Texas, California, and Missouri are moving forward with plans for mid-decade redistricting to gain partisan advantage—with other states threatening to follow. They are not hiding the motive: President Trump asked Texas to gain Republicans seats and Governor Newsom is saying he needs to retaliate. Just how much has gerrymandering gained the parties in Congress? And how much is likely to change now? Eric McGhee finds that both parties are increasingly extreme in gerrymandering but that prior mid-decade redistricting gains have been small. Daniel Kolliner finds that Republican control of redistricting has led to large increases in seat share, with Democratic control gains limited to large states.
Guest: Eric McGhee, Public Policy Institute of California; Daniel Kolliner, Kenyon College
Studies: “Districts for a New Decade;” “Partisan Control Over Redistricting and the Partisan Balance in Congress.”
Transcript
Matt Grossmann: How much will partisan redistricting tip the scales? This week on The Science of Politics. For the Niskanen Center, I’m Matt Grossmann. Texas, California and Missouri are moving forward with plans for mid-decade redistricting to gain partisan advantage, with other states threatening to follow. They’re not hiding it. President Trump asked Texas to gain Republican seats before the midterm election, and Governor Newsom is saying he needs to retaliate on behalf of Democrats. Just how much has partisan control over redistricting gained the parties in terms of seats in Congress and how much is likely to change now?
This week I talked to Eric McGhee of the Public Policy Institute of California about his Publius article, Districts For a New Decade. He finds that both parties are increasingly extreme and gerrymandering when they have control, but that non-partisan commissions do work. He says, “We have had mid-decade redistricting before with limited gains, but we are entering a new era.”
I also talked to Daniel Kolliner of Kenyon College about his co-authored NBER paper Partisan Control Over Redistricting and the Partisan Balance in Congress. He finds that Republican control of redistricting has led to large increases in seat share, but that Democratic control has had outcomes in a more limited fashion.
Before we start, I want to invite you to join us for a live episode of The Science of Politics at the Niskanen Center in DC on Thursday, September 25th at 9:00 AM. I’ll be interviewing Matt Yglesias and Stephen Tellis and we’ll have coffee and bagels to discuss whether and how the Democratic Party will revitalize. You can join us in person or online at niskanencenter.org. I think today’s conversations will provide a good foundation, starting with Eric McGhee.
Tell us about the findings from your recent article on the partisan and racial outcomes from the last redistricting cycle.
Eric McGhee: Yeah, so what we did was we looked at all the plans that have been drawn, both the congressional and state legislative plans around the country. To be clear, this is the plans that were first drawn after the last census. So, obviously, there have been some changes. For instance, North Carolina drew a new plan. And now, Texas has drawn a new plan and there may be more to come. So, this is just where things stood at the end of the last official early-decade redistricting process. And we found that overall things were reasonably balanced, though they still favored Republicans to a certain extent. And a lot of the gerrymanders that had emerged in the previous decade had disappeared.
And overall, not super surprising, but the plans that were drawn by partisan actors like state legislatures and governors, all of one party versus ones that were drawn by courts or commissions tended to be more biased, less competitive. Not really any sign that there was any variation there by racial representation. So, we did look at racial representation as well and didn’t really see a lot of patterns there. The main finding for racial representation is just that typically there are fewer districts that are minority-influenced districts or majority-minority districts than the share of that group’s population in the state as a whole. That’s not super surprising because the law doesn’t really require that kind of proportionality in terms of racial representation, but nonetheless, it certainly seems to be the case.
Matt Grossmann: So, Republicans have the national advantage in terms of the number of states in places where they were able to do redistricting because of a state party control, and we have some increases in geographic advantage or the ease with which you can gerrymander due to Democratic urban concentration. So, that all should suggest that Republicans would be able to have a pretty sizable advantage in gerrymandering. But the overall map, at least at the congressional level, looks quite close between seats and votes.
Eric McGhee: I think the interesting piece there is, actually… So, our article really looked at this from the perspective of what you would expect to come out of the plans that were drawn. And then, in the actual elections that we’ve held under these plans, really, if anything, now for the congressional plans, things tip a little bit to the Democrats, which may be surprising to people because the Republicans control the House. They have a majority of the votes, they have a majority of the seats, but the interesting thing is they don’t really have as many seats as you might think that they would have, given the vote share that they got, which was not enormous. So, it’s just a little bit of a smidge of a tilt toward the Democrats, if anything.
There were a number of races around the country in 2024 especially that turned out well for Democrats and they didn’t necessarily have to. A lot of really close outcomes that tipped the Democrat’s way, and that gave them maybe a few more seats than you might expect that they would have, given their overall vote share. So, the bottom line is that the national map is pretty fair. It’s not fair in every state, so there are some Democratic gerrymanders and there are some Republican gerrymanders, but on balance, the whole thing balances out and if anything, it maybe tilts a little bit toward the Democrats.
Matt Grossmann: So, another one of your findings, not surprising, but important, is that where there’s a split control of the redistricting process, either because of a nonpartisan commission or a court or a split control of government, you get less partisan gerrymandering toward one side or the other. So, for those of us not steeped in the details, what are the states that we’re talking about? What are the big trends there? And what are he implications for the effects of these other forms of redistricting?
