Will Democrats retake the House due to Trump’s bumbling?

The architecture of winner-take-all elections make Democrats’ chances in 2026 an uphill climb

[Dear DemocracySOS readers: this newsletter probing the innards of America’s struggling democracy only works because of the support of readers like you. Might you consider upgrading to a paid subscription? Or giving a friend or family member a gift subscription? Here’s the link]

A number of Democratic Party hopefuls see Donald Trump’s current popularity decline and his administration’s policy turbulence and confusion as providing an opening for the Democrats to retake the US Congress in the November 2026 elections. With the current House majority at 220-215, the second narrowest controlling majority in House history, Democrats only need to pick up three seats to flip control.

Rep. Suzan DelBene (D-Wash.), the chair of the Democratic Congressional Campaign Committee, says she believes her party will take back the House majority in 2026. Moderate Dem Matt Yglesias over at Slow Boring wrote, “Republicans have a very narrow majority, Donald Trump is unpopular… I think Hakeem Jeffries is quite likely to be Speaker of the House in 2027.” Political scientist Larry Sabato’s Crystal Ball has already concluded, “Our bottom line assessment of the battle for Congress is that Democrats should win the House.” Even libertarians like John Stossel and Maxim Lott at their website Election Betting Odds have pegged Democrats’ chance of re-taking the House at almost 80%.

In part, their predictions are based on a well-founded historical pattern in which the non-presidential party most always improves both its share of seats and share of the popular House vote in the non-presidential midterm elections. Other than 2002, when the GOP and President George W. Bush held the House during the midterm election, the non-presidential party has either flipped or maintained control of the House in every midterm since Ronald Reagan.

The case (so far) against Trump

In addition to historical precedent, much of their optimism is also based on an emerging reality that Trump’s presidential performance has fallen well short of expectations. Consider that President Donald Trump has promised, either as a candidate or as president, that his policies would bolster economic growth, boost domestic manufacturing with more products “made in USA,” reduce the price of groceries “on Day 1,” and make America “very rich” again. Instead, so far he has delivered pretty much the opposite – amidst the turbulence and uncertainty of the White House’s chaotic tariff policy, economic growth has declined, manufacturers are cutting orders instead of producing more, many small businesses with Chinese suppliers are facing bankruptcy, the dollar has declined and interest rates on government debt are rising, and with the stock market decline, Americans’ retirement savings are dwindling.

As a result of both the poor economic performance and the perception of bumbling incompetence, large swaths of the US public are losing confidence in the Trump team. For the first time since 2001, the Gallup poll found that more than half of Americans say their financial situation is getting worse and President Trump’s approval ratings are declining. In a CNN/SSRS poll, 66% of Americans said they are pessimistic (29%) or afraid (37%) about the economy, with just 34% feeling enthusiastic or optimistic. Donald Trump’s 100-day approval rating is the worst for a US president in 80 years. Even small businesses that had high hopes for Trump’s second term show declining confidence.

So the stars may well be aligning for the Democrats. However it is still very early in Trump’s term and too soon to predict the ultimate fate of the Trump economy.

However, there is one other crucial factor that Democrats are not accounting for, perhaps the most important of all. That is the limitations created for their success by the use of the district-based “winner take all” electoral system to elect the House. That electoral architecture presents Democratic strategists with an increasingly perplexing dilemma.

Winner-take-all has reduced competitive seats

In district-based, winner-take-all elections, one partisan side “wins all” and the other side loses. So gaining representation becomes a zero-sum game: “if you win, I lose.” This winner-take-all system has produced a stark landscape of legislative districts—indeed, entire states—that are little more than one-party fiefdoms, with ‘flat-line’ levels of competition in the vast majority of races.

Look at what happened in 2024. About 70% of US House seats were won by lopsided landslide margins (20 points or greater) or were uncontested by one of the two major parties, and about 89% were won by noncompetitive ten-point margins or more (or were uncontested). Only 36 seats — a mere 8% of all 435 House seats — were true toss-ups, won by fewer than five points. Thirty-eight US House races did not even have major party opposition. That means only a handful of seats had a competitive race in which the outcome was uncertain, and where candidates from both parties had to work hard to win votes. The rest of them were predictable snooze-fests.

What will happen in 2026?

