that it was a positive step forward and that it was going to represent an important material, a material benefit in many people’s lives. But as far as what it actually did to their world it brought in a lot of additional challenges. And so some of the agendas that those policy advocates had pre ACA, they really had to change post ACA because now they were fighting over ACA related things. And so policies that get introduced, even if they have net benefits that, that folks understand or perceive change the political context and environment for lots of different actors and can make things really difficult.

So now for example, we’re fighting over work requirements, which was not an issue that emerged before. Now there’s no way really that the folks who wrote the ACA could have predicted that. But I think it’s worth at least trying to think through the various pathways that a policy might take and understand how those pathways might affect future politics.

Kogan: Health care policy is really complicated, and so voters don’t always get it right. They observed the end result. They know where they’re better off than they would have been, or at least they know whether they’re better off than they could have been but linking that to specific policy decisions, and linking that specific actors in government who are responsible I think they don’t always get that right, which again affects the incentives, right? If you are an elected official and you make people happy but don’t reward you, then you probably are not going to win a lot of votes by making them happy. If you were elected official, that’s doing a lot to make people worse off, but they don’t know that you’re responsible, they punish your opponents, then that creates a perverse incentive, right? You get rewarded for doing a bad job.

I think Obamacare implementation is a great example of that. Republican state legislators, Republican governors who sabotaged Obamacare probably helped their party win more votes than next election, which is not the kind of incentives we want democracy to create. But I think in this policy area, this is the kind of incentives that we see. So I think that’s one lesson. You know, just because you’re making people better off doesn’t mean that, that’s going to be a vote winner for you because voters don’t always kind of see through the policy implementation process. And I think the second one is that voters don’t really understand these policies very well.

They did well, and they often react in a way that’s irrational, and again, I would point to the premiums people were responding to the sticker price and even though they really shouldn’t have it wasn’t going to affect them directly. So making people better off versus people understanding that they’re better off and rewarding you. Those are two very different things, and I think it’s important for like the official state kind of realize that just because you make people better off doesn’t mean that they’ll realize that you did it and B, reward you for it.

Grossmann: And they agree on the new political risks the Medicaid expansion mattered missioner says, but that could change as policy is restricted.

Michener: We have had medicaid expansion and there’s great research by scholars like Jake Hassles Worth and Michael Sanchez and Joshua Clinton. That suggests that those Medicaid expansions actually led to a boon and political participation. And so at least in the short term, in the immediate wake of Medicaid expansion, it looks like more people are voting. The question of sort of who they’re voting for, who that’s going to benefit, how it’s going affect the broader dynamics of our political system. I think those are questions that still haven’t been fully answered and that as scholars we’re sort of working through and need to continue thinking about.

I think that the largest kind of directional issues in terms of where we’re going, that I have spent my time thinking about and have sort of questions and ideas about are really around two things. One is the trajectory of the Medicaid programs specifically and how that trajectory is going to be effected by the new politics emerging around Medicaid expansion. And so states that are expanding but they’re doing so both with work requirements and with other kinds of requirements that are contingent on the behavior of Medicaid beneficiaries, healthy behavior programs or behaviorally oriented programs. You know, the challenge and the risk with those kinds of policies is that they’re really doing more work than what we imagined and part of the work that they could be doing. And I have some research projects right now that are kind of thinking about and assessing other, this is really going on is shaping the way that people think about Medicaid such that they think about Medicaid more like they think about cash assistance programs than they do think about social security and Medicare.

So levels of support from Medicaid are relatively high and that’s one of the reasons why it is a difficult program to tinker with too much without some political consequences because the public more broadly, even beyond beneficiaries themselves supports Medicaid. And one of the things that work requirements and other kinds of what some people might call punitive or social control elements of Medicaid policy might do is start to shift the way that American’s writ large are thinking about Medicaid. And so they’re thinking about it as the kind of program where we need to make sure these people are working, we need to drug test these people, we need to make sure these people have healthy behaviors and the focus is much more individualistic and much more about the deservingness of the beneficiaries.

Grossmann: Trump is also reducing the laws of effectiveness, but Kogan doesn’t know who voters will blame.

Kogan: It’s going to be interesting to see how voters update their beliefs, right? So we know after 2016 the Trump administration has basically done all it can to undermine Obamacare. I think on the assumption that hey, if voters don’t like it, they’re going to blame Obama and I think that’s an empirical question, right? Our voters are going to realize going forward that maybe Democrats are not responsible for some of these issues and now they’re going to correctly attribute credit and blame. I think that’s an open question. It’s worth and I look forward to studying it and I think the other interesting question is again about sticker price versus actual price. And so the president Trump, when he came in, one of the first things he did is he got rid of the cost sharing reductions, which is really complicated part of Obamacare that essentially added more subsidies for low income people.

And the direct effect of that was to increase the sticker price. But in a way that kind of backfired because of the way the subsidies are structured by increasing the sticker price, it actually made insurance more affordable for a lot of people. And so again, here is going to be an empirical question. Did voters respond to this and how do they respond? Were they responding to the sticker price, which is you know, things are bad. Or were they responding to their personal price, which was actually they were made better off by these changes. And so yeah, I think that’s another case where what prices are voters paying attention to you? Are they reacting rationally? Is another area where this can play out in an interesting way.

Grossmann: Kogan is now looking at what happens when you take away health care.

Kogan: Here we’re looking at people who gained health insurance, and I think the next question is, well how do people react when you take away their health insurance? And I think that has clear implications for things like work requirements for Medicaid. And so we do have at the state level in the early 2000 several cases of just really large mass scale Medicaid, dis enrollment. And so I have another paper that I’m finishing up now that looks at the political consequences of that and asks do voters punish elected officials who take away their health insurance? And the answer as well kind of depends. In some states they do and some sites they don’t. And it’s kind of a mess to make sense of. But in some ways very interesting and intriguing results.

Grossmann: Michener thinks there might be effects of even talking about the policy differently. She’s also looking at effects on more specific vulnerable sub populations.

Michener: I’m also interested in how policy discourse might matter. So the fact that we’re talking about Medicare for all the fact that we’re talking about Medicaid work requirements, the work requirements are happening in some places. They’re not happening in others, even in the places where they’re not happening, but they’re up for discussion. That kind of discourse matters for how we think about a program. And so I’m trying to sort of get a handle on that. And the other thing that I’m working on as much more policy specific. One of the things that I realized in the course of writing the book is that there are a lot of really specific health care… our health constituents, people in need of particular kinds of help in the realm of health policy who we may not think of.

So justice involved populations, people who are in prison or who have just been released from prison. A lot of times those folks have very specific healthcare needs and there are very specific policies that can shape your access to health benefits, for example, while you’re in prison or immediately after you leave prison. Immigrant populations who are largely not covered by programs like Medicaid, and so states have to make a decision how they’re going to treat that population. And there are many other populations that fall into that category. Former foster care children, people with substance abuse disorders. I call this health policy at the margins. These folks generally don’t have political power.

Grossmann: There’s a lot more to learn. The Science of Politics is available biweekly from the Niskanen Center and on iTunes. I’m your host, Matt Grossmann. Thanks to Jameela Michener and Fled Kogan for joining me. Please check out their work and then listen in next time.

Photo credit: Fibonacci Blue from Minnesota, USA [CC BY 2.0 (https://creativecommons.org/licenses/by/2.0)]