A noteworthy caution to Democratic voters on a day when Biden’s lead in the RCP poll of polls has slipped to just 4.2 points, easily the smallest margin he’s enjoyed since entering the race. Progressive flamethrower Elizabeth Warren is now hot on his heels and has even led narrowly in a couple of polls taken over the last 10 days. Maybe the exciting left-wing candidate who can mobilize Dem voters is the way to go against Trump!

Or maybe not, says researcher Alexander Agadjanian:

The experiment’s procedure was simple. A random half of participants read a news snippet illustrating the leftward shift, while the other half read about unrelated topics, such as the schedule of election dates. The news item was a few sentences that included policies discussed by the candidates: decriminalizing unauthorized border crossings; expanding undocumented immigrants’ access to government services; replacing private health insurance with a government-run system; and establishing free public college for all children from working-class families. The content was drawn directly from real news coverage.

Both sets of respondents then indicated how they planned to vote in 2020 (whether for President Trump or the eventual Democratic nominee), how strongly they were considering voting Democratic, and how motivated they felt to turn out and vote for or against the Democratic nominee. Because of the random assignment — with some reading about the policy positions and others reading innocuous, unrelated information — the difference in responses between the groups can be attributed to the effect of reading about the leftward shift.

When deciding between Mr. Trump and the Democratic nominee, voters in the middle — the independents who could ultimately tilt things in Mr. Trump’s favor — became six percentage points less likely to vote Democratic after reading about the leftward turn compared with the independents who had read the innocuous content.

The good news for Warren is that *Democratic* voters who read about the leftward shift were *more* likely to say they’d strongly consider voting for the eventual nominee — but only by a margin of three points, less than the share of indies who were repelled by the leftist trend. If you’re a true-believing progressive, maybe none of that matters. Better to place a risky bet on meaningful change in the form of Warren or Bernie Sanders as nominee than to place a safe bet on the status quo by backing Biden. Or maybe you think the “alienated independent” effect here will be weaker than Agadjanian expects. For instance, if Trump ends up bogged down in impeachment and scandal, or if the economy slows down, indies who might be reluctant to vote for a leftist might nonetheless be more reluctant to vote for Trump. And maybe Warren as nominee will so energize Democrats, including far-left Democrats who sneered at Clinton three years ago, that Democratic turnout will blow the roof off and more than compensate for the loss in independents.

But alienating independents is a big risk, obviously. And the result here buttresses Trump’s strategy to run against socialism next year no matter who the nominee ends up being, even the not-particularly-socialist Biden. That would have been an interesting follow-up question for Agadjanian — do candidates with a rep for being far-left, e.g., Sanders and Warren, lose more independents than a moderate like Biden does when indies are made to read about the party’s leftward shift or is the effect uniform across all candidates? If it’s less pronounced for Biden, obviously that supports his electability pitch.

Another possibility: Maybe Warren as nominee would simply reposition herself as kinda sorta centrist in the general election. All nominees move towards the middle a little, after all. Perry Bacon argues today at FiveThirtyEight that that’s unlikely to happen this time, though. In recent history, party nominees have tended to pander to the center of the general electorate more through “tone” and their VP choice than by backing down on policy promises they made during the primaries. Watering down one’s policies for the general election might not even amount to effective pandering:

The previous Democratic presidential nominees were all in some ways following a kind of “median voter’ model, imagining that there was a set of voters whose views were basically in between the positions of the Democrats and the Republicans. But there is a lot of evidence that moderate, swing and independent voters aren’t particularly centrist, but hold a lot of different views, some of which are conservative, some of which are liberal.

So maybe Sanders or Warren, in a general election, keep their populism pretty amped up in an effort to woo voters who may swing between the two parties but would prefer an unabashed economic populist. Maybe Harris, instead of choosing a centrist white man as her running mate, picks Warren — or Warren chooses Harris — and they run a campaign with strongly liberal stands on issues of race and identity, hoping to win the election by energizing voters particularly turned off by Trump’s racialized and racist appeals.

Any “alienated independents” who are lost now aren’t likely to be wooed back next year in the general, in other words. Democrats are going to have to beat Trump with a version of Trump’s own strategy from 2016, with sky-high turnout among their own base and indies concluding that the other party’s candidate is just a little bit more reprehensible on balance. Every election is a “lesser of two evils” election now!

Speaking of centrism and electability, go read this story about Pete Buttigieg trying to reposition himself as a centrist in the primary now that Biden’s grip on the frontrunner role has begun to slip. Moderate Dems are going to demand an alternative on the ballot next year to Warren and Sanders, even if Grandpa Joe suddenly appears unelectable. Right now, with Harris having collapsed and none of the other also-rans showing signs of life in the polls, Mayor Pete is the only game in town.