Partisan election administrators don’t tip the scales

In the aftermath of the 2020 election, local election officials became objects of unfounded conspiracy theories and attacks. But local clerks, even those elected in partisan elections, do make and implement key decisions about voting opportunities and election procedures. Do they tip the scales to favor their party? Daniel Thompson finds that electing a Democrat vs. a Republican as a county clerk does not affect subsequent election results or turnout. Thompson says reasonable concerns about the partisan effects of election law changes often do not materialize in real advantages.

Guests: Daniel Thompson, UCLA

Studies: “How Partisan is Local Election Administration?”

Transcript

Matt Grossmann: Partisan election administrators do not tip the scales, this week on The Science of Politics. For The Niskanen Center, I’m Matt Grossmann.

In the aftermath of the 2020 election, local election officials became objects of conspiracy theories and attacks. Reviews found mostly commendable work under pressure, but local clerks do make and implement key decisions about voting opportunities and election procedures, and the US system is weird for electing them mostly in partisan elections. You don’t have to be a conspiracy theorist to wonder whether putting powers in their hands might tempt them to benefit their party’s candidates, but that doesn’t mean they do.

This week I talk to Daniel Thompson of UCLA about his new co-authored American Political Science Review paper, How Partisan is Local Election Administration? He finds that local partisan county clerks do not affect subsequent election results or turnout. Even those that barely win elections over their partisan opposition do not use the office to advance their party’s cause. Democratic and Republican clerks don’t have large effects on turnout or policies.

Thompson has also done other work showing that concerns about the partisan effects of election law changes often do not materialize. That is some good news as we head into another election season where election administration will be under a lot of scrutiny. I think you’ll enjoy our conversation.

So your new paper looks at the influence of local partisan clerks on subsequent election results, on turnout, and on election-related policies. What are the main findings and takeaways?

Daniel Thompson: Yeah. So thanks so much for having me. I’m excited to be here and to share this work. It’s joint work with two excellent graduate students, Joshua Ferrer and Igor Geyne, who have contributed greatly to this project, and I’m glad to be here representing the three of us.

As you said, the project studies directly elected partisan county clerks who are responsible for running elections in about half of the country. We find that the typical clerk is not using their authority to provide their party a substantial advantage. So we might be worried that these clerks, with the authorities they have, are using those authorities to give their party a boost, and we don’t see much evidence for that.

Matt Grossmann: So let’s take a step back and ask what these county clerks do and how unique this US system is and why it might be controversial.

Daniel Thompson: Sure. So these county clerks, they were a new, interesting institution to me to try to dig into and understand what it is they do. That’s been something I think many of us have learned a bit about over the last few years as election administration has become increasingly a hot button topic.

And so, just to give you an overview of what it is that they’re doing, I think we bucket into before election day, on election day, after election day. So the way to think about it is before election day, they have a bunch of tasks that they have some responsibility for, though it’s generally a little bit lower grade.

So these are things like they need to manage registration, manage voting lists. If somebody passes away, oftentimes they have to remove them from the list, these types of things. They’re responsible for hiring and training poll workers in the run up to the election, purchasing and maintaining equipment, citing polling places. These are all things they have to do to prepare for election day.

Then also before election day, but related to the process of voting more directly, is they often are responsible for early voting, convenience voting methods. So these are things like also citing and scheduling early voting in person, administering absentee voting, making sure that the absentee ballots or mail ballots reach the people who need them, and staffing all of those operations as well. And so, they often can make decisions around exactly how many hours to run those, where to put them, all of that kind of stuff.

Then, finally, they have some authorities on election day administering the election. This is, again, related to staffing, running the in-person election day, but also doing things like counting votes, handling provisional ballots, even in some cases certifying the results. And so, that captures the broad scope of types of things they could do. There are many things that I’ve not mentioned there that some clerks do, but it gives you a big overview.

The key thing to note is that clerks are local officials. And so, their responsibilities, while I’ve just described many, many things that they could do, not every clerk does all of those things. Some clerks only do some smaller share of those things. This varies from state-to-state. It also even varies from county-to-county within the same state.

And so, we spend a bunch of time in the paper trying to understand what those things are and isolate, make sure we’re not leaning too heavily on one set of institutions. But the big picture is that they have broad authorities to manage and administer elections at the local level.