Eric McGhee: Yeah. So, the big change that happened there… So, there’s quite a number of states often that have some kind of split control in government. It’s usually the legislature matches the overall partisanship of the state, and then the governor is randomly of another party. You can think of what these states are. It can be places that are otherwise very tilted toward one part or the other, like Massachusetts and Maryland have had Republican governors, Louisiana and Kansas have had Democratic governors. So, it definitely creates that interesting dynamic, and that can constrain what a partisan actor can do. The individual rules at any given state level can certainly also impact things like in North Carolina for instance, the governor doesn’t have any say over the redistricting process. So, just all that matters is who controls the state legislature.
But the interesting trend that’s been happening since the last redistricting cycle, and it was around 2011, is there’s just a lot more commissions that are out there, and these commissions are groups of… It kind of varies. Sometimes they’re appointed by legislators, sometimes they are appointed through a complicated process that chooses average citizens to serve on the commission, but it’s meant to be a body that’s separate from the legislature and draws the maps instead of the legislature, which historically, it was legislatures that always drew this.
So, there’s a number of these now that didn’t exist before. So, Colorado and Michigan are a couple of prominent examples. They have one in Ohio that’s run more into some trouble. Virginia has one, New York has one, right? There’s a number of states that have these commissions, but they vary a lot in terms of their independence from the legislature and their ability to enforce their decision once they’ve made it.
Then, there have been a number of court decisions that have taken line drawing out of the hands of the legislature because at the state constitutional level, the courts decided that there was a right to avoid partisan gerrymandering in that state, and the courts then stepped in and redrew the maps. The most prominent example of this is in Pennsylvania, where the court stepped in and redrew the congressional map, and now has something that’s pretty fair. Whereas in the last redistricting cycle, it was pretty tilted toward the Republicans.
There’s a similar thing that has happened in Wisconsin, but without quite the same dramatic results in terms of the partisan balance, but nonetheless, the courts have stepped in and forced some redrawing of the maps that wouldn’t have happened otherwise. So, there’s been this trend toward more states where the state legislature doesn’t have control of the process compared to the way things were in 2011.
Matt Grossmann: So, we’re now entering this more aggressive mid-decade redistricting cycle. As you mentioned, you never quite get to take 10 years off. There’s always a few plans that are redrawn from courts or changes afterwards, but it does seem like this is a new era with the Texas redrawing and the California threat to do so, and some of the other ones on the horizon. So, what’s new here and what’s not? And what kind of historical precedents are there for this activity?
Eric McGhee: Yeah, so mid-decade redistricting actually have a pretty long history in the United States. They go back to the late 19th century. There was a period of time when a number of states would do this. It fell out of favor over for basically the entirety of the 20th century, during a period of time when control of the House of Representatives wasn’t really at stake and the parties themselves were more muddled in terms of their political identities, and there just didn’t seem to be much incentive, I think, from the perspectives of the parties, to try to push for maximum redistricting advantage. That has changed just in the course of the first few decades of the 21st century as control of the House of Representatives has just always been in the balance. It’s always been pretty close by historical standards, and the perceived consequences of that control have just seemed enormous to partisans on either side. And so, there just seems to be this pressure to try and draw the maps in their favor.
The modern era of mid-decade redistricting actually goes back to about 20 years where ironically, Texas wanted redraw its map in 2003 to give itself more favorable for the Republicans to give themselves a more favorable map. And there was an issue at that point where the question was whether you could do that given that you’re supposed to be trying to draw these districts to be equal in population and you no longer had census data that was up-to-date. It had been collected at the beginning of the decade, and now you were three, four years in and you were using the same numbers to try to balance the populations of the different districts. And so, it went all the way to Supreme Court. And Supreme Court said, Nope, that’s fine. We don’t see any problem with this. Go ahead. So, they did did that mid-decade redistricting. There hasn’t been a lot of that activity since then though, so nobody really took advantage of that in the decade since.
However, now Texas is doing it again. There’s no longer any constitutional question about whether that’s okay, so they can go ahead. And North Carolina had also, of course… The courts had ended up stepping in North Carolina and invalidating the map that the legislature drew and drawing a different one. And then, the membership of the Supreme Court of North Carolina changed by election. And then, the court made a different decision. And now, the Republicans in North Carolina decided to step in the middle of the decade to redraw the maps in pursuant to that state Supreme Court decision.
So, there have already been a couple of these mid-decade redraws since the last census in 2020. And as you say, it’s looking like we’re going to have quite a few more. California wants to do it and we’ll see whether it will happen. Missouri is already going down the process, and a few other states are contemplating taking that step as well. Ohio actually has to redraw because there’s some complications with its commission, and there was going to be a redrawn plan there anyway, and now it’s just creating an opportunity for Republicans to extract a few more seats from Democrats as long as they’re going ahead with the redrawing process.
Matt Grossmann: So, there’s a narrative among some Democrats that Governor Newsom is relying on here, that Democrats have been played into supporting nonpartisan rules, going for nonpartisan redistricting, while Republicans have moved forward with gerrymandering and that’s not going to be a winning political strategy. To what extent is that that true that Democrats have been for these independent processes? And what’s the role of that versus things like state court decisions, referendums, and just the overall ease of gerrymandering in different states in producing these imbalances?
Eric McGhee: Historically, the idea of redistricting reform has not been necessarily a Democratic thing, so it’s gone back and forth. I know the California case really well. And in California, for several decades, it was primarily a Republican talking point and something that Republicans pushed for and the Democrats resisted it, even though they didn’t typically get an opportunity to draw a gerrymander in their favor. They had one example where they managed to pull it off and then it got referended and then replaced with a different plan. So, I wouldn’t say that it’s consistently or uniformly been a Democratic thing.
