paywall

paywall

Photo-Illustration: Intelligencer; Photos: Getty Images

As you probably know, Democrats lost a rather consequential election in 2024. They (and the country) have been suffering mightily from the many excesses of Donald Trump’s second term as president. So Democrats have been spending a lot of time analyzing their mistakes and weighing their options for recovering in the next election cycle.

Republicans, led by their perpetually self-aggrandizing leader, have spent a lot of time inflating their 2024 win into a landslide of world-historical proportions. They suggest it represented a permanent realignment of the major parties, and now the GOP can be expected to romp to victory after victory. They’ve been unwittingly assisted in this effort by some mainstream-media journalists who seem determined to find data showing calamities without end for the Democratic Party.

Now the New York Times’ Shane Goldmacher has penned an article parlaying short-term voter-registration trends into a desperate crisis for Democrats. His language is remarkably lurid. Democrats are “hemorrhaging” voters, who are in a “stampede away from” the party. And “Democrats are divided and flummoxed over what to do.” The sources he cites seem equally sure they’ve seen the future in a few years of limited data:

“I don’t want to say, ‘The death cycle of the Democratic Party,’ but there seems to be no end to this,” said Michael Pruser, who tracks voter registration closely as the director of data science for Decision Desk HQ, an election-analysis site. “There is no silver lining or cavalry coming across the hill. This is month after month, year after year.”

What these Cassandras are discussing is major-party voter-registration trends from 2020 to 2024 in the 30 states that allow registration by party. So right off the bat, the data exclude 20 states. Goldmacher’s analysis also mostly ignores the largest group of new registered voters, those who choose not to register as members of either party. Also missing is any reflection on the specific state laws and practices that influence registration decisions, particularly the ability of independents to vote in party primaries. Given the catastrophic conclusions reached, it would have been a good idea to take at least a brief look at data points that don’t have these limitations, such as surveys of party self-identification. Here’s what Gallup had to say about that just a few weeks ago:

In the second quarter of 2025, an average of 46% of U.S. adults identified as Democrats or said they are independents who lean toward the Democratic Party, while 43% identified as Republicans or said they lean Republican.

That three-percentage-point Democratic advantage compares with a tie between the two parties in the first quarter of 2025, after a four-point Republican lead in the fourth quarter of 2024. Until now, the Republican Party had led or tied in most quarters since 2023.

To be clear, there is no fixed relationship between party registration and voting by party. Democrats had a regular registration and self-identification advantage for many decades thanks to relatively large numbers of conservative Democrats. As the two parties have sorted themselves out ideologically, they’ve come much closer to parity. Focusing obsessively on 2020-through-’24 trends can be very misleading, particularly given the exclusion of data from the 20 states without party registration. But let’s say they are as indicative of the strength or weakness of the major-party “brands” as Goldmacher thinks they are: Of course Democrats lost ground during these four years. America experienced a pandemic that killed over a million people and closed down much of the economy; the first major inflationary price spike since the 1970s; and (like much of the developed world) an equally dramatic spike in immigration. It would be shocking if Republicans didn’t make voter-registration gains over this period. And Trump’s 2024 victory after Democrats executed a mid-campaign candidate change was hardly surprising.

The 2020-through-’24 GOP surge could very well end in 2026, particularly since Trump’s job-approval ratings are cratering precisely in those voter categories where he made such impressive gains in 2024.

One point Goldmacher makes that really is relevant is that Democratic voter-registration strategies may not be working:

For years, the left has relied on a sprawling network of nonprofits — which solicit donations from people whose identities they need not disclose — to register Black, Latino and younger voters. Though the groups are technically nonpartisan, the underlying assumption has been that most new voters registering would vote Democratic.

Mr. Trump upended that calculation with the inroads he made with working-class nonwhite voters.

Again, these “inroads” may not last even two years. But without question, past Democratic registration and voter-mobilization strategies have been lazy and could use a lot of fine-tuning and better targeting.

More generally, party politics in this country has gone through multiple dizzying turns in the 21st century within an overall framework of intense competition. Democrats have been declared dead in 2002, 2004, 2010, 2014, and now 2024, and Republicans have been written off in 2006, 2008, 2018, and 2020. The recent spate of obituaries for a doomed Democratic Party are no more reliable than any of these confident predictions based on short-term trends. Let them rest in peace.

Obituaries for the Democratic Party Are Premature

Obituaries for the Democratic Party Are Premature