Matt Grossmann: Yeah. So to a foreign audience, this sounds crazy that we have Republican and Democratic elected officials taking on these tasks. So how unique is the US system and why is it controversial?

Daniel Thompson: Yeah. So in terms of its uniqueness, the three of us are American politics scholars. So we’re relying here on just what we have read from other comparative politics scholars. But our impression generally that is quite unusual, there are certainly a handful of cases of institutions that bear some similarities to the US. I haven’t seen any that are exactly the same.

But as I think we’ll talk about … Maybe we can discuss a little bit later, is what other possible institutions there might be. It’s a little bit tricky to … I think generally just tricky to insulate election administration from partisan considerations. And so, I don’t know that it is unique, as unique as it may initially appear, but I would say that this particular constellation of institutions is unusual.

You have a second question, which I think is important for us to think about, which is why is it so controversial? Why are so many people worried about this institution? I think that comes … I mean it can come from many places. There’s a great quote in a recent interview, a political scientist at UNC Charlotte, Martha Kropf, said something that I really like and I think is relevant here. She said … Talking about this institution, she said, “It just seems fishy.”

I think that’s exactly why most of us have an instinctual response is like if a directly elected person who seems to be … Should be, you would think, responsive to voters is directly elected as a Democrat, as a Republican, it seems like they might be exactly the kind of person who would be oriented toward trying to advantage the party that elected them.

I think there’s also political science reasons to be worried about this, just even … Not even theoretical, just empirical. If we look at our legislatures at Congress, at the state legislatures, one of the common findings that we have in recent years is that if we have two people who have the exact same electorate are serving the exact same district but come from different parties, they vote in the legislature in radically different ways on bills that matter quite a lot.

And so, when we look at these legislative institutions, we get this intuition that there’s high levels of polarization in policymaking, so that Democrats and Republicans make radically different decisions even when facing the same voters on very important bills. And so, the natural expectation is that perhaps this would carry down to the level of election administrators who are responsible for administering our elections.

Matt Grossmann: So you use a regression discontinuity design. So I want you to talk a little bit about why you do that and why we can’t just compare all Democratic and Republican clerks. But in doing so, maybe you could talk about just the basic associations. Is it true that Democratic or Republican elected clerk areas would have higher turnout or higher results in favor of their party? Is there any other thing we can do? Can we just compare Democrats who replace Republicans or anything like that?

Daniel Thompson: Yeah. I think the first question, like you say, are places that elect Democrats as clerks and places that elect Republicans as clerks different from one another? Just to begin with, just looking at the raw data without using any research design, and if we were to have just gone and found all the places that elected Democratic clerks, all the places elected Republican clerks and asked how different are they, we would see that they’re quite different.

So places that elect Democratic presidential candidates receive about 10 percentage points more of the vote in counties that have Democratic clerks than in counties that have Republican clerks. But we have a lot of strong reasons to expect that to be the case in advance without … For reasons completely unrelated to the party of the clerk. And so, these would be things like …

First and foremost, we think that those places are different because of their preferences about how everything from how the federal budget might be used and how large the federal budget ought to be to differences around how elections ought to be run. Those things are related in American politics. And so, places that vote for Democrats for a president might also vote for Democrats at the clerk level.

This wouldn’t mean that the clerk is using their authority to advantage their party. It just means that the voters, when they enter the ballot box, prefer a Democrat in both offices, or a Republican.

This association we see in the raw data. It is, I think, one thing to note that’s interesting to us in writing this paper is that we were surprised that it wasn’t even bigger than it is, that there’s a lot of recent work documenting how tight the relationship is between Democratic vote for Congress and how much an electorate votes for Democrats for Congress and how much they vote for Democrats for president, or for state legislature and president.

One of the things that’s striking here in our data is just that that correlation is quite a bit smaller between … The correlation between Democratic vote share for clerks and Democratic vote share for president is quite a bit smaller than it is for the US Senate and president or something.

That’s not terribly surprising ex post. Now that we have seen that, you can say, well, clerks do something that’s quite different from the president and senators, do something that’s a little … Maybe are voting on policies that are a little bit more similar. But we thought it was interesting in the loss given the nationalization of American politics.