In recent years because really Republicans, they took control of a lot of state legislatures in 2010, listeners may recall that it was just a really, really great year for Republicans two years into Obama’s first term, and they just won a crazy number of seats in the House of Representatives and almost took control of the Senate and they won a bunch of state legislatures. And they saw this opportunity in a way that maybe Democrats hadn’t appreciated or wanted to take advantage of. They saw this opportunity to use this newfound state-level control to draw a bunch of gerrymanders in a bunch of states.
And so, that created, I think, this dynamic of the last maybe 15 years where Democrats have been pretty favorable to reform and Republicans have not. There was a series of court cases at the federal level that went to the Supreme Court to try to establish a national standard for partisan gerrymandering, and that would’ve basically provided a common set of boundaries for every state in the country, but the court declined to do that. Eventually, vacating itself from the area entirely and saying, “We don’t have anything to say about this question, and so go-“
We don’t have anything to say about this question, and so let’s let the states decide what they want to do. They can go ahead and make changes. And that, I think really encouraged this commission movement to pick up steam and also encouraged a bunch of court cases at the state level using what I mentioned before, at the state level using state constitutional standards to try to pursue the same ends that way.
But it was always this kind of reaction to the greater use of partisan gerrymandering on the Republican side. Now, whether Democrats have buyer’s remorse or not is an interesting question. We’ll have to see how Democrats in California respond, because they were the ones who voted… Voters voted in the commission system that California uses, and so voters would have to get rid of it even temporarily, and that would be a sign maybe that times had changed if…
Obviously, it’s different voters. The commission was created 15 years ago or so. And so it’s not the same group of people precisely, but it would be interesting if we saw a movement among Democrats to overturn the thing that they had once supported.
So, I think whether this is just a democratic or Republican thing, I think it is a democratic thing now, it’s not clear whether it will be moving forward. In the last maybe 20 years, there has been this natural advantage that Republicans have in redistricting, where the that gets one party more seats than the votes that it has is if that party is wasting fewer voters, it’s a harsh term, but nonetheless wasting fewer voters in the seats that it wins.
If you’re a partisan actor, you don’t want your party to be winning the seats that it wins by large margins, because those voters above the 50% plus one needed to win the seat, those extra voters could be moved into some other seat to help you win that seat as well. And you also really don’t want to be losing a bunch of seats by narrow margins, because then you’ve got a whole bunch of voters in those seats who are voting for the loser.
So, that can happen if a plan is drawn to try to produce that result, where one party is winning its seats by large margins and losing the seats that it loses by narrow margins. It can also just happen because of where people happen to live. And historically, Democrats have had these really high concentrations of democratic populations in the central cities, and that means in those areas they win by super large margins.
And you could draw a plan that was blind to where the partisans lived and just drawing little boxes and circles and so forth, and still end up drawing a lot of districts in central cities that had really big margins for Democrats. And so it was, you can think of it as wasted votes in those places.
That is not quite as true as it once was. In part because we’ve seen this shift among voters of color in cities toward Republicans in the last election, and in the rural areas, they’ve become so heavily Republican that there’s almost this countervailing effect of wasted Republican votes in rural areas that didn’t used to exist. It used to be the rural areas, they were more Republican, but it was a little bit more mixed, so not overwhelmingly Republican.
So, that shift in the basic political geography, if you will, has in fact undercut that natural tilt toward Republicans. And now more of what’s going on I think is, it’s not that there isn’t a natural tilt in favor of Republicans, it’s just a lot smaller than it used to be. And now a lot more of it is about the exact lines that you decide to draw.
Matt Grossmann: So let’s talk through some of these specific states, congressional maps. How much change is there from the before and after of the Texas maps? What is the California change likely to look like and what other states are catching your attention as likely to have consequential changes?
Eric McGhee: Yeah, so I think the first thing to understand is that there are constraints to this process. So, if you open the floodgates and allow as many states to redraw as possible, it’s not like an unlimited number of seats can be… I mean, I think Donald Trump at some point said, “We’re going to get 100 seats.” That’s not possible. It just can’t be done.
So the constraints are, in the first case, it’s just the thing that we’ve been talking about throughout this conversation, which is some states have split control. Some states, the courts have intervened and drawn the maps and said that they have to be drawn a certain way and put guardrails. Some states have commissions, and those commissions are not always to get easy to get around. And that political geography that we were just talking about can make it easier or more challenging to draw a plan with more Democratic, more seats, more Republican seats.
The other piece is just how many seats your party already holds in each state. And if you hold seven out of eight seats, the most seats you’re going to get is one. So, there are these limits. In terms of the Texas redraw, there’s about three new Republican seats that are slam dunks for Republicans.
Then other two are much more dubious, especially if you think that 2026 is going to be a really good year for Democrats, which historically a midterm election with an unpopular president with bad polling numbers is going to be a good year for the opposite party. So, Democrats probably will do well, can’t guarantee it, but historically that would be the case. So, you would expect that maybe those last two seats would have a better chance certainly of remaining in Democratic hands.
In the case of California, the plan that has been drawn is really… The funny thing about California is that the Democrats were already outperforming in California as it is. There were three seats that the party picked up in California in 2024. Even though statewide, the party lost vote in California, it picked up three seats and it picked them up by really narrow margins.
