Nate Cohn is the chief political analyst at the New York Times, where he also oversees the newspaper’s polling operation. I wanted to speak with him this week about another subject he writes about, though: the redistricting wars. They have once again tilted in the Republicans’ favor in recent weeks, after a controversial Supreme Court decision and a setback for Democrats in Virginia, where they were blocked from pushing through their own gerrymander in the hopes of offsetting successful redistricting efforts in Republican-majority states across the country. I also wanted Cohn—with whom I worked at The New Republic, and who remains a friend—to give a sense of how battles over redistricting are likely to play out in the next several election cycles, and what those battles may mean for 2028 and beyond. Our conversation, edited for length and clarity, is below.

Over the past month, we have had a Supreme Court decision about the Voting Rights Act, and several moves by Republican states to create safe congressional districts.

So, as you may recall, in 2025, Donald Trump began a highly unusual, if not unprecedented, mid-cycle campaign where he sought to encourage Republican states to redraw their maps to the advantage of Republicans. A whole series of Republican states followed through on that over the ensuing months, with Texas, North Carolina, and Florida chief among them. And this will likely net the Republicans a meaningful number of seats in the House. What I think the Republicans didn’t expect was that, earlier this year, the Democrats seemed like they were mostly able to cancel out those likely-to-be-lost congressional seats with their own gerrymanders in Virginia and in California. But, over the last two weeks, that basic picture of Republican gerrymanders mostly cancelled out by Democratic gerrymanders has changed.

First, the Supreme Court substantially narrowed the power of the Voting Rights Act and specifically gave states the ability to dismantle minority-majority districts, as long as it was being done for a “partisan” purpose. That’s allowed a number of Southern states to move forward with an entirely new wave of gerrymandering, beyond what they were able to attempt previously. We’ll see what the final tally is, but it’s possible that as many as five or more seats currently occupied by Democrats could be flipped by this effort.

And then the second half of this is that the Virginia Supreme Court struck down the state’s newly gerrymandered map, which has the potential to cost Democrats up to four seats. So between those two things, the balance of gerrymanders has lurched pretty abruptly toward the right. The stalemate that existed has been broken, and the Republicans now seem likely to obtain a meaningful edge heading into the midterms.

There are a couple of states still working things out, but you calculated that, as it stands now, Democrats will likely need to win the national House popular vote by about four points to capture it, right?

That’s right. In that calculation, the Republicans will have followed through with dismantling Democratic-held majority-Black districts in South Carolina, Alabama, and Louisiana. In all three of those states, it’s not necessarily a foregone conclusion that the Republicans will eliminate every one of those Democratic seats, but given the conduct of Republicans at this point, I would expect that at least some of those Democratic districts will be eliminated.

There is only so much more that can happen before the midterms, because we are so close to the election, but is there more that Southern Republican states can do in future elections to erase other minority-majority districts and give themselves even more seats?

There are many more districts that the Republicans can try to eliminate ahead of the 2028 election that they will not attempt to eliminate before this one. One reason is that there are a number of Southern states where Republicans control the redistricting process, but where primary elections have already been held, like Texas and North Carolina.

A second factor is that the 2026 election is shaping up to be a pretty good one for Democrats, which may induce some amount of caution in Republican gerrymandering efforts to this point.

Right. In Texas, Republicans gave themselves what they thought would be five more likely seats, but polls suggest that there may be a Democratic wave this year, so it’s possible that not all five of those seats will vote Republican. And other states might not want to take similar risks.

Yeah. They could have gone further if the Supreme Court had narrowed the power of the Voting Rights Act requirement before they had redistricted. They could have insured that several of the state’s majority-Latino districts were even more Republican and put them essentially out of play. And then the second thing to consider is that because the 2026 midterm election is poised to be so Democratic, the Republicans might not want to stretch too far, out of fear that doing so would endanger some of their own seats.

What can Democrats do going forward, and what do you expect them to do? I’ve heard ideas suggesting that blue states could go even further in places like California—which already created more blue districts last year—could try to pass a map where all fifty-two districts are likely blue. How realistic is all this?

It’s a tough question to answer because, in most blue states, the Democrats face meaningful constitutional obstacles to drawing even more serious gerrymanders. Just to take the California example you mentioned: California had to amend the state’s constitution already in order to create more blue districts last year. There are other blue states where more aggressive redistricting efforts would take similarly aggressive steps. New York and Colorado are good examples of states where the Democrats could potentially create more blue seats, but to do it, they would have to amend the state constitution, and we don’t know whether they can do that, and how soon they could do it, because many states have different provisions about how to amend their state’s constitution. Some of those processes don’t allow them to do it quickly.