That’s the landscape of America’s winner-take-all politics all across the nation, election after election. So now let’s apply that reality to 2026, as we assess Democrats’ chances of retaking the US House or Senate.

Looking at a map of the 435 winner-take-all House districts, one disturbing reality is starkly clear – the number of flippable seats is few and far between. The Cook Political Report currently shows 189 seats as Solid GOP, 11 seats as Likely GOP, and nine more seats as Lean GOP (209 total); and 174 as Solid Democratic seats, 19 as Likely Democrat, and 12 seats as Lean Democrat (205 total). That leaves only 19 seats as Tossups, according to Cook. Political scientist Larry Sabato’s Crystal Ball analysis similarly forecasts 209 seats rated Safe, Likely, or Leans Democratic, 207 seats rated Safe, Likely, or Leans Republican, and 19 Toss-ups.

So according to Sabato, Democrats would need to win 9 of the Toss-ups to reach the 218 seats necessary for a majority, Republicans would need 11; Cook says Republicans would need to win 9 Toss-up seats and Democrats 13.

Whichever numbers are right, with only three dozen seats within five points, the landscape of opportunity for Democrats to flip sufficient numbers of GOP seats is exceedingly small. Don't forget that, of the 36 swing districts, Democrats won – and therefore must defend – 22 of those. That leaves just 14 competitive seats held by Republicans for the Democrats to flip and gain anywhere from 9 to 13 seats.

Sounds hard? Then think of this: of those 14 seats, only four of them are districts carried by Kamala Harris in 2024. So Democrats would have to sweep a number of GOP districts in which Trump beat Harris. Moreover, Democrats must defend 13 districts carried by Trump.

But that’s not all. Republicans also will likely benefit from a mid-cycle redistricting in Ohio. The Buckeyes State’s redistricting process will allow Republicans to gerrymander Ohio’s 15 House districts to draw a map that could well allow them to win three of the five seats currently held by Democrats.

So there are no slam dunks here, no matter how goofy or authoritarian the White House acts or how badly Trump’s poll numbers dip. What that means is that Democrats will need to run a maniacally strong effort to defend what they already hold, as well as to win a few more seats to gain the House majority.

Impossible? No. Likely? I’ll let the fates decide.

It is always amazing to me that the really smart people who are constantly crunching the numbers and analyzing this don't ever account for these realities of our winner-take-all electoral system. It's the elephant in the living room, yet somehow in their partisan palavering they seem to do their best to ignore it. Which is kind of foolish, because it doesn't change reality and it just gives false hope for the wrong reasons. It’s better to have an accurate picture of the actual election landscape to help inform strategy, it seems to me.

Certainly, nothing is set in stone. Trump is such a volatile character that almost anything could happen. Plenty of possibilities over the next two years, including a recession or a war with Greenland, Canada or Panama (!) could tip the scales one way or another. All the more reason to factor in the glaringly obvious winner-take-all incentives.

For more about that, read on.

Winner Take All realities are destroying US democracy

Beyond having a more realistic assessment of Democratic chances in 2026, this “winner-take-all blind spot” on the part of so many pundits prevents them from grasping how the failure of America’s antiquated 18th century electoral system is the root of so much of what ails US politics.

The dirty little secret is that the US House elections are so predictable every election cycle. Just by looking at the presidential vote in each district in the previous election, political analysts as well as savvy campaign consultants can predict not only which candidate is going to win, but even the margin of victory in most races. Think about that for a moment, and its ramifications.

That means party leaders and campaign donors focus like a laser on trying to win those 8% of competitive seats, because in the 50-50 nation those are the races where the legislative majority will be decided. Most voters all across the nation, living in the 92% of non-competitive districts, are peripheral to this democratic rite of meaningless participation.

Moreover, when we add up all the winner-take-all factors and incentives, we discover that election results mostly are a by-product of partisan residential patterns (i.e., where people live), combined with the winner-take-all district system. Geographic demographics have become destiny, in terms of which party wins each district seat. Redistricting reforms and campaign finance reform, while having their merits, can not greatly overcome the defective architecture of America’s 18th century, geographic-based political system. True, in a handful of states, partisan gerrymanders have reduced competition, and in the handful of close contests campaign inequities can play a key role.

