Pollsters and prognosticators have had a rough eight years. With the exception of Nate Silver, who faced blowback for daring to suggest Donald Trump was a standard deviation away from victory in the polls, the consensus was that Hillary Clinton would win comfortably in 2016. In 2018, while Democrats did well overall, pollsters badly overestimated their strength in Florida, Ohio, and Wisconsin. In 2020, polls showed a Biden landslide, while in 2022, high turnout among white voters and registered Republicans disguised precisely which white voters were turning out.

This has not stopped prognosticators from trying to discern the future from the more than 20 million votes that have already been cast in this year’s election. The best that can be said about this approach is that it is imperfect. It only tells us who has voted, not who may vote, and it also does not tell us who they are voting for.

It also falls victim to the lack of a clear baseline: Was 2020 a normal year, or an unusual one, taking place as it did in the middle of a pandemic? We don’t know, and more importantly we do not know to what degree.

All of that aside, while the data cannot tell us what will happen, it can tell us what is not happening right now. There are certain things that if they were true, we would have expected to see them already. If Kamala Harris was viewed as a second Obama among the African American community we would expect African American voters to show up as soon as they could in order to show their support. This has not happened.

The fact that this has not happened does not mean African Americans will not vote in large numbers, or that they will not vote for her. It does, however, mean they do not seem enthused enough to vote for Harris as soon as they can.

In turn, if there were some substantial form of Trump fatigue in rural areas, we would expect to see them lagging in turnout. Instead, rural early vote turnout is breaking records.

A closer examination of the states where a substantial portion of the electorate has already cast ballots, including Georgia, where almost a quarter of registered voters have voted, can perhaps provide a glimpse into where both candidates stand just over a week out from Election Day.

Kamala Harris’s African American Enthusiasm Problems are Real in the South

Of all the assumptions surrounding Kamala Harris’s replacement of Joe Biden on the Democratic ticket, the idea that her nomination would halt or even reverse the decline in African American support for Democrats seems to have aged the worst. Barack Obama’s candidacy hyper-charged African American turnout in both his races, even in states he didn’t win. The 43.6% he won in Mississippi during his 2012 reelection remains the best Democratic performance there in thirty-years.

Donald Trump’s observation back in August regarding Harris’s lack of ties to the community was widely mocked in a period in which it was impossible for Harris to get bad press. But now Trump’s remarks look prescient, with even Barack Obama rebuking young African American men over their lack of enthusiasm for Harris.

That lack of enthusiasm has shown up perhaps most dramatically in Georgia. As of October 21, 23.8% of registered voters had cast their ballots. But while 27.8% of white voters had done so, only 22.3% of African Americans had statewide.

The gap appears even more extensive among Democrats, or at least those inclined to support Harris. Fulton County covers much of downtown Atlanta, and while it has a slightly larger black than white population (42%-37%) Joe Biden carried it 72%-26%, implying he probably won a substantial portion of the white vote.

Currently, 26.4% of Fulton County voters have cast their ballots, above the statewide average, but the gap between African American turnout and white turnout is double that for Georgia as a whole. Over a third of white Fulton County registered voters have cast their ballots, compared to only 23% of African Americans.

This pattern repeats across the state. Not only is African American turnout lagging, but it is inversely correlated with the presence of white liberals. The more liberal the white vote in an area, the lower the African American turnout. In Dekalb County, where Joe Biden won 85%-15%, and which is 50% African American, African American turnout is 23%, while white turnout is 32%. By contrast, in Forsyth County, where Donald Trump won almost two-thirds of the vote, there is virtually no gap – white turnout is 35% and African American turnout is 33%.

Statewide, African Americans make up just under 27% of current turnout. For comparison, they constituted 31% at the end of Early Voting in 2020, and 29% in the final results. The trend is for that percentage to fall over time, not rise. At this point in 2020, they represented 34% of all early votes cast.

A similar pattern is evident in North Carolina and Virginia. In the former, African Americans currently constitute 19.4% of early voters compared with 24% at this point in 2020. In Virginia, the plurality-African-American 3rd congressional district has the lowest turnout in the state both in absolute terms, constituting only 5.5% of the statewide vote, and relative to its 2020 early turnout, having cast only 29% of its 2020 total. Biden won the district 67%-31%.

