In all the coverage of revived Democratic enthusiasm in 2017, Democrats are beginning to realize there is a problem in the party’s most loyal demographic group, African-Americans. I wrote about it this week at New York.

As Ron Brownstein observes after looking at now-available census- and voter-file-based evidence — which is more accurate than the initial exit polls — African-American turnout took a big hit in 2016:

“In 2012, African Americans holding at least a four-year college degree voted at a slightly higher rate than whites with advanced education, and African Americans without degrees turned out at notably higher rates than blue-collar whites. But in 2016, turnout in both categories dropped so sharply that it fell below the levels of college-educated and working-class whites…. In 2016, turnout sagged to about 73 percent among college-educated African Americans (down from nearly 80 percent in 2012) and to about 56 percent among those without degrees (down from over 63 percent in 2016). Overall, the Census data showed turnout among eligible African Americans dropped fully 7 percentage points from 2012 to 2016, the biggest drop over a single election for the group since at least 1980. In the battlegrounds that tipped the election to Trump, state-level Census data show black turnout plummeting in Wisconsin; skidding in North Carolina, Florida, and Ohio; and declining more modestly in Michigan and Pennsylvania.”

The big question is whether this drop-off in African-American voting is the “new normal” — at least until such time, if ever, that a figure as magnetic as Barack Obama is heading up a party ticket. Or was it a temporary phenomenon attributable in part to Hillary Clinton’s alleged problems in connecting to voters?

Writing at FiveThirtyEight, conservative analyst Patrick Ruffini offers some evidence that the black voter drop-off could be with us for a while, as shown by turnout patterns in the very hot special election in Georgia’s sixth district:

“[O]n April 18th in Georgia, black voters did not necessarily join their white counterparts in a surge of Democratic enthusiasm against Trump. Compared to turnout levels in the 2014 midterms — which, like this special election, was an off-year election where Democratic enthusiasm was low and Obama was not on the ballot — black Democratic turnout in Georgia’s 6th lagged around 10 points behind that of white Democrats, though black voters still turned out at a higher rate than Republicans as a whole did.”

This did not matter all that much, and may not matter in the June 20 runoff in Georgia’s sixth, because of the suburban district’s demographics: The African-American share of the vote is about the same as the Hispanic share, and barely above the Asian-American share. And as Ruffini noted, even the depressed African-American turnout in the primary was higher than that among self-identified Republicans.

Down the road, in 2018, these trends could matter a great deal, particularly in elections where Democrats are more heavily dependent on African-American voters. Perhaps Republicans will help the donkey party with this problem by aggressively pursuing the kind of voter suppression and mass-incarceration policies to which the GOP is already prone. But it won’t be automatic. Democrats will have to earn the kind of black turnout that seemed to come so easily when Barack Obama was running for president.