Matt Grossmann: Tell us about the regression discontinuity design, and maybe in doing so, tell us how close these character elections tend to be and how many elections we’re looking at then.

Daniel Thompson: Yeah, great. So given that we know that places that elect Democratic clerks are already more likely to vote for Democrats just because of this preference for Democrats for both offices, we need to do something that is going to allow us to isolate the effect of Democratic clerks or Republican clerks on the election outcome. Ultimately what we want to be able to say is had these places had a Republican clerk running the election rather than a Democratic clerk, or the opposite, that we would have gotten different election outcomes.

And so, what we do in this paper is compare using this regression discontinuity research design. What we do is compare places that essentially randomly elected a Democratic clerk rather than a Republican clerk, and ask whether the places that elected the Democratic clerk have higher Democratic vote share.

The way that we do this, the way that this design does … This is a common design now, is to use data that use clerk elections where the election is relatively close. So some number of … In our case, we’re often using elections where the election is decided by 5% or 10%. I think 8% is our standard margin.

Then what you do is you use regression to learn about what’s happening in tied elections. So places where the election would … If we had data, what would we have seen when the election was literally tied and it was actually a coin flip where the Democrat won or the Republican won? And using this, we can say that those places, because it was essentially a coin flip, the places that elected Democrats and the places that elected Republicans are identical before having elected those people, and so any differences that we see in how they vote in the presidential election or how elections are administered are a result downstream of the decision to have elected a Democratic clerk rather than a Republican.

Matt Grossmann: You also look at turnout and don’t find a strong association, but it might be the case that we would expect a difference even if the clerks weren’t out there trying to rig elections in favor of their party, that there would just be differences in policies selected that might impact turnout. So why doesn’t that occur?

Daniel Thompson: Yeah. So I can share a little bit about how we think about this. I think we see this as two separate questions that are closely related empirically, but important to distinguish for the purposes of understanding what’s going on. One is that just knowing a bit about American history and politics, we know that even though there are many counter-examples, and I don’t want to paint with too broad a brush, we can say that I think Democrats have in the last few decades more often advocated for expansions of the franchise and policies that lower the cost of voting relative to Republicans, though again, like I say, there are many counter-examples. I don’t want to paint with too broad a brush.

And so one prediction is that Democrats might put in place policies that increase participation simply because the Democratic places prefer those policies, elect Democratic clerks, that even in the regression discontinuity design, that Democratic clerks have that orientation as compared to Republican clerks, like what we see in Congress where Democratic legislators might be more likely to support franchise expansions and policies that lower the cost of voting than Republicans coming from the exact same district.

And what we see in fact is that Democratic clerks and Republican clerks, once you account for the district characteristics like we do with the regression discontinuity design, we don’t find much evidence that Democratic clerks increase participation. We have a bunch of potential explanations for this that we try to work through. I think one classic explanation, or one, I think, very plausible explanation is simply that a lot of the things over which they have authority don’t dramatically change participation. So even if they change policies, the changes in policies don’t lead to a dramatic increase or decrease in participation.

Matt Grossmann: So one possible problem in finding a relationship is that Democrats and Republicans don’t actually implement different policies that might have an effect on a turnout. They don’t try. Another possibility is that they do, but those policies don’t necessarily end up increasing or decreasing turnout. Do we know which one it is?

Daniel Thompson: Yes. Yeah. Yeah. Okay. So I think, as you say, one explanation is that they implement different policies and those policies do not result in different outcomes, different levels of turnout. Another possibility is that they implement similar policies, and so ultimately, that has no effect on different outcomes. We have a little bit of evidence that can help us answer this question, but I wouldn’t say it’s definitive.

So what we find is that Democratic and Republican clerks seem to implement similar policies. Our evidence here is noisy. So it could be that Democrats and Republicans implement slightly different policies and we just can’t pick up those differences. Still, we don’t see evidence for dramatic differences in the policies that Democrats and Republicans implement as clerks.

Matt Grossmann: So your research suggests that some of these concerns about partisan election clerks are overblown, but some people might say that that’s just the past and we’re moving into a set of cycles where now there are real crazy folks that still think that the last election was completely stolen who are running for these offices, and maybe you can’t even win as a Republican clerk in a county without agreeing with Donald Trump that the elections were stolen and that you might be more motivated than past actors to take these kinds of actions. Maybe we haven’t seen it, but that doesn’t mean it won’t come.

