Democrats can win the redistricting war

by Nate Silver Democrats are fighting mad about Trump-led Republican efforts to conduct mid-decade redistricting in Texas and other states in advance of next year’s midterms — and, perhaps for a change, they’re in a mood to fight back. Late last week, a UC Berkeley poll for the Los Angeles Times¹ found 55 percent of “voters who regularly cast ballots in statewide elections” favor California Gov. Gavin Newsom’s plan to remove control of the state’s redistricting process from a nonpartisan commission, which would allow the Democrat-dominated state legislature to aggressively target the 9 of 52 seats that elected Republicans to Congress last year. In the same poll, 34 percent are opposed. The 21-point margin in the survey almost exactly matches Kamala Harris’s victory margin over Trump in the Golden State last year. That’s not a coincidence, with support breaking down strongly along partisan lines. So Newsom might not have a lot of work left to do to persuade his base of the necessity of the measure. It has been endorsed by party heavyweights like Barack Obama and even conceded to by Common Cause, the plaintiff in a 2019 Supreme Court case that deemed partisan gerrymandering nonjusticiable. That decision held that it was up to Congress if it wants to place limits on gerrymandering — but that isn’t going to happen any time soon, in part because gerrymandering itself means the FDR-like supermajorities that might once have produced large-scale change in the governance of elections are all but out of reach. According to Polymarket, there’s an 87 percent chance that the initiative, Proposition 50, passes, which would place us in a new paradigm, with Democrats joining Republicans in a race to the bottom as parties with firm control over their state legislatures do whatever they can to maximize their expected number of seats in the U.S. House. That 87 percent figure is high already — but I’d probably take the over considering that the initiative will be on the ballot in a special election this November, a circumstance under which Democrats have repeatedly overperformed in recent years. The initiative marks the end of a decade-plus of a “when they go low, we go high” attitude among Democratic leaders, which the party base has increasingly soured on. And here, the base has the right strategic instincts. Actions like Newsom’s could help the party keep the playing field level in the long run — or even eke out an advantage. Upgrade to paid A Reuters headline on Sunday claimed that “Trump’s redistricting push could bring decades of Republican rule” in the House. But that’s probably wrong. As Eli wrote last week about Texas, many of the estimates we’ve seen of projected seat gains are overconfident, understating how many supposedly safe seats could still be competitive in the event of a blue wave (or a red one) next year. But the bigger issue is that if both parties pursue a maximalist strategy, there’s no particular reason why Democrats should expect to be at a structural advantage. They just need to win elections to statewide offices that determine control of the redistricting process — and then be just as ruthless as Republicans have been. True, Republicans will probably gain ground on net in advance of the 2026 midterms. Although California might indeed roughly offset Texas if Newsom’s plan is approved, Republicans are also expected to pursue redistricting plans in Ohio, Florida and other states. But Democrats will have additional opportunities to fight back in advance of 2028, particularly in New York, where the party wound up adopting a relatively fair and risk-averse map amid court challenges following the 2020 Census. The redistricting race to the bottom is bad for representative government. But the rules under the Roberts Court are clear. For more background on the history of gerrymandering, I’d recommend last week’s episode of Risky Business with Maria. I’m under a lot of deadline pressure this week², so I’m not going to repeat all of that context here. Where the Puck Is Going: Why Maximum Gerrymandering May Be Inevitable Pushkin Industries Episode But if you listen to the episode, you can probably detect a hint of exasperation in my voice. That’s because I don’t think Democrats have been playing the “game” very well. “When they go high, we go low” is a losing strategy in the high-stakes prisoner’s dilemma that redistricting has become. They’re now applying the right tactics, but it’s been a long time coming. Because none of this is new, exactly; it’s one of the more obvious examples where Democrats ought to have been skating to where the puck was going. The country has gone through spells in the past where mid-decade redistricting outside of the usual decennial Census has been more common — but it’s been less frequent in recent decades, except in response to maps overturned by courts. As someone who has to do a bunch of extra legwork for our Congressional forecasts whenever new districts are drawn, however, I can also tell you that it’s hardly been outside of the Overton Window — particularly not in Texas, where Republicans aggressively redrew the map under Tom DeLay in advance of the 2004 election as the state was becoming redder. The attitude that Democrats take whenever Trump does virtually anything is endless screaming. And I understand — indeed, I’m one of those softies who believe that all of this is quite bad, actually. But clicking your heels together three times and saying “democracy!” isn’t a solution. True, for certain corrosive actions that Trump undertakes, like undermining America’s relationships with our longstanding allies or threatening the integrity of government-produced economic statistics, there’s essentially no good recourse apart from winning back the White House in 2028. This is different: Democrats have a “trifecta”, meaning control of the governorship and all branches of the state legislature, in states representing roughly as many seats in the House as Republicans do. And if Democrats think that Trump and Republicans represent an existential threat to democracy, their actions have consistently failed to match their rhetoric. Initially seeking to renominate Joe Biden last year when his age-related decline was obvious to voters for months in advance was an unforgivable error. (The party should probably still go further in ostracizing those who supported that effort.) Harris gave Democrats a fighting chance, but she was still a suboptimal replacement. Democrats are by no means above a bit of gerrymandering of their own: check out Illinois’s 13th Congressional district, for instance. But when it’s been their turn in power, they’ve often pulled their punches. Redistricting commissions like the one Newsom is hoping to overturn in California might not be quite as nonpartisan as they seem.