How did Jews vote this year? It depends on which pollster you ask, apparently.

Exit polls from this week’s presidential election show that a large majority of Jews voted Democratic, as they have done in the past, choosing Joe Biden over Donald Trump by a wide margin. But results also indicated that more Jews — as many as 250,000 nationwide — voted for Trump than they did four years ago.

However, the exact breakdown of Jewish votes has stirred disagreement.

An exit poll of 800 people released by J Street, a liberal Jewish Middle East advocacy group, showed Jews supporting Biden at 77 percent vs. 21 percent for Trump.

The Republican Jewish Coalition released its own exit poll showing numbers that were more favorable for their candidate. Of 600 people polled, Jews supported Biden 60.6 percent vs. 30.5 percent for Trump, with 8.8 percent either voting third party or declining to disclose how they voted.

The difference does matter: A 10 percent variance in the Jewish vote represents more than 500,000 voters, according to estimates from the Associated Press. It also could determine whether one party or the other gained or lost Jewish voters since the election in 2016.

So, who is right?

To start off, the authoritative Associated Press’ VoteCast survey, which is different from an exit poll, found Jews voting for Biden at 68 percent and Trump at 30 percent.

Arielle Levites, a social scientist who works at George Washington University, attributed the discrepancy to each poll capturing a different population of Jews.

“A very obvious reason why these polls may have produced different results is that they screen-in respondents using different criteria,” Levites said.

For example, if a poll includes Jews who describe themselves as “culturally Jewish,” then the data will skew left. That’s because nonreligious or religiously unaffiliated people are much more likely to vote Democrat.

Levites noted that the J Street exit poll used the same methods as the Pew Research Center’s 2013 “A Portrait of Jewish Americans” study, with its more expansive definition of who is a Jew. That’s the major reason, she said, why the data for J Street’s exit poll may show more support for Biden.

RJC’s results didn’t include great detail about methodology; the individuals surveyed “describe themselves as Jewish.” However, one small factor may reveal why the RJC results skewed slightly more to the right. RJC’s poll included phone calls, which are known to push the sample population toward older voters and thus create a more conservative data set. J Street first texted voters, asking if they would complete the survey by phone or online.

While AP does not share how it determines who qualifies for a Jewish poll, Levites suspects the data came from a narrower range of Jewish voters, thus showing different results.

Is it possible to tell which one is most accurate?

Ken Goldstein, a politics professor and polling expert at the University of San Francisco, agrees that polling methods that include explicit parameters about who is a Jew will certainly result in different outcomes.

But Goldstein, who has worked decision desks on election night for the last two decades, does believe that the AP survey, which is not party-affiliated and shows an approximately 70-30 Biden-Trump split among Jewish voters, may be closest to the mark.

“From data I’ve seen over the course of the year, my sense is that the answer is closer to that AP VoteCast,” Goldstein said.

If that is the case, then Trump expanded his base of Jewish voters from four years ago by roughly 5 percent. One other data point that supports this theory is the overall shift among minority voters, who moved slightly toward Trump, said retired professor of Jewish communal studies Steven Windmueller of Hebrew Union College in Los Angeles.

“Whether we would insert Jews into that consideration, it depends on how we define ourselves,” Windmueller said.

He was surprised by how many Jews voted for Trump in this election, especially because he anticipated a defection of “Never Trump” Jewish Republicans.

“But now as you look at some of the data from these surveys,” Windmueller said, “it appears he’s picked up some additional support in the Jewish community. The first question is, what happened to these ‘Never Trump’ Republican Jews?” He speculated that another group of Jewish voters might have replaced them. “Democrats? Independents? New voters? Or a growing number of Orthodox voters?” he asked.

The growth in numbers of Jews voting Republican was discussed at a Nov. 4 RJC press conference to provide its poll findings. Media consultant Ari Flesicher, who served as White House Press Secretary under George W. Bush, said the “conclusion” suggested by these numbers is that “it’s not just the embassy that moved. Jewish voters have moved, too. And they’re moving Republican.”

He was referring, of course, to this administration’s decision to move the U.S. Embassy from Tel Aviv to Jerusalem, which was very popular among Jews in this country and in Israel.

The Jewish vote for Trump also was historically high in Florida, where both the RJC poll and a New York Times/AP exit poll showed 41 percent voting Republican.

That validated the RJC’s outreach efforts in Florida, said RJC executive director Matt Brooks. The group spent half of its $10 million national budget in the state, and Florida’s Jews helped win it for Trump, he asserted.

“Going into this election, we said there are three big states for us,” he said. “Florida, Florida and Florida.”

J. Editor Sue Fishkoff contributed to this report.