Daniel Thompson:Yeah. I think this is an important question, and trying to figure out what the evidence that we have in the paper can tell us about this takes some thoughtful consideration. Of course, we do not have a crystal ball. We can’t run this regression discontinuity design for 2028, not yet, but we can do some things now that I think that at least are informative. So we’ve got a few things, some in this paper and some in research we’re currently conducting.

So just to focus on what’s in this paper first, I think as the discussion we just had starts to point toward, there are two explanations for why we’re not finding that clerks advantage their party, and they point in different directions in terms of how worried we should be about future elections. So let’s pause it for a moment that everything you said is right, that in fact there is a dramatic increase in the number of people running elections at a local level that have intentions to advantage their party.

If that’s the case, it really matters why we found that clerks historically have not advantaged their party. So one explanation is that in the past, or at least up until today, clerks have by and large not intended to advantage their party and that they did this because of some public-service motivation, or the selection process that led to the types of people who are in office leads to weak, smaller differences between Democrats and Republicans than we would expect by just looking at public-opinion surveys about how elections ought to be run.

And I think there are good reasons to expect something like this. I have earlier work studying sheriffs, which has a similar finding, where what we see is that Democrats and Republicans have wildly different views on immigration enforcement if you just ask them on opinion polls, and even on particular things around immigration enforcement in their communities, they have wildly different views. But when you look at sheriffs and how they implement immigration law at the local level, Democratic and Republican sheriffs implement very similar immigration and do immigration enforcement in very similar ways. And one thing I see in that paper is that when you survey Democratic and Republican sheriffs, they have very similar attitudes about immigration enforcement, which is striking given how different the attitudes are in the mass public.

One of the explanations that I walked through in that paper is just that the types of people who have the experience and background to be elected sheriff may have developed different opinions or may have needed to have different opinions to achieve that level of experience as a sheriff in order to successfully win the office. And so it’s possible something like that is going on here, where Democratic and Republican clerks have similar preferences because of the selection process; the types of folks who have the background and experience to end up in that office actually hold similar views. And we see some evidence for that in our paper, but not quite as much as I saw in the sheriffs paper.

And if that were the explanation for why county clerks don’t seem to be advantaging their party, I think it would be concerning, because it would say that if we replaced these folks with folks who have wildly different views on how elections ought to be administered, then they could plausibly do something very different and shape how elections are administered at the local level.

There’s another explanation that I want to just quickly touch on, which is that there are institutional constraints to how they behave, and that’s more closely related to what I was describing before around this possibility that maybe election officials know that in fact many of the decisions that they could make to advantage their party, that is commonly thought to be potential opportunities to advantage their party, really don’t lead to much difference in outcomes.

And there’s some costs associated, whether those are legal risks or actual financial costs, or maybe they’re constrained by… They have the people who hold the purse strings, who are often county commissioners, that some external constraints are actually reining them in. And so even if you replace a bunch of sensible clerks with clerks who have strong partisan motivations, who seek to dramatically changed the outcome of elections in their community, that there are all kinds of constraints that they can’t really overcome them.

And so that’s all to say that going forward, we can’t say too much. We can say that up until today and even in 2020 we don’t see evidence for Democrats’ and Republicans’ clerks advantaging their party. We can’t say too much directly about the future, but reasoning about it from the evidence that we have, I think the real question, the things to look out for, is how much do we believe that there are constraints on this office, and if there aren’t, is it really the case that there’s a new wave of people taking over this office?

Matt Grossmann: But some of the norms and constraints that we thought were fairly obvious seemed to be tested after the 2020 election. Two examples is that election officials were pressured not to certify election results after the fact, and then Trump campaign lawyers actually asked judges to invalidate whole counties’ votes, to have them not count for the statewide totals, because of a dispute related to whether the local officials had implemented the correct policy that the state legislature had created.

So I know you didn’t study this directly, but that does seem to indicate that maybe you’re just not thinking wildly enough, I guess, about what might happen in two-partisan election administration in an era in which one of the national parties might be willing to do everything possible under their control. Now, obviously, all those efforts were pushed back by federal judges. Everyone ended up certifying elections. But can we be sure that that will happen next time?