³ But they don’t give you the same leverage as if you explicitly draw the maps to maximize partisan benefit. In New York, meanwhile, Democrats have sometimes prioritized incumbent protection over more aggressive maps. Meanwhile, even under Obama’s large majorities, Democrats were reluctant to pursue strategies like statehood for Washington, D.C. or Puerto Rico that could rectify the skew of the Senate toward low-population, rural states. To be clear, excessively partisan districting is bad — as are districts that change every two years instead of every ten. If a genie gave me three wishes to change the U.S. Constitution, putting some curbs on this would be one of them. The issue isn’t really that it puts Democrats at a partisan disadvantage, however — it doesn’t necessarily, and yes, we’ll get back to that soon — so much as that if there are only a few dozen competitive districts out of the 435 scattered throughout the country, more voters are deprived of casting a meaningful ballot every other November. Also, you wind up with a lot of wackjobs in Congress. Already, a considerable majority of members of Congress have more to worry about from party primaries than in a general election, and this is only going to get worse as the redistricting wars escalate. And yet, what Republicans are doing is pretty clearly within the rules as they’re going to be enforced by American courts. SCOTUS’s Rucho v. Common Cause decision held that questions about partisan gerrymandering are outside of its purview. If you want to draw up new rules, you have to do it through Congress — or win enough elections that you can change the composition of the Supreme Court.⁴ I’m slightly sympathetic to the Roberts Court in the sense that I don’t think there’s any magic algorithm to ensure “fair” districts — although I also think there’s a degree of strategically advantageous “learned helplessness” in Roberts demanding legislative solutions that aren’t going to feasible anytime soon given the turn toward partisanship in American elections. Still, the rules are the rules, even when they’re biased toward entrenching the existing regime’s hold on power. (Which rules aren’t like that, honestly?) And mid-decade redistricting, which has a long history, has consistently been tolerated by courts, too. Aggressive partisan gerrymandering doesn’t intrinsically favor Republicans The assumption that any degree of partisan gerrymandering is bad for Democrats took root during the 2010s, but this is largely an accident of timing. Democrats had an awful midterm in 2010, losing 63 seats in the House amidst a backlash to Obama. That set the GOP up to draw favorable maps for itself for the next decade. While there was considerable erosion in the GOP edge by the end of the 2010s as demographic patterns shifted in the suburbs, producing a large Democratic majority in the House following the 2018 midterms, much of the reason Republicans didn’t pursue mid-decade changes in the 2010s was because they were already pretty much maxed out. One concern Democrats had during this period was that aesthetically pleasing maps — districts are required to be contiguous, so it’s hard for Democrats to take advantage of blue enclaves surrounded by red territory like Asheville, North Carolina — were inherently likely to favor Republicans because the vast majority of geographic territory in the United States has been red in recent years.⁵ This formulation is imprecise: congressional districts are required to be equally weighted by population, not by territory. (The entire Upper Peninsula of Michigan has fewer people than Lower Manhattan, where I live.) Rather, the mathematical issue is that the bluest parts of the country were bluer during this period than the reddest parts were red, which means they’d have more wasted votes if you drew contiguous maps. In 2016 — not exactly a great election for Democrats — what was then New York’s Bronx-based 18th Congressional district voted for Hillary Clinton over Trump 94-5. Trump’s best district that year, in rural Alabama, voted for him “merely” 80-17, by contrast. Back when Democrats were the party of the underclass and not the college-educated bourgeoisie, you could always find a few blue voters even in the reddest areas — Black voters, gays and lesbians, sometimes unionized workers. This has changed, however, amid a substantial decline in support for Democrats among the nonwhite working class. Last year, the largest margin Kamala Harris carried anywhere was in Pennsylvania’s 3rd District in Philly, which she won 88-11: still huge, but notably less than 94-5. The reddest parts of the country are, if anything, getting even Trumpier, however, although this effect has been less pronounced. We’re way overdue for some actual data in this post, so here is some. Let’s line up each district from #1 to #435 based on the margin it gave Democrats on the presidential ballot in 2016 and 2024. These margins are adjusted relative to the national popular vote: Clinton did win the popular vote, after all, while Harris did not. The Congressional