If you look at the map that we have now, and you just kind of predict based on the presidential vote in each district as a measure of partisanship, which is a thing that a redistricter, a line drawer, has the most control over, the Democrats probably should win about 39 seats, and they currently hold 43. So, if you reran 2024 a million times, you’d probably have on average about 39 seats for Democrats. Instead, they have 43. If you look at the new plan that’s been drawn, you run that 2024 election a million times, they will average about 43.
So, I think the best way of thinking about this plan is that it turns what was a happy accident for Democrats or favorable, something that was fortunate, into something that’s expected. And again, if you think of 2026 as being a good year for Democrats, they’ll probably pick up those seats easily, plus a few more besides, and you’re going to end up with a very, very Democratic plan in California.
So, I think those two probably about balance each other out. There may even be an argument that in 2026, the plan in California, if it’s adopted, will be even better for Democrats than the one in Texas is for Republicans. And then there’s a few other states, like Missouri is already talking about redrawing its lines. Indiana has been contemplating it. Kentucky is another possibility. Florida is another possibility. These are all for Republicans.
Florida has a little bit of a hitch in that in its constitution, it has a provision that says that the legislature can’t draw a plan for partisan intent to produce a particular partisan result. But that’s interpreted by the state Supreme Court. And we don’t know what the state Supreme Court is going to do in Florida, but that would be kind of the main constraint there.
There are some states on the democratic side, but not nearly as many. Illinois already has a gerrymander and it’s a pretty aggressive one. So the idea of extracting more Democratic seats out of that plan might be tough. Oregon also has a bit of a Democratic gerrymander. It’s going to be probably a little tough to pull more Democratic seats out of that state, but it might be possible.
New York is often talked about as another state, but New York has rules that prevent a mid-decade redistricting, and it would have to take it to the voters and also have a couple of different votes of the state legislature in order to make that change. And so it’s not clear that it’s even possible by the 2026 election to redraw the New York map. So, there’s more opportunities on the Republican side than there are on the Democratic side.
One of the interesting big constraints here for the Republicans is that the entire Deep South, really the Democratic seats in the Deep South are majority Black seats for the most part. Currently, the Voting Rights Act prevents redrawing those seats to get rid of them. It would be really tough to do.
There are some cases before the Supreme Court that could make that easy to do again, but whether the court will make a decision that will undermine the Voting Rights Act in time to redraw plans in those states, whether they will even decide to undermine the Voting Rights Act. It’s not a 100% clear. But it is possible. And if they did, that would unlock the possibility of getting more Republican seats in all of those states too, which is currently, would not be allowed.
Matt Grossmann:
So, most of the discussion has been about Congress. What is the status of redrawing for state legislative maps? Is there any crossover in these discussions? And where do we stand in terms of the level of gerrymandering and state legislative maps?
Eric McGhee:
So, the interesting thing about gerrymandering is that the focus is almost entirely on Congress. People are just obsessed with trying to extract more seats to get a congressional majority. The potential for shenanigans at the state level seems at least as extreme.
If you manage to draw a plan that protects your party from losing a majority, you can lock in your control of state government and have a lot of power at the state level. And yet, it just isn’t something that state legislatures typically do, which is I find very interesting. It’s not to say that there aren’t some gerrymandered state legislative maps. Wisconsin is perhaps the best example, where the Republicans have lost the popular vote several times in the last 15 years or so, and yet they have never lost, never have fallen below, I think, 60% of the seats in that state.
So, that’s the kind of thing you might expect would happen more often. And yet, it just kind of doesn’t. And it’s so funny because really at the congressional level, the leverage to control that national majority is so much weaker for Congress. Most the biggest states are 38, 52 seats.
And you really can never be certain that what you’re doing at the state level with your congressional plan is going to end up producing the result you want at the national level, whereas you have a lot more control over what’s going on in the state legislative plan. You can be much more certain that you’re protecting your party for state legislature. And yet, there’s just been this overwhelming focus and obsession with Congress and there really isn’t as much activity at the state legislative level.
Matt Grossmann: So, you said you’re not finding a whole lot of interaction between the racial and partisan representation. Obviously, we had this traditional pattern where Republicans could benefit from racially homogenous districts, because they would concentrate Democrats, but as you mentioned, we’ve had a couple of Southern redrawings that have actually led to Democratic seat, and then we have this trend with Republicans gaining among minority voters. So, what’s a current relationship between racial and partisan outcomes, and how much is it changing?
Eric McGhee: Yeah, I think the first thing about the changes in minority voting is that they’ve much, much more significant among Latinos than among Black voters. So, there was a shift toward Republicans in 2024 among Black voters, but it wasn’t nearly as large as among Latinos. So, this is going to be something that has more consequence for a state like Texas or a state like California than for a state like Alabama or Georgia or what have you.
And certainly among Black voters in the Deep South, party and race are very wrapped up with each other. There’s not a lot of daylight between the two. And so when you’re drawing a district that is majority Black, you’re drawing a district that’s going to be Democratic. If you’re drawing a district that’s majority white, it’s almost certainly going to be, in most Southern states, it’s going to be a Republican district.
And there was this dynamic early on in the ’90s when racially conscious redistricting first became a thing, where, as you say, there was a sense that maybe you were creating these overwhelmingly Democratic districts, and that was ending up wasting votes in the way that we’ve been talking about, wasting Democratic votes and preventing them from being allocated elsewhere to win more Democratic seats. It just doesn’t seem like that’s quite as much the case as it used to be for whatever reason.