Daniel Thompson: Yeah, I don’t think that’s something that we can directly speak to. I mean, I think everything you said is right. And the question, the way that I think we have thought about these kind of broader sets of concerns, things outside of … We’re focused on this particular office of partisan-elected clerks, and we’re studying their typical behavior. We can’t really look at these, our research design doesn’t allow us to say that much about extreme cases or about this kind of rare but really important examples.

We’re more asking, does this institution on average tend to produce these kinds of advantages for the people who hold the office? But I think it does speak to some of these questions.

So I think there’s, one part of what you were asking about is, and the kind of setup that you laid out there, the set of facts that you laid out there is just that there, even despite a lot of effort being put into trying to change the certification and counting of the votes, et cetera, that federal courts constrained those efforts.

I think that’s some of what could be happening happening. Our case. And I think just the evidence that we see in our little case bears on these questions a bit, because it suggests that maybe … It’s not direct, certainly not direct evidence for these constraints functioning, but at least it’s consistent with those constraints functioning.

Another part of this is that also, it suggests that going in, I think we were genuinely open to any possible set of findings. That partisan clerks meaningfully advantage their party or didn’t. That was why we felt it was an important thing to investigate, and go get a bunch of data, and go do the best job we could to answer the question.

I think having now seen that, I at least update a bit in the direction of being less worried about this in other areas where I would also have expected it in an advance. So that’s not to say … That’s only weekly, associated with these other types of concerns. And one of the challenges in the US today is that because elections are so incredibly close, even small things can make big differences in who wins.

But I definitely come away from this exercise of doing this, writing this paper, feeling a little bit less concerned than I was the day I went out.

Matt Grossmann: So as you mentioned, one potential explanation for some of these null findings is that there might be policies implemented, but the policies don’t make the differences that we think they make. And you have several other pieces of research that point in favor of that explanation in other contexts, looking at both vote by mail and extended absentee voting in 2020. Both places where at least partisan actors were acting like they should be making a partisan difference and didn’t.

So explain what you did there, and why those policies might not have made the difference political actors expected them to.

Daniel Thompson: Yeah, so this is also collaborative research, in this case with Jesse Yoder, Andy Hall, Jen Wu on one paper. And then on another paper, also Andy Meyers, Sandy Handan-Nader and some others. I’m sorry if I’m forgetting anybody.

But these projects were both intended to try to answer this question: how much does expanding absentee voting in 2020 and before lead to higher participation and advantages for the Democratic Party?

Exactly as you said, there’s this kind of idea that they advantage Democrats and that’s why we have so much partisan contest over these issues. In fact, what we do is kind of, in these two papers is two completely different research designs, completely different data, completely different settings and even different policies, but we find pretty similar things.

Which is that there is only a modest, if any increase in participation by switching to universal vote by mail from not sending everybody a mail ballot, and also from expanding options to vote by mail. We don’t see a large increase in participation, even in 2020. And that these changes do not seem to dramatically advantage Democrats.

The two research designs, one is studying counties that shifted from having more traditional, in-person and mixed in-person by mail elections to universal mail balloting, where everybody is sent a mail ballot. You can return it in person, but that everybody is sent a mail ballot.

That study looks at before 2020, and we find again that there’s a modest increase in participation and that Democrats don’t seem to do better in elections administered by mail.

And then in a separate paper, because we were worried that perhaps 2020 would be different. Because in 2020, Democrats would be especially worried about Covid, more than Republicans. And so if Democrats couldn’t vote by mail, they might be less likely to participate. Whereas Republicans would be just as likely to participate whether they could vote by mail or not.

And so we were worried that maybe this pattern of effects that we saw in previous years wouldn’t hold. So what we did in 2020 is looked at Texas and Indiana, where in both states there’s this policy where you can vote by mail if you’re 65 or older, but it’s very difficult to vote by mail if you’re 64 or younger.

And so we compared people who are 64 to people who were 65 and said, did this policy that changes discreetly at that age cutoff lead to higher participation when people could vote by mail in 2020? And did this differentially affect Democrats and Republicans?

And we just don’t see much evidence that this had a differential effect for Democrats and Republicans.

Matt Grossmann: This would seem to introduce a puzzle. Which is, why is it that Democrats and Republicans seem to agree that easier registration, and easier voting, and more options for voting is going to be good for Democrats if that doesn’t actually end up being the case?