But for the vast majority of states and legislative seats, liberals and conservatives live in their own demographic clusters, with liberals dominating in cities and conservatives dominating in rural areas and many suburbs. When those demographics cast votes via the single-seat district system, the vast majority of districts are branded either Republican red or Democratic blue before the partisan line drawers draw their squiggly lines or big campaign donors write their checks.

The situation in individual states is particularly stark. In Texas, just 2 of 38 congressional races were decided by 10 percentage points or fewer, and Democrats did not even bother running a candidate in 5 races. In North Carolina, only one of the state’s 14 congressional districts was decided by fewer than five points. In Democratic-dominated Illinois, zero congressional races out of 17 was within a five-point margin, and just two were within 10 points. In Maryland, just one district out of eight was within a five-point margin, and Georgia did not have a single congressional district out of 14 within a 10-point margin.

That means most of these US House contests are being decided in partisan primaries in which the dominant party selects its nominees. Partisan primaries are notorious for having low voter turnout in which partisan activists are more disproportionately motivated to vote, so winning candidates must appeal more to the partisan extremes. The party nominee then easily wins the lopsided district. As a result of this winner-take-all dynamic, one study of the 2024 elections found that 87% of US House seats were effectively elected by only 7% of Americans. Our primary system of nominating party candidates not only is contributing to minority rule in the US, it is also a breeding ground for extremists.

State legislative races are even worse, with a mere 7% of state legislative races all across the country decided by fewer than 5% in 2024, according to a New York Times analysis. A whopping 38% of seats were uncontested by one of the two major parties. That’s over 2200 state legislative races in which there was no contest at all. In 2020, 27% of the 7383 state seats were uncontested, including 75% in Massachusetts, 61% in Wyoming, 58% in Rhode Island, 57% in Arkansas and 51% in Georgia.

Nationwide, most statewide contests for president, the US Senate and governor were just as noncompetitive as the US House or state legislative races. Entire states now can be categorized as safe Red Republican or Blue Democrat. The Electoral College vote for president famously came down to only seven battleground states. The rest of the nation was ignored by the presidential candidates if you didn’t live in one of the Swing Vote Seven.

This persistent reality of our geographic-based, winner-take-all system demonstrates that monopoly politics has become hardwired into our representative democracy, which at the end of the day is not all that “representative.” The sad reality is that most voters have been rendered superfluous. Their “choice” on Election Day where they live, does not reflect even a two-party system, it’s simply whether to ratify the candidate of the lone party that dominates their district or state. That’s a level of choice we expect in China’s authoritarian system, or in the old Soviet Union Politburo.

For real democracy, dump Winner Take All

There is a solution to these anti-democratic tendencies of our Winner Take All electoral system. The most profound reform would be to get rid of geographic-based, single-seat, Winner Take All districts and change the method for electing all our legislatures to a system founded on the bedrock of proportional ranked choice voting elected inside multi-seat “super districts.” With proportional representation (PR), as it is sometimes called, a range of political parties across the partisan spectrum will win seats. Voters will win representation based on what they think, instead of where they live (though there are different configurations, including hybrids that allow both geographic and ideological representation). Virtually all voters will be able to cast a vote for a winning candidate or party, and tiny pockets of swing voters in a handful of swing districts and states will have far less influence because suddenly all voters will be swing voters.

Monopoly representation by one party in any state (with more than one House representative) would be a thing of the past. Parties would not be so beholden to their own fringe extremes, and the ideological diversity within each party would not get strangled in the crib by scheming, unscrupulous party boss leaders. A well-organized minor party and independents would have new opportunities for winning representation and holding the major parties accountable. Based on the experience of the many democracies using PR, more women, young legislators, and minority perspectives, both racial and geographic, would get elected. Will more choice and broader representation, voter turnout would increase dramatically.

Congress itself could authorize this method for electing the US House, no constitutional amendment is needed. Last November the city of Portland, Oregon started using PR for the first time to elect its city council, and the overall public response was positive. A number of US cities use other methods of PR (many of them put into place to settle voting rights lawsuits to facilitate diverse representation).

That’s a brighter horizon that America desperately needs. It’s important to understand how our political system really works if we ever hope to improve it.

Steven Hill @StevenHill1776 bsky.social @StevenHill1776