In theory, the number of votes cast in each presidential election year should always be the highest on record. After all, the population of the United States keeps climbing, so you figure the pool of voters would similarly increase and, therefore, the number of votes cast. But that's not how it works.

Votes are still being counted from last week's election, including 4.3 million in California alone. But we know now that the total number of votes cast in 2016 will exceed those cast in 2012, with about 130 million banked so far vs. 129.2 million four years ago. It's almost certain that, as Hillary Clinton's campaign predicted, the total number of votes cast this year will be the highest on record, surpassing the number cast in 2008.

So far, though, we're talking about relatively little growth in the number of ballots cast over eight years, even as the population increased by 18 million people.

Michael McDonald, a political science professor at the University of Florida, has compiled turnout figures dating to the dawn of the republic. McDonald compares the number of ballots cast to the number of people eligible to vote — excluding those too young or those legally proscribed from voting. His estimate, without all of the votes in, is that 58.1 percent of the eligible population will have voted by the time all ballots are counted (and that the total number of ballots cast will near 134.5 million, setting that record). That's a slight dip from 2012.

(If McDonald's turnout estimate is accurate, the first graph above will look like this.)

The change in turnout patterns wasn't uniform across states. Data from the U.S. Election Atlas suggests that there was a slight correlation between the change in how a state voted and its turnout, with states seeing more turnout since 2012 voting more heavily Republican.

A few outliers are worth noting. There's Utah, where turnout dropped because of a lack of interest in Donald Trump's candidacy — and the result shifted strongly to the Democrats. There's Vermont, where turnout increased, but the Democrat did worse than four years ago (you are welcome to speculate why).

Trump likes to crow about the fact that he got the most votes in Republican primary history. It's possible that he'll still get the most votes of any Republican in general-election history, too, but so far he's doing about as well as the past few Republican candidates. The big increase this year was among those casting a vote for someone other than the Republican or Democrat, which hit levels last seen during the candidacies of Ross Perot in the mid-1990s.

It's a very odd result. Turnout up slightly in terms of raw numbers, but down as a percentage of those eligible. A likely drop in votes for the Democrat and a spike in votes for third party candidates, with the Republican holding steady. More votes for the Democrat, but the Republican becoming president.

But, then, a weird result was the only logical way 2016 could end.