map has become fairer Democratic presidential vote net margin relative to national popular vote by congressional district, 2016 and 2024. You can see the presence of some really, really blue districts in 2016 that aren’t really extant anymore — not so much because the lines were redrawn but because of that Republican shift among working-class minority voters away from Democrats. In 2016, Democrats had a pronounced left tail of wasted votes; now they don’t so much, and instead the spread of the vote across districts looks pretty much like a good, old-fashioned symmetrical bell curve. We can also zoom in on the effect this has on the minority of districts that are actually competitive. In 2016, the median district was 5.5 points redder than the country overall. Last year, that was down to just 1.5 points. Zooming in: 100 most competitive districts Democratic presidential vote net margin relative to national popular vote by congressional district, 2016 and 2024. And the median doesn’t tell you everything. In 2024, Democrats lost the popular vote for the U.S. House by 2.6 points, and yet nevertheless nearly won a House majority in an otherwise bad year for the party. Basically, the same situation held in 2022. Our fancy probabilistic modeling suggests there isn’t much of a Republican advantage based on the maps as contested in 2024 — maybe even a slight Democratic edge if anything — although there are a number of complications here, such that very blue districts have lower turnout on average⁶ and that Democrats are sometimes smarter about allocating fundraising and other resources to marginal seats.⁷ The Reuters article predicting perpetual doom for Democrats also pointed out that Republicans currently have a “trifecta” in 23 states, versus 15 for Democrats. But, honestly, this is pretty lazy framing, because in the House — unlike the Senate — it’s seats rather than states that matter. Furthermore, control of the redistricting apparatus doesn’t matter at all for federal elections if you’re in a 1-member state; there’s nothing either party can do to prevent the sole representative from Vermont from being a Democrat, or the only one from Wyoming from being a Republican. If you look instead at the number of marginal seats that each party controls under a “trifecta” — which I define simply as one less than seat its overall number apportioned in the 2020 Census — it’s pretty close: 153 for Republicans and 148 for Democrats. And this is after a bad election for Democrats last year. If, say, Georgia has a divided government instead of a GOP trifecta after next year, the advantage in the number of seats under the control of a trifecta would actually shift toward Democrats. Republicans control more states, but not many more seats Current state government trifectas by number of marginal U.S. House seats There’s one other advantage for Democrats, too. Perennial warfare over districting, even in off years, probably benefits whichever party can consistently turn out its voters in obscure elections, particularly special elections and state supreme court elections. And that’s been the Democrats in recent years. Democrats have also generally had higher turnout in recent midterms, which means they can use those to shore up boundaries in advance of presidential years when more Trump-leaning marginal voters show up at the polls. Even in this long post — longer than I was planning on — there’s a lot I’m leaving out. Districting isn’t quite a zero-sum game; incumbents have a selfish, if understandable, interest in protecting themselves. There are also questions about the Voting Rights Act and Trump’s efforts to exclude illegal immigrants from Congressional apportionment that I’ll leave for another time. And maybe, in the very long run, the parties will also have some interest in a more representative House if they calculate that the redistricting wars will inevitably be fought to a draw — just like maybe the headlines tomorrow will be that we’ve discovered a cure for cancer and an unlimited supply of unicorns for every child. Until that happens, the strategic imperatives are clear. Democrats don’t merely have to reciprocate — tit-for-tat might be a game-theory optimal strategy over the long run, but it doesn’t work as well if you’re always a half-step behind. They should also consider acting preemptively, something few Democrats other than Beto O’Rourke of all people have dared to suggest. In the long run, voters could turn against aggressive gerrymandering, especially if they see their longstanding moderate members of Congress replaced by partisan hacks. But to change that requires Congressional action, a new court, or a constitutional amendment — and all of those require winning elections. Democrats can’t be too proud to play by the rules of the game as they’ve so clearly been established under Roberts: they might even come out ahead. Silver Bulletin is a reader-supported publication. To receive new posts and support our work, consider becoming a subscriber. Upgrade to paid 1 The Berkeley survey confirms previous internal polling from Newsom’s camp, but you mostly shouldn’t put a lot of stock in those. 2 And I’m also supposed to be on vacation; hello from Bergen, Norway. I have a bad habit of timing these things for when there’s a lot of newsletter work to get done. 3 Groups of experts often have a subtle or not-so-subtle bias toward Democrats 4 Ruth Bader Ginsburg’s failure to retire while a Democrat could have replaced her— and Democrats failing to urge Sonia Sotomayor to do so when she could have last year — are also examples of ceding power. 5 Well, provided you ignore the vast, low-density swaths of rural Alaska that are fairly blue. 6 Partly because they contain more non-U.S. citizens, but the effects of this are considerably overstated. 7 And that Democrats have been more likely lately to leave seats uncontested, which allows the GOP to run up the score in the House popular vote.