Is there such a thing as an “off year” for U.S. elections? The elections in 2025 were not nearly as all-encompassing as last year’s presidential race, nor as chaotic as what is expected from next year’s midterms. But hundreds of elections were held in dozens of states, including local contests, mayoral races, special congressional elections and two highly anticipated governor’s races.

Many of the elections were seen as early tests of how lasting President Trump’s 2024 gains might be and as a preview of what might happen in 2026.

Here are five takeaways from the 2025 election cycle.

In Elections Seen as Referendums on Trump, Democrats Won Big

Democrats did well in nearly all of this year’s elections, continuing a pattern that has played out across off-year elections for the last two decades: The party that wins the White House routinely loses ground in the next round of elections.

Elections in these years are often viewed as referendums on the president’s performance. And Mr. Trump’s approval ratings, after months of holding steady, took a dip in November.

A notable shift came in New Jersey, where the majority-Hispanic townships that swung toward Mr. Trump in 2024 swung back to Democrats in the 2025 governor’s race. That contributed significantly to the victory of Representative Mikie Sherrill, the Democratic candidate, over Jack Ciattarelli, the Trump-backed Republican.

The leftward swing was viewed by many political commentators as a reaction to Mr. Trump. If that is the case, it remains to be seen how much of it will carry over into 2026.

Progressive and Moderate Democrats Are Both Claiming Victories

Democratic strategists continue to debate whether the party should embrace progressive candidates or more moderate ones. And in 2025, the election results had both sides feeling emboldened.

In New York City, Zohran Mamdani, a democratic socialist who struggled to garner support from the Democratic Party, defeated former Gov. Andrew Cuomo by nine points. A similar story played out in Jersey City, where James Solomon, a progressive, crushed former Gov. James McGreevey of New Jersey in a mayoral runoff. Progressives also prevailed in cities like Detroit and Seattle.

Centrist Democrats, meanwhile, came away with arguably the two biggest wins of the year against Trump-endorsed Republicans. Abigail Spanberger and Ms. Sherrill, both Democrats, outperformed their polling estimates and decisively won the high-profile governor’s races in Virginia and New Jersey.

The debate will continue among Democrats as several 2026 primaries have prominent progressive and moderate candidates going head to head.

In Texas, Representative Jasmine Crockett, a progressive, entered the primary race for a U.S. Senate seat against the more moderate James Talarico. A similar situation has developed in Maine, where Graham Platner has pitched himself as a more progressive alternative to Janet Mills in the party’s attempt to unseat Senator Susan Collins, a Republican. Other progressives, like Julie Gonzales in Colorado and Brad Lander in New York, are challenging incumbent Democrats in primary races.

A Record 14 Women Will Serve as Governors in 2026

Virginians elected Ms. Spanberger as their first female governor. In New Jersey, Ms. Sherrill became the second woman to secure the position. Both women significantly outperformed Vice President Kamala Harris’s margins from the 2024 presidential race, improving on her results by almost 10 points.

Female candidates also did well down the ballot. Eileen Higgins will be the first female mayor in Miami after defeating Emilio González, who had the support of Mr. Trump. And, in Seattle, Katie Wilson defeated the incumbent mayor, Bruce Harrell.

Come 2026, a record 14 women — 10 Democrats and four Republicans — will serve as governors, with six of them expected to run for re-election next year. (More than a dozen states have yet to elect a female governor.)

In New York, it is likely that both candidates will be women: Representative Elise Stefanik, a Republican, began a campaign last month against the incumbent, Kathy Hochul.

Special Elections Are Still Very Special (for Democrats)

Despite not flipping any House seats, Democrats outperformed Ms. Harris’s 2024 results in every House special election this cycle. Their wins, however, offer limited insight into what might happen in 2026.

Special elections, which happen outside of regular election cycles to fill vacated seats, draw fewer voters than those in midterm or presidential years. Special election voters tend to be older and highly engaged politically, and they are more likely to be college educated. That has given Democrats a distinct advantage in recent years, and 2025 was no exception.

Democratic strength in special elections extended to lower-profile races held this year. In Virginia, Democrats secured 64 out of 100 seats in the House of Delegates. In Georgia, Democrats won two seats on the Georgia Public Service Commission, the first time the party won a non-federal statewide office since 2006. Pennsylvania Democrats swept the major Bucks County contests, electing a Democratic district attorney for the first time. And, in Mississippi, Democrats broke the Republican supermajority in the State Senate.

Odd-Numbered Years Are Still Very Odd (for Election Polls)

Polling in off-year election cycles is challenging because it’s hard to know who will turn out to vote. This year, the polls significantly overestimated the Republicans in the Virginia and New Jersey governor’s races, which both had particularly high turnout for an off year. In 2021, polls had the opposite problem, as they overestimated Democrats.

Polling misses don’t necessarily carry over from cycle to cycle: Despite the leftward bias of the polls in 2021, they performed very well in 2022. After each election, pollsters look at the result and evaluate their performance, and then note where they went wrong. Analysis from groups like the American Association for Public Opinion Research frequently indicates that errors come from an incorrect sense of who shows up to vote. Pollsters then try to adjust for this error in the next election cycle.

The errors of 2025 may prove largely irrelevant, however, as the midterm elections will feature a larger, very different pool of voters with a new set of races, and a new host of lessons for pollsters to learn.

Off years are weird, and the polling errors they produce often are as well.