Virginia’s 4th Congressional district, with the second highest African American population at 39%, is doing a bit better, having cast 35% of its 2020 early vote. By contrast, the 9th Congressional District, which Trump won 70%-28%, has cast 54% of its total 2020 early votes, and accounts for 8.5% of all early and mail votes statewide.

The MAGA Base is Voting

The strong turnout in the heavily Republican 9th district of Virginia illustrates a pattern across the wider south. The strongest turnout is located in some of the most heavily Trump rural areas. In Georgia, Towns County, where Donald Trump defeated Joe Biden 80%-19%, has the highest turnout in the state, with 42% of registered voters having cast ballots as of October 22. In Forsyth County, where Donald Trump won two-thirds of the vote, more than a third of registered voters have cast their ballots. In North Carolina, where Democrats normally have a healthy lead on registered Republicans, the numbers are a virtual tie, while in Florida, a single day of in person early voting erased the Democratic advantage in mail-in ballots.

This is not to say that the turnout picture is uniformly good for Republicans. While the highest turnout areas are heavily Republican, not all Republican strongholds are turning out. While Towns County and other parts of Georgia’s heavily Republican 9th Congressional District have led the state in turnout, the 14th district, also heavily Republican, has lagged behind. In Whitfield County, home to the city of Dalton, which Trump won 70%-29%, turnout is running at 18.8%, 7% below the statewide average. While this is driven by non-white voters sitting out the election, even among whites in Dalton less than 23% of those registered have voted. In Bartow County, which Trump carried 75%-24% in 2020, turnout is also under 23%.

However, even these figures are only marginally below the statewide average, and the worst GOP numbers tend to resemble what Democratic operatives are brandishing as their most promising: the urban turnout in the north.

What About Detroit?

Having given up pretending that Harris’s African American issues are imaginary, the media has switched to trying to argue that they are confined to the South.

In contrast to the stories of panic in the South, panic deep enough to lead Harris to transform her sixtieth birthday into a get-out-the-vote event at an Atlanta African American Church, the media has been filled with Democrats gushing about the astronomical vote-by-mail operation they have been running in Detroit, and only slightly less impressive numbers in Philadelphia.

But just how do these tales compare to reality?

As of October 21, voters in Detroit’s Wayne County had cast a combined 187,905 votes through mail and in-person early votes, accounting for 16.4% of all votes cast statewide.

This may sound impressive when Wayne only cast 15.8% of the early vote in the 2020 presidential race. But it is easy to do well if you are the only runner in the race. Early-voting began in Detroit on October 19, a full-week before it opened statewide on October 26. This has allowed Democrats to concentrate all of their resources on turning out their base voters in the city. Republicans too have the opportunity to concentrate their resources, but when the only place voting is eight-to-one Democratic, there are just fewer supporters to turn out.

Even with that head-start, just 15.5% of Wayne County registered voters have voted early, lagging the statewide level of 15.9%. The situation is better for Democrats than in Georgia, North Carolina, and Virginia, at least in the sense that overwhelming resources and skewed changes to the voting rules have managed to produce a result that is only marginally worse than 2020. But it still fails to indicate enthusiasm.

In Pennsylvania, Democrats are trumpeting their massive lead in mail-ballot requests and a favorable gap in ballot returns. However, the lead is both reduced from 2020 – 63-28% in 2024 versus 73%-18% at this point in 2020 – and turnout is down 12%. Compared to the same time in 2020, 55,000 more registered Republicans have returned their absentee ballots, while 193,000 fewer Democrats have returned theirs.

Where does that leave us?

The early voting numbers go a long way to explaining the recent behavior of the Harris campaign which appears to be abandoning Georgia and North Carolina in favor of making a final stand in the Midwest. They know from the numbers that they are not getting the African American turnout they need, while Donald Trump is currently getting what he needs. Nor do they appear to put much faith in Barack Obama’s ability to turn the situation around by publicly rebuking African American men.

Instead, they seem to be putting their faith in a 2022-esque strategy of winning enough white Republican defectors to make up for the gap. In 2022 this strategy helped Democrats outperform expectations, but they still lost the House and national popular vote by three points. It remains to be seen whether Democrats can perform better this year. Yet under the current circumstances, trying to repeat 2022 may be the best bet Democrats have left.

Walter Samuel is the pseudonym of a prolific international affairs writer and academic. He has worked in Washington as well as in London and Asia, and holds a Doctorate in International History.