His claim that he “did great” with black voters is untrue; he did slightly better with black voters than did John McCain in 2008 or Mitt Romney in 2012 but worse than every other Republican before those two. Exit polling suggests that black voters made up 12 percent of the electorate in 2016, down slightly from 2008 and 2012.
Trump’s claim that many black voters stayed home, though, is correct.
On Sunday, the New York Times published research from a group of political scientists and data analysts that breaks out how voters who supported President Barack Obama in 2012 behaved in 2016. Most of them, unsurprisingly, voted for Hillary Clinton. Nine percent voted for Trump. Seven percent didn’t vote.
Those percentages aren’t distributed evenly by race. According to the analysis, 12 percent of white voters who had backed Obama in 2012 voted for Trump four years later. Eleven percent of black Obama 2012 voters stayed home.
Those are small percentages of the total pool of voters, but it means that the Obama-to-Trump voter pool was overwhelmingly white — and the Obama-to-nonvoting pool disproportionately black.
We see this effect in other ways, too. The U.S. Elections Project tracks Census Bureau data on turnout. In 2016, black turnout was down eight points from 2012, helping contribute to that lower percentage that black voters made up of the overall electorate.
2016 was an election cycle in which Trump’s margin of victory was one of the narrowest in U.S. history. It came down to about 78,000 votes in three states, including Michigan, Pennsylvania and Wisconsin. It’s hard not to wonder, then, how the decrease in turnout among black voters might have affected the outcome. In Michigan, where 14 percent of residents are black, Trump won by 10,704 votes of 4.8 million cast. In Pennsylvania, he won by 44,000 of 6.2 million cast — with blacks making up more than a tenth of the population. Clinton wins those states, and the 2016 race is essentially a tie.
Again: The analysis published by the Times, and conducted by Sean McElwee, Jesse H. Rhodes, Brian F. Schaffner and Bernard L. Fraga, shows that more white Obama 2012 voters than black voters stayed home. But we also know that the black vote was targeted by suppression efforts from Russian trolls (as made clear in ads released by social media companies) and, according to Bloomberg News shortly before the election, by the Trump campaign itself.
Clinton’s “1996 suggestion that some African American males are ‘super predators’ is the basis of a below-the-radar effort to discourage infrequent black voters from showing up at the polls — particularly in Florida,” Bloomberg’s Joshua Green and Sasha Issenberg wrote in October 2016.
There is another reason that black turnout might have declined from 2012 to 2016. In 2008 and 2012, the Democratic presidential candidate was Obama, the first African American president of the United States. Without Obama on the ballot in 2016, turnout fell. That is not the only time that effect was apparent. In 2008, 69 percent of registered black voters cast ballots, compared with 65 percent of whites, according to the U.S. Elections Project. In 2010, with Obama not on the ballot, 45 percent of whites voted, a drop of 20 percentage points. Only 42 percent of black voters did — a decline of 27 points.
Again, most Obama 2012 voters cast ballots for Clinton. The chart above is misleading; if we scale the bars by the size of each group, we get a better sense for how things changed from 2012 to 2016.
The block of white voters who backed Trump is much bigger than the block of black voters who stayed home. (About 5 million white Obama 2012 voters supported Trump; about 1.6 million black voters stayed home.) But in an election where tweaking any number of dials might have changed the result, it’s interesting to consider the ramifications in the shift in black turnout.
Again, it’s not just us considering that.
“Thank you to the African American community,” Trump said in December 2016, covering all of those who decided not to head to the polls or who were unable to cast ballots.