The national exit poll shows that President Donald Trump has pushed up his vote share of minorities and women, say political activists.

Left-wing activist Matt Bruenig posted the 2016 and 2020 exit poll numbers late Tuesday evening before the actual results were reported:

The exit poll suggests that one-in-three Latino men voted for Trump, along with one-in-six black men.

The exit poll also suggests that women and non-whites are a 59.6 percent majority of Trump’s coalition, in contrast to the media suggestion that white men form a majority of Trump’s coalition.

Bruenig added:

“After four years of ‘white supremacy,’ Trump doubled his vote among black women, and saw gains among all Latinos,” responded Andrew Sullivan, a noted critic of the “woke” progressives’ push for identity politics focused on people’s race and sex instead of their income and education.

The philosophy of “Wokeness is unfalsifiable. But this comes close,” Sullivan tweeted.

“Whatever happens next, this has been a crushing defeat for left racial identity politics. If Trump wins because of Latino voters, it strikes the heart of an entire worldview,” Sullivan added.

“We are witnessing a dramatic and historic realignment,” said a tweet from Thao Nguyen, an election expert. He pointed to the poll results from Starr County, Texas, a district dominated by Latinos. Trump lost the district by only five points, marking a huge shift from his 60-point loss to Hillary Clinton in 2016.

The exit poll data comes from an exit survey funded by multiple media outlets. According to CNN, the exit polls:

…are a combination of in-person interviews with Election Day voters and telephone polls measuring the views of absentee by-mail and early voters, and were conducted by Edison Research on behalf of the National Election Pool. In-person interviews on Election Day were conducted at a random sample of 115 polling locations nationwide among 7,774 Election Day voters. The results also include 4,919 interviews with early and absentee voters conducted by phone. Results for the full sample have a margin of error of plus or minus 2 percentage points; it is larger for subgroups.

The 2016 exit poll is here.

The Associated Press also ran an election-day survey. AP says its data is not from an exit poll, but instead from “a random sample of registered voters to carefully calibrate a massive poll conducted using opt-in online panels.”