In California, for instance, the act of drawing more majority Black and Latino, well, not Black, but Latino seats, and in a couple cases, even some majority Asian American seats, has actually ended up producing more Democratic wins. So, you can compare the draft maps of the commission versus the final maps. And there’s a lot more Latino representation in the final maps than in the draft maps. And there’s also more Democratic representation.
And I think the dynamic there is just that there’s these areas of the state that are Republican but not overwhelmingly Republican but have large Latino populations, and if you go into the-
… if you then go in but have large Latino populations. And if you go into those parts of the state and deliberately draw a Latino district, you are creating a democratic district out of an area that would otherwise probably be mostly Republican, right? And there may be a similar kind of dynamic going on in other parts of the country. It’s not entirely clear, but it certainly does seem to be the case that that link is weakened. And now drawing Voting Rights Act districts seems to be a net benefit for Democrats. Maybe not a huge one, but on balance kind of in that benefit for Democrats. And one of the things that one of my co-authors, Nick Stephanopoulos have found is that in states that were Voting Rights Act states, and so had to draw a certain number of black districts, the ones that were controlled by Democrats and went through this process, ended up producing kind of balanced maps and a certain number of black districts.
And the ones that were controlled by Republicans produced very heavily Republican maps that also produced in black districts. So a lot of it is kind of the judgment about how to draw the lines. There’s a lot of wiggle room in terms of how you produce those black districts. You can produce them so that you end up with everything else being Republican, you can produce them so that some of the what’s left is democratic and there’s a lot more flexibility than the kind of traditional story of packing democratic supporters in certain districts would lead you to think.
Matt Grossmann: So the other paper for this episode finds that Republican control of redistricting has influential outcomes in terms of increasing their seat share. Democratic control can, but it’s a much more… It’s not in most states. And so I wondered, you don’t have comment on the paper specifically, but whether that kind of general idea matches your view, and if so, what might explain it and if anything is likely to change?
Eric McGhee: Yeah, I think in the recent history, Democrats have, as we discussed, had gotten into a good government vibe. The charitable way of looking at that is that they just had this kind of ethos of fair government. The less charitable way of viewing it is that they felt like ultimately they would end up with better plans that would favor them or at least not disfavor them. And that the status quo was a bunch of plans that worked to their disadvantage. And there certainly was this phenomenon in the 2010 redistricting cycle, as we mentioned, where Republicans took advantage of all these new state legislatures and governors, governorships that they claimed in 2010, the so-called trifecta, state legislature and governor and use that to their advantage to draw a bunch of very aggressive gerrymanders.
One of the interesting phenomena that I’ve noticed about redistricting in the United States is that it is kind of unpredictable which states decide to go whole hog on partisan gerrymandering. There seems to almost be kind of a political culture behind the idea. There are states that could probably be more aggressive than they are and they just don’t. But there are these states that… North Carolina is just a classic example. North Carolina has not had a calm and uneventful redistricting cycle in decades and decades. It’s just every single time they try to do something, there’s just this kind of political culture of monkeying with the process that they just can’t resist. And it is something that they almost feel is expected of them that they have to make the process interesting. They have to try and do a little bit something extra. And I think it’s kind of on both sides. They view it as a tool to try and gain some kind of advantage.
And so North Carolina is just constantly in litigation. And then the last… In the 2010 cycle, there were a number of other key states. It is like Ohio and Michigan and Wisconsin and Pennsylvania. A lot of them are big, but it was really just these certain states that seemed to constantly come up again and again in the conversation. And in that sense, it’s not really clear what predicts whether a particular state is going to be a kind of gerrymandering state or not. And it could be that if we continue down this road that we’re on right now, that every state is going to become a gerrymandering state because everybody believes that the stakes are too high for congressional control. And so they’re all going to start playing this game. But I think it is true that in the recent history, it’s been more Republican. I just don’t know that moving forward if that isn’t just sort of a historical accident and if we aren’t kind of heading into a world where both sides do it.
Matt Grossmann: So part of the current context is that this mid-decade cycle is occurring at a time of increased concern about democratic backsliding in the United States more generally. And there has been found to be a relationship between state gerrymandering and other state policies that try and restrict voting. And we obviously are now at a time when we have questioning of election outcomes and we have other things that might make us worry that this is part of a broader pattern. On the other hand, obviously gerrymandering has been with us throughout US history and we shouldn’t be making ahistorical claims about something now being the sort of end of democracy. So how do you balance thinking about those two?
Eric McGhee: The thing that’s different now is that it’s just a lot easier to draw highly precise gerrymanders because we have a lot more data. We have computers, and so the tools are just a lot better than they were in the 1880s and 1890s when this was the last time this was a much more common occurrence. Even back in the late 19th century, it wasn’t every state that did it. There were certain states that had a habit of doing this kind of thing. Ohio, for instance, would constantly be redrawing its lines mid-decade when every time the state legislature would change party hands, they’d redraw the map.
But so I think that’s more the big concern. I think the numbers that we’re talking about in terms of the number of seats that Republicans could conceivably get if everything went right for them, they got all the states that could draw to draw and draw the maximum number. And California isn’t successful with its attempt to redraw its map. No other democratic map is redrawn. The numbers are… They could be the difference between the Democrats taking control or not taking control in 2026. In a lot of conventional scenarios, if you believe the polling that we see now and some of the special election results, it’s not crazy to think that Democrats will get more seats than that anyway. And will be able to take control of the House of Representatives even if the kind of worst-case scenario for them happens.