Daniel Thompson: I don’t have a great answer for why this is such a highly contested issue. I think one thing is that, and one thing that I’ve heard from folks, which I think is completely reasonable, is that in some ways this is just different ethical views on how elections ought to be conducted. Genuinely different views on what is the right way to conduct an election. And that people who hold these views have sorted into the different parties accordingly.

And so we might think that even one vote, which would not be … These statistical tests cannot pick up the single vote that a person is not able to cast because they don’t have the proper identification in a voter ID state, or that the small number of votes that are cast because somebody doesn’t need to get off work and they can vote early in person.

It might be that these effects are small, but that they are meaningful just on ethical grounds. That Democrats tend to, more Democrats think that people ought to have access to the ballot, and more Republicans think that the process ought to be secure, and not allow for any questions, any potential fraud. And that these two views are just not very compatible, and they’re hard to square and easy to fight over.

And so it’s not clear to me, I think this is an open question that we ought to be asking. Is, where do these differences between Democrats and Republicans on preferences come from? But it’s not clear that they ought to arise from the outcomes of these policies.

There is one potential explanation that I don’t want to just breeze past. Which is that, while a lot of this literature is I think pretty consistently finding that these advantages, that changing these policies don’t tend to affect election outcomes all that much, our elections are very close in many places. And policy matters a lot to many people.

And so it could be that the effects are so small that we can’t really observe them, they matter in a very small number of cases. But that people who are worried about this care about those small number of cases. And so that would be a perfectly reasonable way to interpret the evidence.

It’s just that I don’t think it would be fair to interpret the evidence as being consistent with very large effects that would swing elections that would not otherwise be close.

Matt Grossmann: Sorry to be skeptical of the ethical differences explanation, but it seems like the literature that tries to explain these policy differences, say across states or localities, and even interviews with elected officials themselves, there’s often some open discussion that this is being done to advantage their political party.

If the implications are switched, like in military ballots, we often see differences that don’t seem to be consistent with other cases. There’s also literatures that have found effects on adoption of these policies having to do with potential racial differences.

So I guess I would guess that they really do think these will make large effects, but that we may just not find those effects. So I guess, is it possible that elected officials just believe that these decisions have major implications? They don’t, and they might learn that they aren’t as influential as we think?

Daniel Thompson: That’s totally possible. Yeah, that’s totally possible. I mean, it’s completely plausible. What you just laid out is completely plausible to me, and I have no strong argument against it.

I think what that tells me is that you’ve convinced me there’s something really interesting that we ought to be trying to understand better. And so I think there are many possible explanations, and we ought to go try to figure out what actually is consistent with the evidence.

Matt Grossmann: As you mentioned, this was a somewhat sleepy area of political science to think about election administrators, and election administration policy. But it’s become much more popular in recent years with, I don’t want to over-characterize the literature, but a lot of defense of what administrators are doing, and some suggestions that these kinds of policies might not be as partisan as we expect.

So how would you characterize that research? Have we benefited from the increased attention to election administration and what have we learned?

Daniel Thompson: Yeah, I think we’ve learned… I’ve learned a lot. A lot of these facts while they were in the literature, if you went and went to look for them, you might have come to this conclusion before. I know plenty of people would say that, though when I look at the previous literature in this area, I find that it’s often quite mixed and it’s hard to parse. I think it’s good that we’ve done this work. I think it’s important that we understand what are these effects and try to… Hopefully it’ll help us as we continue to try to design these institutions. One thing I would think is important to point out is that while it might seem like these institutions, we’ve studied one particular institution here, while it might seem like these institutions are just handed down and the role of the political science is just to understand how they work… In fact, just today, right now across the country, many counties are deciding whether to directly elect an election official, whether to have a partisan or a non-partisan election official.

And they’re switching back and forth in all kinds of different directions, and these decisions are often motivated by some of, or often in the discussions about these decisions, there’s often questions around how these institutions function. Is it the case that democratic clerks advantage their party or that clerks advantage their party at Democrats advantage Democrats and Republicans advantage Republicans? And in fact, my co-author on this project, Joshua Ferrer, has a really neat working paper on this question asking how much does electing these clerks versus appointing them lead to more participation, higher registration rates, et cetera, and is studying a bunch of these institutional changes. I think there’s a lot of reasons that we… There’s good reasons that political scientists ought to be focusing in this area, and also as you say, that because election administration has become increasingly contested, understanding what are areas of contests that are reasonable and what is not as well-founded is really important. And so it’s something that I hope to continue to see more work in this area.

