The final votes of the 2020 election are being counted, and we’re confronting two realities that are somewhat at odds:

  1. Republicans had a surprisingly good night in House and Senate races, including probably keeping the Senate despite a difficult map, and emerging with the largest House minority in 20 years, and …
  2. Trump appears pretty likely to lose anyway.

The latest mail ballots show Biden surging to leads in Michigan and Wisconsin. That puts him on course for an electoral college victory built on close results — somewhat like in 2016, when it was Trump who carried three decisive states by less than one percentage point.

But it didn’t necessarily have to be that way for Trump. With the caveat that it’s not over yet, there is evidence that Trump might have cost himself a winnable race, given that other top-of-the-ticket races showed Republicans doing marginally better.

Nine competitive Senate races were held in states that were also in play at the presidential level. In seven of them, Trump is running behind the GOP Senate candidate’s margin. He even ran behind challengers to incumbent Democrats in two key states.

The situation is particularly acute for Trump right now in Michigan, where Biden just pulled ahead by 0.2 percentage points. But even as that happened, the Republican Senate candidate in the state, John James, retains a small lead on Sen. Gary Peters (D-Mich.), by 0.7 points.

There’s an even bigger gap in another undecided state in which Trump is fighting for his political life, Georgia. He leads Biden by 2.2 points there with much of the heavily Democratic Atlanta vote yet to come, but that’s about two points shy of Republican Sen. David Perdue’s four-point margin. It’s even more possible here than in Michigan that Trump could lose a key state even while the GOP Senate candidate beats his Democratic opponent. (The state would go to a runoff if Perdue slips below 50 percent, though.)

The situation is similar in Georgia’s other Senate race, a special election with an open, “jungle” primary. There, the GOP candidates are combining to beat the total share of votes for Democratic candidates by even more: five points.

A third undecided state is North Carolina, and yet again Trump runs behind the GOP Senate candidate’s margin. It’s closer there, though, with Sen. Thom Tillis (R-N.C.) leading by 1.8 points and Trump up by 1.4 points, and both appear likely to win.

Trump’s margins also run behind GOP Senate candidates in Colorado (by 4.1 points), Minnesota (1.9 points) and Texas (3.9 points). The two states where he’s doing better than them are Arizona (by 1.8 points) and Iowa (0.7 points).

It remains to be seen what the final electoral college vote tally is. But it’s logical to think Trump could lose the decisive states by around a percentage point, and we have readily available evidence of other Republicans on the same ballots running ahead of him by that much or more.

That doesn’t necessarily mean it was Trump’s fault; candidates matter, and perhaps the Democratic candidates were bad and/or the GOP candidates were good. It’s also worth noting that in some of these cases, Trump’s raw share of the vote is actually bigger than those of the Republican Senate hopefuls, thanks to third-party candidates taking more of the major-party vote.

But even in those cases, Biden was able to exceed his party’s share of the major-party Senate vote in a way Trump simply wasn’t — perhaps in part because Trump so motivated his opponents to defeat him. To the extent that these Senate races reflected people’s more general partisan predispositions on Election Day, Trump might have underperformed the fundamentals in these states in a slight but costly way.

The president and his allies have often referred to “shy” Trump voters — i.e., the idea that they wouldn’t admit to supporting Trump when a pollster called them. These numbers suggest GOP-inclined voters might have been a little too shy when it came to actually supporting Trump alongside GOP Senate candidates.