Ripple or Wave? What Early Indicators Say About 2026
Breaking down generic ballot polling, turnout shifts, and retirements.
Don’t look now, but the midterms are nearly in full swing. Most big-name candidates have launched their primary campaigns, even as we await final decisions from a few gamechangers (looking at you, Governor Kemp).
The initial playing field, too, is largely set. On the Senate side, Georgia and Michigan look most vulnerable for Democrats, while North Carolina and Maine are dicey for Republicans. Staring down a 53-47 deficit, though, Democrats are likely to remain the minority, barring surprises in red states like Ohio, Iowa, or Texas. Over on the House side, a relatively small battleground means neither party is likely to command an overwhelming majority come January 2027. Even if Democrats keep all their seats and win every Republican-held tossup, they’re at just 224 seats.
Still eighteen months from Election Day, however, a whole lot can change. At this point in 2021, Joe Biden was still enjoying an approval rating well above 50 percent! Only after the chaotic Afghanistan withdrawal in August did his numbers slide underwater. Then came Glenn Youngkin’s off-year win in a Biden +10 state and 2022 looked like it had the makings of a red tsunami. Six months later, backlash to the Dobbs decision jolted the political environment left again. In short, we’re at the very beginning of the midterm rollercoaster.
At this early juncture, head-to-head polling is sparse and not especially informative. Name recognition for non-incumbents is low and campaigns are still in the early stages of fundraising. But other early clues can still give us a sense of how 2026 is taking shape. Can Democrats repeat their Trump 1.0 blue wave? Or will Republicans hold steady and minimize losses?
After a couple months of parity, Democrats have built a small lead in generic ballot tests. In April’s New York Times poll, for example, 47 percent of respondents preferred a Democratic candidate and 44 percent preferred a Republican. A three-point popular-vote win for Democrats would very likely hand them the House—but this is hardly indicative of a wave environment.
More interesting than the topline, however, are the voters who say they don’t know who they’ll vote for. The Times poll found that even after a “push” from the interviewer, 16 percent of Hispanic voters remained undecided. As did 15 percent of non-white non-college respondents and 11 percent of 18-29 year olds. For comparison, just 7 percent of white respondents, 6 percent of white-college respondents, and 6 percent of 45-64 year olds indicated the same. The disparities closely align with margin shifts in the 2024 election: the demographic groups that shifted the most towards Trump are the groups most indecisive about their 2026 vote.
Though non-white voters often give higher “don’t know” responses, the size of the indecision gap is new. A review of Times polls from early in the 2020 and 2022 campaigns reveal differences of only a few points between age and ethnic groups. This cycle, it seems, Trump’s newest voters are unsure if they’ll complete their Republican conversion or return to the Democratic fold. Most of these voters hopped aboard the MAGA train in 2024 because of economic concerns—essentially a nostalgia for 2019 and the roaring economy of Trump’s first term. One hundred and five days in, these voters aren’t feeling a return on their investment. Poll after poll shows Trump’s approval rating crashing with both young and Hispanic voters.
On the other hand, these swing voters are hardly pining for a Democratic return to power either. The party’s brand remains squarely in the gutter and elected officials have yet to coalesce around an effective out-of-power strategy. So we’re left with a sizable bloc of young and non-white voters who are up for grabs in 2026—that is if they vote at all.
The emerging (and expanding) Democratic advantage with high-propensity voters has left party leaders optimistic about off-year and midterm elections, regardless of political currents. Split Ticket estimates, for example, that the electorate that showed up for Wisconsin’s Supreme Court election voted for Harris by 7 points. Turnout, not persuasion, powered the liberal victory. This promises to be an essential distinction as 2026 coverage heats up: margin swings between presidential and midterm years often hide that the most important difference is who shows up.
Even with votes still months away, we can reasonably predict a few things about the 2026 electorate. Turnout from 18-29 year olds will almost certainly drop and probably by a considerable margin. In 2020 and 2024, young voters were 16 percent of the electorate. In 2022, the share was 10 percent. Who benefits from the drop is another question. Conventional wisdom has long held that lower youth turnout hurts Democrats. But after winning young voters by just 4 points last year, a rethinking is in order. If low-propensity young voters stay home, it’s actually probably a net gain for Democrats.
The 2026 electorate is likely to be more college educated than the 2024 electorate. Since 2008, the non-college share of the vote has dropped an average of 3.75 points between presidential and midterm years. This drop is most notable among white voters, though broader demographic trends mean the white-college share has been growing and the white non-college share shrinking for a while. If Democrats maintain or expand their generic-ballot lead, 2026 could see the smallest gap ever between the white college and white non-college electorate shares—all while the partisanship of each group heads in opposite directions.
Congressional retirements are another early indicator of how insiders see the cycle shaping up. In 2018’s blue wave, 34 House Republicans did not seek reelection compared to just 18 House Democrats. Four years later, the disparity flipped in favor of the GOP. Thus far in the 2026 cycle, the number of overall retirements is outpacing 2018, 2020, and 2024, while falling just short of 2022. Interestingly, however, the traditional partisan gap has not emerged. Though most expect 2026 to be a blue year, more Democratic Senators have announced their retirements than Republicans. After the Biden fiasco, older Democrats appear particularly sensitive to the age issue—which has, at least in part, prompted the retirements of Dick Durbin, Tina Smith, and Jeanne Shaheen.
More worrying is the number of Senators returning home to run for governor. Colorado’s Michael Bennet has already announced a gubernatorial campaign and will leave his Senate term early. Alabama’s Tommy Tuberville is also considering a gubernatorial bid, as is Tennessee's Marsha Blackburn—even though she just won reelection last year. A handful of House members, too, are ditching Capitol Hill to seek their respective governor’s mansion. Legislators in both parties clearly do not view Congress as a pleasant place to work these days, which might prompt a rash of retirements unrelated to the political environment.
As the political winds swirl, intergenerational conflict and congressional dysfunction could play an outsized role in determining who controls Congress next year. Early indicators still look promising for Democrats, but expect more twists and turns ahead. Less than eighteen months to go!
Good observation on the number of Senators ditching the Senate. Long ago---perhaps 15 years ago---the House ceased to be a real institution because it quit passing budgets, which was and remains its #1 Constitutional function. Then Pelosi's idiotic two impeachments made the House further irrelevant. The Senate has moved increasingly slowly due to Mitch McConnell's Trump hate, making IT irrelevant.
Meanwhile, voters ARE speaking, and a lot of people, including this column, just don't want to hear it. It's called voter registrations. Last week, IA, NJ, and NM all (again) moved further to the GOP, the first two by a net 1500 or so, NM by 738, but it's a much smaller state. Much more important, beginning around mid 2024, we have 44 weeks x 11 states where we have tracked voter registration changes (AZ, FL, IA, NC, NH, NJ, NM, NV, NY, PA, and CA---I don't bother with OH because it is so extraordinarily red) and of these 484 observations, only ONE TIME (PA for a week) have the Ds out-registered Rs. Now, it doesn't matter if this is due to voter roll purges (declines), both parties gaining, or a combo. The Rs in 483 weeks in 11 states have seen NET REGISTRATION GAINS.
This puts the lie to most Hoax Polling that Trump is not liked. The issue for Rs is to "be" Trump. But it also shows that universally Ds are intensely DISLIKED, so much so that even in D heavy states such as NY, CA and NM they cannot manage to out-register Rs. Thank God no Democrat analysists are looking at this.
Win or lose, it doesn't matter if Democrats don't stand for something and I don't mean stand against someone.