Matt Grossmann: Despite that bright future, this research is entering a very politicized context in which nearly half the country believes that the previous election was stolen, and in the face of a pretty repeated set of not just research findings, but media and public official and even own party confirmation that it was not. Given that context, can we hope that this kind of finding will break through, or is everyone continue to believe that the other party is rigging things in their favor?

Daniel Thompson: My view is maybe a little bit… I think there are competing perspectives on how this type of research should be used. My view is that the goal of this kind of work is to develop a set of facts that can inform our theory around how institutions should be designed in order to achieve the best possible outcomes, and accordingly, any individual paper is just one part of how we can learn about this space. And that it doesn’t really tell… No single paper should be especially convincing to any public official, but that hopefully over time as we assemble more and more evidence, start to reach consistent set of findings using good research designs that this would shape the public discourse, whether that’s through the papers themselves, which I’m maybe a little bit more skeptical of, or through the types of things that you and I are doing all the time, speaking to undergraduates about research in American politics, speaking to the public about research in American politics, that that process can over time shape debates.

And so we see this all the time in our discussions of, for example, of polarization, a lot of policies that might be directed at reducing polarization, we find that maybe they’re not as effective as we thought, then advocates lean in different directions. And I think we can see might some of that in the selection administration space as well, but I don’t think it happens with any individual paper or even a series of papers. Hopefully it’s the collection of research in an area that does that, and I’m more hopeful about that.

Matt Grossmann: Your chance to tout anything that you’re working on now that might help inform these debates or anything we didn’t get to that you wanted to include.

Daniel Thompson: Sure. I’ll say one thing that is related to just the particular subject that we’re talking about here, which is that Joshua Ferrer and I also, again, one of the authors of this paper that we’ve been talking about mostly today, is the two of us have been working on trying to study this question of how different our election administrators today than they were four years ago, eight years ago. And so Joshua has led a really great effort of collecting lists of these officials across dozens of states over time for the last 30 years to extend what we have, which is just the partisan elected officials, be able to extend this to places that don’t have partisan officials, that have appointed officials rather than elected officials, et cetera. And then we’re working on linking this to campaign finance records and the voter file to be able to say, “Who are these folks and how have they changed over time? What does turnover look like? Is it really the case that these folks have fled the… A lot of the folks who have been working in this position for a long time have fled the office.”

And so hopefully we’ll have some findings on that soon in the coming six months or so. And then another project that I think is related here to this subject is that I’ve just completed a paper with [inaudible 00:49:56] studying another way in which election administration has become highly contested, which is these $350 million in private grants to local election officials from the Center for Tech and Civic Life, which is funded by Mark Zuckerberg and Priscilla Chan. And what we’ve found is consistent with a lot of the stuff we’ve talked about today, which is that there were dramatic differences in places that received the funding, that didn’t, that Democratic counties were much more likely to apply for and receive funding from these to help support the administration of elections, and people are worried that this led to increased turnout and support for Democrats in those places, and that this ultimately contributed to Biden’s win in 2020.

And what we document is that despite this pattern of Democratic places being more likely to receive this funding, that the effect of this funding on Democratic vote share and on participation was so small that we can’t distinguish it from having no effect at all, and that even at the top end of our 95% competence rules, the effect is not large enough to be consistent with having substantially affected the election. And so we take this as consistent with the theme of the work that I’ve been describing here, which is that there’s a consistent pattern that a lot of these things, even when we dramatically increase funding for elections in some places and not in others, we don’t see it dramatically changing the outcomes of the election.

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 linked on our website, How Early Voting is Changing American Elections, How Campaign Money Changes Elections Before and After Citizens United, Interpreting the Results of the 2020 Election, Does the 2022 Election Show how Democratic campaigns Win, and Do Democrats and Republicans get different results. Thanks to Daniel Thompson for joining me. Please check out How Partisan Is Local Election Administration and then listen in next time.