We just saw that in Virginia. Virginia attempted to amend its state constitution, and the state Supreme Court intervened and said that they didn’t follow the correct procedure in doing so, so Democrats will go back to the drawing board. They are appealing to the Supreme Court, but I think they have an argument, which is at least interesting, that the Virginia Supreme Court’s decision depends on an interpretation of what “election day” means; there’s maybe a one-per-cent chance or a tenth-of-a-per-cent chance that the Supreme Court would hear out their case on it. Much more likely is that the Virginia Democrats will have to try again in the next election cycle with another vote to amend the state constitution and another referendum to try to get voters to draw a more aggressive map.

You have mentioned red and blue states, but not purple ones. What about states like Michigan, Wisconsin, or Pennsylvania?

In the swing states, there are at least three factors that immediately come to my mind. One is whether a party can take full control of the state government. To take an example that would be potentially devastating to the Republicans, imagine that the Democrats took full control of the state government in Georgia. If they did, maybe the Democrats would attempt to redraw that state’s congressional map, and if they did, they could easily flip four or more Republican-held seats there.

In Georgia, Governor Brian Kemp just announced that he’s going to call a special session later this year to draw new congressional maps for 2028.

I suppose that’ll bring me to my second point about purple states, which is that it is harder to draw aggressive gerrymanders there than it is in a solidly red or blue state. It is possible to draw a congressional map with fifty-two Democratic seats and zero Republican seats in California by drawing fifty-two districts that were approximately Harris plus twenty. If you try to draw a Georgia map that is fourteen Republican seats to zero Democratic seats, you would draw fourteen districts that are just Trump plus two, and Republicans would be in danger of losing many of them or all of them in that particular scenario. So the Republicans will be somewhat limited in a state like Georgia in terms of obtaining too many more seats beyond what they already have. And the third thing is that we still have the state constitutional issues to consider. Michigan, Arizona, and Pennsylvania: these are all states that, in various ways, have either passed state Supreme Court decisions or constitutional amendments that have set restrictions on opportunities for partisan gerrymandering.

It seems like blue and purple states have more restrictions on gerrymandering than red states. Is that accurate?

That’s mostly true. There are some exceptions. Florida and Ohio, both red states, each have limitations on gerrymandering that have been enforced to varying degrees, but, other than that, most of the red states give the state legislature full control to draw the map however they’d like. And the same can’t be said for most blue states.

If Democrats lose the House because of gerrymandering, or even if they don’t, it seems to me that there will be incredible, incredible pressure on Democrats around the country to do whatever they can to match Republicans in terms of gerrymandering blue states. Do you agree?

That seems entirely right. I think that if Donald Trump’s initial effort to redraw maps in the G.O.P.’s favor didn’t insure that, then certainly the Supreme Court’s decision on the Voting Rights Act insured it. The advantage that Republicans might hope to obtain by this November, could, hypothetically, grow even more if the Democrats did nothing before 2028. Looking back, I think that 2008 is the last time that a party won the House popular vote by more than four points in a Presidential-election year. So, if the Democrats were to do nothing, the Republicans would be pretty clearly favored to win control of the House in 2028, even in an election they might have otherwise lost reasonably clearly. I find it very hard to believe that there won’t be an enormous amount of pressure for Democrats to retaliate. And they have every incentive to retaliate, too, because otherwise they might be shut out of government.

Look, America’s not a stranger to one party winning the popular vote and losing the election. The Democrats in particular aren’t strangers to that, but I think this would feel pretty different because it would come on the heels of a deliberate effort to redraw the national map in the G.O.P.’s favor, and it would come on the heels of a Supreme Court decision that made it possible for the Republicans to eliminate majority-Black districts. And I think that would be a very difficult pill for the Democrats to swallow.

How would the Democratic caucus likely change if it started losing seats in the South and winning them in New York and Colorado? Would it get more white?

It does seem to me that the number of Black representatives in Congress will take a big hit. It’s not as obvious to me that other non-white voting groups will see a significant decline in their representation because Democratic districts can routinely elect non-white representatives even when non-white voters are not a majority of the population. So I think it’s at least conceivable that the total amount of non-white representation in Washington doesn’t take an enormous hit even as the number of Black representatives falls, and the number of Southern Black representatives plunges.