The real question is Congress has the authority under the Constitution to regulate this process and to pass legislation that would put guardrails on every state. Democrats take control. And if this process gets really messy, maybe the parties will come together and say, “You know what? Let’s just put some constraints on this.” They won’t apply until the next census. So we can kind of blind ourselves to who’s going to benefit, but we’ll put some guardrails on this process for the future. I’m not optimistic that that will happen. And I think the biggest danger of what we’re doing now is that it creates an incentive for Democrats to want to continue with this process themselves, right? They will now be bought into the idea of drawing lines to get more seats. And there will no longer be a natural constituency for trying to put some guardrails on this process. And I think it very much can produce some really distorted outcomes and could be another example of the kind stuff that’s going on right now that a lot of people are worried about. It could just be yet another example.
Matt Grossmann: So we’ve now heard about the potential effects of mid-decade redistricting from Eric. Let’s now turn to new estimates of the long-term impact of partisan control of redistricting from Daniel [inaudible 00:47:05]. So tell us about the findings from your new NBER paper about partisan control of redistricting.
Daniel Kolliner: Okay. So there’s really two main things that I want to emphasize from the paper. So the first is thinking about the main exercise that we’re trying to do in the paper, and that’s thinking about defining partisan control of the redistricting process. So one of the things that it sounds like it’s very obvious to do, but it’s not quite as easy because you need to read the state documents to figure out who has control is determining legal control of the redistricting process, which basically for us, what this means is that a single political party is going to be able to pass the redistricting bill without the single vote from an opposing party. So to put that in a current context, there are times when a political party can’t pass a redistricting bill because they gave the party to a commission. That’s very popular. That happens in the 2020 cycle, Michigan, California, New York, all these other states that are very interested in this.
But it’s happened before the 2020 cycle with a number of states. It can also happen though, because the parties of the state government, the house, the lower house, the upper house, and the governor, they may not all be aligned, in which case then they have to sort of make an agreement that’s somewhat bipartisan to get a map approved. Whereas if one party controls all the houses and they’re the ones that are making the redistricting plan, then of course you don’t need any of the minority party to pass a bill. And those are going to be the cases that we think about being partisan controlled maps, essentially, right? Okay. So the goal is then we want to think about what happens when there is partisan control in the map. And I think obviously the thing that you would start with doing is say, “Okay, well, I’ll compare the places that have partisan control versus the places that don’t have partisan control. And then we’ll just sort of see what happens.”
And that should be obvious in determining the impact that this has. And so the problem is that our main outcome that we’re interested in is the partisan balance in Congress. So who your state sends to Congress. And it’s a little bit different than what a lot of other people are interested in that we’re not interested in trying to categorize gerrymandering or come up with some measure of what is gerrymandered, what isn’t gerrymandered. We just want to know what’s the treatment effect or what’s the consequence, if you will, of giving partisan control to a state to make a registering map. So the challenge that exists when you do this is that, okay, let’s think about a traditional, very large democratic state or very large Republican state. If that state gained legal control, that means that the governor, the lower house and the upper house, they’re all the same party.
Well, of course that state’s going to also want to send a lot of that particular party to Congress. That just means that people in that state tend to favor that party. So what we have to do in our paper is we come up with a number of methods where we try to come up with some sort of counterfactual to think about what would happen to places that gained legal control in the event that they didn’t gain control. So essentially trying to take into account the fact that some states are just going to send more of a particular party to Congress and what would happen in the situations when that state had low control versus it didn’t. And so that’s our idea of what we’re trying to do. And the main findings are that we do find evidence that this legal control definition does matter.
We find that in terms of the partisan balance of Congress, it really has an impact in the last two decades, or sorry, I should say the 2000s and the 2010 cycle, we don’t have the 2020 cycle because the cycle hasn’t completed for our research design. So we don’t have that included yet. And we find that there’s an asymmetry in the results. So the asymmetry is that in all cases, we find that Republican states send more Republicans to Congress when they have legal control. But we find that democratic states don’t have necessarily that same blanket definition in terms of the effect. What we find though is that democratic states that are large and our definition of large is more than nine representatives of the state. They do act very similar to Republican states.
And so that’s been a big question that we’ve been trying to tackle in this paper is, “Well, why is this asymmetry exists?” And so the real asymmetry is, “Okay, why is it that states that are relatively small that are democratic, why do they act differently than states that are relatively small and Republican?” And that’s been sort of the ongoing debate that we’ve been trying to think about. And a lot of it, I think just comes down to the types of states that are gaining control, and also just the fact that not many states of particular sizes are gaining control on the democratic side.
Matt Grossmann: So you do come into an active literature with, as you say, different methods. So talk a little bit about what you find in terms of the size of these effects in the context of close national partisan competition and how your estimates differ from others.
Daniel Kolliner: So there’s a lot of work in this space. We’re hardly the first people to be interested in redistricting or gerrymandering or any of these topics, especially on the partisan balance of Congress. So I think the original literature, everything sort of before the 2000s, which makes sense given our results, it more or less the prevailing idea was that redistricting wasn’t necessarily a partisan tool at that point. So you can redistrict for other purposes besides looking for to gain partisan advantage. And I think a lot of other papers had found one of the main motivations at the time was protecting incumbents, for example. And so what we show is we document, at least in the 2000, 2010s, we show that there is this trend in partisan redistricting as a result of legal control. And so one of the reasons that this is different is that our sample is different than a lot of the original research because we’re looking at things in the 2000, 2010s.
