The New York Post’s Jon Levine is very confident that, should Democrats nominate Sen. Bernie Sanders (I-Vt.) to oppose President Trump in November, most Jewish voters will vote for Trump. He offered that prediction on Twitter, encouraging people to take a screenshot of it for some reason. (Maybe he plans on deleting it?)

So, here:

Why is Levine so confident? He explained in a subsequent tweet.

“All of my reporting, and that of my colleagues, and so many countless unreported personal anecdotes I have seen make this conclusion inescapable,” Levine wrote. Sadly for Levine’s odds of being correct, this is akin to that apocryphal quote from Pauline Kael about not knowing any Richard Nixon voters in 1972. Just because a reporter for the New York Post (or an opinion columnist for the New York Times) is running into a lot of Jewish Trump backers doesn’t really mean that Jewish voters broadly will support him.

Despite relying on anecdotal data, might he be correct anyway? In theory, sure. It’s just not very likely.

The last time the Jewish vote in a presidential race was even close was 1980. That year, Jewish voters preferred Jimmy Carter over Ronald Reagan by a relatively narrow six-point margin. Since 1992, though, the Democratic candidate for president has beaten the Republican among Jewish voters by no less than 39 points. In 2016, Trump lost to Hillary Clinton by nearly 50 points.

Why? In part because Jewish voters identify as Democrats or Democratic-leaning independents by twice the margin than they identify as Republican or Republican-leaning. In the most recent data available from Pew Research Center, a majority of Jewish respondents identified as Democrats, three times as many as identified as Republicans.

Earning a majority of the Jewish vote means either a lot of Democratic-leaning Jewish voters don’t vote or they have decided to vote for Trump. Or really, given the numbers here, both.

It seems awfully unlikely that a lot of Jewish Democrats will vote for Trump simply because nearly half of them identify as either extremely liberal or liberal in General Social Survey data since 2000. In that biennial polling, Republican-identifying Jewish respondents were more likely to describe themselves as moderate than were the Democrats.

Might they stay home? That seems slightly more likely because Jewish Americans are still largely skeptical of Trump. Last August, Gallup reported that fewer than 1 in 3 Jews viewed Trump’s job performance with approval. More than two-thirds disapproved of the job he’s doing. You’ll notice that those numbers align with party registration — suggesting that there doesn’t seem to be anything unique about how Jewish voters in particular are considering the Trump presidency.

Polling doesn’t really tell us what’s likely to happen with Jewish voters in 2020 at this point. The Jewish population in the United States is fairly small, meaning that polls generally don’t interview enough Jewish voters to actually offer much insight into how they plan to vote. A new poll of voters in the state of New York conducted by Siena College, for example, breaks out Jewish preferences in hypothetical 2020 matchups, but the margin of error for those results is in the double digits.

To summarize, Trump could certainly improve on his 2016 performance with Jewish voters. Perhaps he could pull even with a Democratic candidate among Jewish voters. But hitting a majority? That’s a really, really, really high bar.

Though it is one that might seem significantly lower if most of the Jewish voters with whom you’re speaking happen to be fervent critics of Sanders.