As a team of nonpartisan political analysts under professor Larry Sabato at the University of Virginia pointed out Thursday, we haven't seen partisan numbers like this since the 17th Amendment was ratified in 1913, allowing senators to be elected by popular vote (as opposed to being elected by state legislatures).
In 1920, more than 95 percent of states voted for the same party for president as Senate. In 2016, of 34 states with Senate races, all 34 states voted for the same party for president and U.S. Senate.
This data will undoubtedly serve the narrative that split-ticket voting is dying (or perhaps already dead). It's yet one more example for people to point to of how partisan the United States has become.
As The Fix's Chris Cillizza lamented in September about the decline of split-ticket voting: "[I]t suggests we have become increasingly tribal in the way we think about elections. You are either with 'us' (whoever that 'us' may be) or against us. There's no room for a candidate-by-candidate assessment up and down the ballot."
If we fit the 2016 results into what we know about split-ticket voting, that may be the case. Even before 2016, split-ticket voting was at its lowest point in decades. The Fix's Philip Bump calculated that in the 1990 presidential election, 63 percent of states with Senate races voted for the same candidate for president as for Senate. In 2014, 91 percent of states did.
In 2016, that metric was 100 percent.
A straight-ticket vote was supposed to be terrible news for Senate Republicans. Most analysts figured Hillary Clinton was going to get more votes than Donald Trump in states with competitive races, and if those same voters marked the ballot for all Democrats, well, Senate Republicans' majority was doomed.
That clearly didn't happen. Republicans lost two seats in the Senate, but they'll hang onto their majority. It's possible Trump's strength with the Republican base brought out staunch Republican voters, stemming the tide of any split-ticket voters.
As House Speaker Paul D. Ryan (R-Wis.) said the day after the election: "Donald Trump provided the kind of coattails that got a lot of people over the finish line so the we could maintain our strong House and Senate majorities."
Or, maybe it was the opposite. Trump actually underperformed Republican Senate candidates in most of the competitive Senate races. The day after the election, Bump compared Trump's performance in 12 states to the Senate Republicans' performance in those states. He found that in all but three, the Senate Republican won a higher percentage of the vote than Trump did.
In other words, Trump's coattails were too short for most Senate Republicans to grab on to. It also suggests there are still some split-ticket voters out there, although not nearly as many as we had thought there were.
In New Hampshire, for example, Sen. Kelly Ayotte (R) outperformed Trump by .4 to .5 percentage points, but an average of polls in the week before the election suggested she could outperform him by as much as 8 points. (Ayotte ended up narrowly losing the election to Gov. Maggie Hassan (D) anyway.)
Many things about this election turned out to be the opposite of what was predicted. Sabato's team also pointed out that every one of the major analysts' predictions about who would win the presidency (The Fix included) got it wrong.
Which brings us back to the central question of this piece: Is the death of split-ticket voting an anomaly, or the new normal? The data suggest that crossing party lines at the ballot box is a trend that is on its way out, if not gone already.
But honestly, nobody knows for sure what's going to happen one election to the next. Certainly no one was predicting that one week after the election, we'd be noting the near-absence of split-ticket voting and that Senate Republicans had held their majority. And yet here we are.
It's a strange, unpredictable — and partisan — world out there.