We know that in the 2010 cycle at least, there’s very explicit strategy involving redistricting and gaining more seats in Congress, particularly from the Republican Party. And when it comes to other research designs that are in the political science literature, but also in the economics literature in the 2016 to 2020 period, I would say, there’s a relatively large sort of reckoning, if you will, of one of the major research designs, which was the difference in different strategy. And so not to get too far into the weeds about the shortcomings of difference in differences in particular estimators, but it’s more or less been shown that a staggered design from difference in differences has a number of problems. And those problems are particularly an issue in the effect when there’s differences in treatment effects across time or if there’s some sort of staggered adoption. And so-
Effects across time or if there’s some sort of staggered adoption. And so we show that there are these treatment effect differences across time. And so you need to account for the fact that this happens. And so the research designs that we employ, respect that that can occur. And so what’s the literature has more or less decided to call a stacked difference in difference or a stacked event study approach. That’s the main panel estimator that we employ. But we also have two other estimators that we use that aren’t going to be quite as sort of difference in different style, but they rely on different types of counterfactuals, if you will, or different comparison groups if you will, to try to document the effects.
So, I think what it comes down to is that compared to early literature, we do document results. And then when comparing it to sort of some of the more modern literature, I think our research designs are substantially different than a lot of the other ones. And in the economics literature at least, there’s other papers that do the same thing. And when I say the same thing, I guess I should be clear, there are other papers that ask the same question. And I think what it comes down to in some of these competing papers and our paper is that we’re quite different in how we define the way that redistricting control comes about. We’re very clear that redistricting control for us only means one party can do it without any help from the other party. And some other papers they might use something like unified control instead.
And so unified control is very, very close to legal control, but in some states, you don’t need the governor to pass the redistricting bill, right? North Carolina and Connecticut come to mind there, right? In some states the commission is who’s going to pass the bill. And so it doesn’t matter if you have a unified government in that case because the commission’s still going to be the entity that passes this. And so we take that into account in our decision for who is going to have legal control. And I think that’s a relatively strong contribution just to think very precisely about what we mean legal control, meaning in this case.
Matt Grossmann: So, you do consider several reasons why Republican states might have been more successful at moving redistricting to their partisan advantage. But the one that comes up most often, at least for me, is about geographic concentration. That is, I was involved in the Michigan process and lots and lots of maps could be produced by computers or elsewhere, and it just was always the case that it was much easier to produce a map to concentrate Democrats in a few areas and thus make map more conducive to Republicans gaining statewide than the reverse. Not that you couldn’t do it, but because Democratic concentration was so strong in cities, it was easier for Republicans to do it. How do you think about that as a part of this process and other potential explanations why Republicans might be more effective than Democrats?
Daniel Kolliner: Yeah, that’s a great question. So, I have two ways of thinking about this. So, first, so the big question, I think what you’re sort of getting at is this divergence in our results between Democrats and Republicans and that we’re saying that we Republicans have this effect and that in some cases Democrats don’t have this effect, but it reads as if it’s like Republicans have an effect and Democrats don’t. Right? So, what I should be clear about is that I think one of the main reasons that this divergence occurs certainly could be the geographic advantage, but from our paper, I think that what it comes down to is in the 2000, 2010 cycle, it comes down to the states that were able to obtain control. And so when you look at the distribution here, right? In 2000, you have two small states that gained Republican, five large states that gained Republican status.
And in 2010, that jumps to eight and seven. When you compare this to the Democratic Party, it’s four small states in 2003, in 2010, three large, sorry. And then in 2010 it goes from four to one. So, basically the big difference here is that in the 2010 cycle, the Democratic Party does a very bad job of gaining political control in the traditional maybe blue states, essentially, right? Okay. The question is why doesn’t that happen? So, I mean, certainly part of that could be 2010 is a particularly bad year for Democrats. The Obama wave is not great for the elections there. But it also could be that in 2010, the Republicans really targeted state legislatures as an outcome that they cared about a lot, and it wasn’t something that was as I think thought of in the Democratic Party. And so now, okay, now I’ll turn back to your idea about the distinct advantage that Republicans versus Democrats have here.
So, I want to be clear that part of, I think the main reason that we see this result is just obtaining control. So, what you’re talking about in terms of Democrats having this disadvantage when it comes to redrawing maps, it’s based off of these simulated map exercise. And I think these simulated map exercise is really, really great and also very, very hard to do. I think one of the things that we don’t think about with simulated maps for instance, is that what you’re trying to do is you’re trying to obtain all possible outcomes of the map that are possible. And that’s a really hard problem to solve. There are papers that look at that, the Kenny 2023 PNAS paper, the work sort of behind that.
They have a really rigorous way of thinking about how to design these sets. But I think that that is a very difficult question for us to solve. And so as you pointed out, when you look at the results from these simulated map exercises, a lot of the times you find that yes, it is a lot harder for the Democrats to redistrict. But what I think I also would want to point to is that at least the results that I think you are referencing, especially your helping of the redistricting process in Michigan for the 2020 cycle is a lot of these are based off of the most recent cycle. The Kenny paper is based off of the most recent cycle, and the result, which I’ll talk about later, is that kind of equal and opposite reactions more or less.
But what I think I want to point to for our paper and sort of the response that I would have for how we would say why I’m maybe not as concerned about geographic advantage being what drives the differences in our results is that our probably most rigorous method that we design is we compare states among based off the probability that the state is going to have legal control for one party or the other. So, we do a simulated procedure based off of basically simulating a bunch of different vote shocks. And so what’s the minimum vote shock your state needs in order to flip essentially, right? Or what’s your minimum vote shock you need to become a democratic state?