Something that occurred to me is that one consequence of Democrats matching Republicans in the redistricting game is that the only way you could likely get a bipartisan bill banning gerrymandering through Congress would be if there was a sense that redistricting wasn’t helping either party. Do you agree, and what would such a bill look like?

I think you’re absolutely right. I think it’s pretty hard to imagine that the Republicans, in particular, would accept a ban on gerrymandering if the Republicans had a significant structural advantage in the House. And maybe the Democrats would push it through on partisan grounds in a scenario where the Democrats had narrowly managed to win the House even while the Republicans had a significant structural advantage. But, yeah, I think that it would be hard to pull that off if one party was obviously going to benefit from it.

In terms of what a hypothetical bipartisan bill to limit gerrymandering might look like, there are a number of different proposals for how to accomplish something like it, and there are at least a few different levers that reformers could potentially pull. One is: who draws the maps? Can you take it out of the hands of politicians and put it in the hands of a nonpartisan commission? I’ll note that Republicans have traditionally been skeptical of this. They think these commissions are either composed of liberals or can be influenced by liberal-interest groups.

A second thing you can do is create really strict limits on what kinds of maps can be drawn. You can limit the number of times that you split counties and municipalities, and so on. And you would do that in the hope that even if partisans were able to redraw maps, they would be constrained by these rules in ways that would limit their ability to draw the most extreme gerrymanders.

And, finally, you can change the criteria. You establish legal standards that a map needs to meet in order to be legal. Those could be explicit requirements for a map to appear to be fair by certain statistical measures of partisanship. Historically, Republicans have been fairly skeptical of this as well, partly because they think those measures are, on balance, to their disadvantage, given that they typically enjoy some kind of structural advantage in the House of Representatives. But, also, there are cases where it is challenging to draw maps that comply with these kinds of statistical tests. I think Massachusetts is a nice example from a Democratic point of view. The Democrats have a nine-to-zero map there. It’s really challenging to draw even one Trump district, let alone the three you might think that Republicans are entitled to based on their share of the popular vote in the state.

I think that if you’re trying to cobble together a bipartisan bill, the second option, where you limit what states can do, is the most tenable one for both sides. But given the way politics is going right now, I find it challenging to believe it would happen. Still, you never know, and although all these issues are really fraught, the truth is that it wouldn’t really make too much of a difference which of these options you chose. I think that all of them would yield more or less the same map: one that’s relatively fair and perhaps slightly Republican-tilting, but not one that’s so biased that you would feel compelled to have an interview about it.

Have you taken out a map and looked at every state and thought, “If partisan redistricting efforts are pushed maximally in every state, what is this going to look like nationally in a couple of years”?

I’ve sort of done that exercise. But I think that once you do that exercise, you find that you start asking yourself whether what you’re working through is worth the effort or not, for a few reasons. I’ve already mentioned that so much of this hinges on what kind of state constitutional changes you can make, so I find it hard to believe that the Democrats can pull off what they would need to do.

And the second factor is that it’s very difficult to judge what really counts as a maximum gerrymander. Take the state of Texas, for instance. The Republicans hypothetically could draw a thirty-eight-to-zero map there, but if they did that, they would probably risk some districts by spreading out reliably Republican voters. So what’s their level of risk tolerance? How far are they willing to go? And that’s a matter of speculation.

The main thing I would say is, if Democrats could really seize control of the process in every state where they reasonably could, they would have control over the fate of more Republican-held seats than the other way around. I don’t know whether people realize that or not, but it is the case that the Democrats could hypothetically, with the right combination of legal changes, eliminate more Republican-held seats than the Republicans can eliminate Democratic seats. But whether that’s possible and whether that’s realistic are very different questions.

Taking everything you’re saying, I suspect we’re going to be on the other side of 2028, and Republicans will still have an advantage in the House. Not a gigantic seven-point advantage or something like that, but some sort of advantage, like the one they have now, maybe a little less or a bit more.

I think that’s right. Republicans clearly have more easy opportunities to draw friendly districts going forward. The opportunities for Democrats to do that are more questionable to my mind. I think that if you assume that the Democrats will succeed in some, but not all, of those efforts, they’ll probably be able to cancel out much of what the Republicans would be expected to do between now and 2028, and leave us more or less where we are today. ♦