Matt Grossmann: So, this discussion is occurring at a time of a wider concern about democratic backsliding having to do with all kinds of institutional changes in the United States. And certainly there are examples of gerrymandered maps that cause people’s concerns about democracy, like states where one side never relinquishes control of the legislature, even if the other side gets more of votes statewide regularly. And in the current redistricting cycle, I don’t think we should ignore that it’s being pretty blatantly developed in order to maintain control of the US House of Representatives. So, there’s not much of a pretense here that this is much other than partisan gains. On the other hand, I don’t think we would want to point to something that has a very long tradition in the United States and is kind of built into our process as the sign that we’ve gone over the edge in terms of democracy. So, how should we think about the role of gerrymandering in the sort of quality of democracy in the current context?
Daniel Kolliner: Yeah, that’s a great question. I think it’s an important thing that you really mentioned here is that this isn’t something that’s new. This has historical precedence. I mean, if you go back all the way to 1812, that’s where we get the term gerrymandering. So, in some ways, you can think of the US having a storied history of gerrymandering in the sense. And so if we were doing it in 1812, when we’re doing it in 2025, maybe it’s just part of the system, and if we thought democracy was working well at points in history, why can’t it keep working? Well? So, I think that’s certainly one of the things to sort of think about. And so I think the other thing that’s worth thinking about is scale and scope. So, what we end up sort of finding, at least in the periods that we study, is that yes, we do find this effect.
In the 2000s it doesn’t really explain a ton of the advantage for one party or the other in the 2010 period we start to explain a lot of the house gap, if you will, with the margin. But going forward, it’s hard to think that the only thing wrong with the democratic process is going to come down to the districting that occurs. And you do see a lot of states taking action towards making … Look at Michigan. I mean, you were part of the process in Michigan. States have been taking action towards this. I mean, you look at State Ohio, Ohio was relatively close to enacting a bill, very similar to what Michigan has by voter referendum. It didn’t pass on the first try, but I think democracy sometimes has small steps and maybe doesn’t pass the first time, but maybe the next time it gains more momentum and it does pass or something like that if you think that commissions are a solution in this sense.
And so I think that it’s really easy to get mad about the districting in the way that it’s designed. I think you look at your district and if you’re of the party that your district doesn’t belong to, it’s easy to get really angry about that and say, “Oh, well, my vote doesn’t count. How is this allowing my vote to be counted the same as everyone else’s vote? Because no matter what I do, I’m going to get the same type of outcome essentially.” But I think that on the whole of things you end up with, I mean similar outcomes. And I think that the geography of this is one of the smaller maybe issues of things. Not smaller, but it’s not the reason that democracy, if you think of that’s unraveling, I don’t think it’s because of gerrymandering necessarily, right? I mean, we would’ve had democracy unravel a lot sooner if partisan redistricting was the thing that unraveled.
And so I guess to bring it all together in sort of a very wordy response is I think that partisan redistricting is certainly, from my perspective, something that I don’t think is ideal. I don’t think if you told anyone in the street that this is the sort of outcome that happens. I don’t think a lot of people would think that this is awesome. But I think that in the scheme of things, I don’t think it’s the one underlying thing that if you fix this, all of your problems are going to get solved. I think that maybe a lot of times we think that if we district in a fair way, everything’s going to get fixed.
And so not to jump too far ahead of anything, some other work that at least Ethan and I and another co-author Anthony Fowler have worked on or that we’re working on is trying to think about this political control of redistricting on polarization, for instance, which I think is an interesting question. We actually find very puzzling results in that. And so much as we don’t find a huge polarizing effect of redistricting, and in fact, we find a very small, but still significant in the statistical sense, a depolarizing effect from redistricting. And so our argument there is that, okay, well, essentially what you’re doing is you’re creating, if you think about an extremity curve, the more competitive your district is, maybe the less extreme you are, you’re creating more close elections essentially, if you’re going to follow traditional theory of cracking and packing.
So, you’re going to create more close elections, and those going to be less extreme legislatures. And of course you’re going to make some very extreme cases, but those districts were already pretty extreme. And so ideologically moving that person to a more extreme district, their vote space isn’t all that different from a polarization standpoint. And so on net, what happens then is you end up sort of seeing a small depolarizing effort in this case. And so if that story is actually true, or if there’s results told throughout the rest of the paper, then it certainly doesn’t seem like the districting and the partisan gerrymandering is the real thing that’s sort of pulling at the fabric of this, right? It’s part of the story, but it’s probably not the smoking gun I think that people want to make it out to be.
Matt Grossmann: There’s a lot more to learn. The science of politics is available bi-weekly for the Niskanen Center, and I’m your host, Matt Grossmann. If you like this discussion, here are the episodes I recommend next. Is democracy declining in the American states? Is demographic and geographic polarization overstated? Moderate voters matter. Partisan election administrators don’t tip the scales and have conservatives transformed the states. A reminder to join us for a live episode of The Science of Politics at the Niskanen Center in DC on Thursday, September 25th at 9:00 AM with Matt Yglesias and Stephen Teles on whether and how the Democratic Party will revitalize. Join us in person or online. RSVP at niskanencenter.org. Thanks to Eric McGee and Daniel Kallner for joining me. Please check out Districts for a New Decade and Partisan Control over Redistricting and the Partisan Balance in Congress, and then listen in next time.
