Tuesday marks the end of the 2018 midterm election cycle. Political scientists hesitate to say things like, “Donald Trump is on the ballot” when he is actually, you know, not on the ballot. We tend to see congressional elections as local affairs, driven more by statewide dynamics than the occupant of the White House. At the same time, this is an unusual election cycle where the President has inserted himself into the campaign forcefully. He is polarizing and salient. Like it or not, he is on the ballot within the minds of many voters. Mr. Trump’s agenda is clear. Americans will choose whether or not to support that agenda as they elect U.S. Senators and Representatives across the nation.

In my previous post on the midterms, I looked at the relationship between presidential popularity and aggregate wins and losses during the midterms. Based on history, we might expect the Democrats to pick up around 25-30 US House seats and 4 Senate slots. Most forecasts suggest we will see something like that in the US House, which would swing control of the chamber to the Democrats. These forecasts hinge on the accuracy of somewhat infrequent polls with somewhat limited sample sizes. Also, if the factors that biased the polls toward Democrats are still present in 2018, we could see a surprise result if Republicans improve their poll numbers. If we are witnessing a shift between the parties, where voters are changing loyalties over the long haul, our results could be somewhat confounding for polls built on statistical models from 2014, 2010, and 2006. Even though we measure public opinion in the here and now, the assumptions those polls are built on are historically rooted. We will see.

In Ohio, keep a close eye on OH 1 and OH 12. In 1, Chabot has won some competitive contests and underperformed both Trump (65% for Trump, 59% for Chabot) and Romney (59% for Romney, 57% for Chabot) in his congressional district in 2016 and 2012. In 2014, he won comfortably with 63% of the vote. If Chabot loses or wins in a very tight contest, Republicans in Ohio should start to worry, because it suggests the midterm electorate, which is generally more favorable to Republicans because they vote at higher rates, has shifted.

In Ohio 12, we see a repeat of the recent special election between Balderson and O’Connor. Balderson won the earlier contest in a squeaker (with 50.1%). External money has poured into the district. Pat Tiberi, who stepped out of the seat, which initiated the special election, won his last midterm contest with 68% of the vote. He outperformed Trump in his district in 2016 (67% to 53%). But Tiberi is not on the ballot. If the GOP wins OH 1 and OH 12 comfortably, the GOP will probably have a good night in OH. If they lose one or both, things may go horribly off the rails and Democrats will likely be in for a resurgence in the Buckeye State. I tend to think Ohio is trending red and is moving away from its solidly moderate, middle-of-the-road political status and toward the right. If I am correct, expect the Republicans to do well in the state regardless of national results.

Almost no one at this time believes the Democrats will take control of the US Senate. Let’s look a little more closely at the political dynamics behind these Senate forecasts. Given the presence of two special elections, we have 35 seats up for grabs. Of the 65 seats NOT up for election, 42 are GOP and 23 are Democratic. The Democrats need to win 28 of the 35 elections to take control of the chamber, while the GOP must only secure 8 of the 35 seats since a 50-50 chamber reverts to the GOP given Mike Pence’s tie-breaking vote. Advantage Republicans.

Of these 35 contests, five appear to be locks for the Republicans–UT, WY, NEB, MISS 1, and MISS 2.* That pushes the party up to 47 seats, so they only need to win 3 more to maintain power. The Democrats have 19 safe seats, which pushes them up to 42. They have to win 9 more seats to take control–they need 51 in total. With these safe seats in mind, that leaves us with 11 competitive races. The Democrats must win 9 and the Republicans must win 3. Here they are (with the party that possesses the seat or used to possess the seat if it is currently “open”):

AZ (R), FL (D), IN (D), MO (D), MT (D), ND (D), NJ (D), NV (R), TN (R), TX (R), and WV (D).

On its face, this looks possible for Democrats–hang on to the seats they have plus flip two others. In a typical “wave” election, this is how chambers change hands. One party dominates “close” elections. The reality is much more difficult for Democrats. They are defending 6 seats in “red” states that Donald Trump won, some of them handily, in 2016 (FL, IN, MO, MT, ND, WV). When we look at all 11 together, Trump won 54.5% of the vote in 2016. Democrats would need to endure difficult political environments just to hang onto these seats, much less flip others.

I think the Republicans are in an excellent position to pick up North Dakota and defend in Texas. That would leave them with only one more victory to maintain power. The Democrats are likely to win New Jersey. Assuming Republicans win in North Dakota and Texas, that would require them to sweep the remaining contests to get to 51 seats. Advantage Republicans.

I think the most likely outcome is for the Republicans to win in ND, TX, TN, and MO. If this happens, it is a status quo election. If they manage to win any of the other contests, they will build their majority. If the Democrats overcome these obstacles to take control of the Senate, we will have witnessed a significant political event. If Democrats pull an upset in Texas and hang onto North Dakota, anything is possible.

An interesting follow-up would be to measure the extent to which these Senate Republicans mimicked the politics of Trump. I have followed some of the campaigns closely (like Renacci’s challenge to Brown here in OH), but not enough to characterize the tone and method of the campaigns themselves, at least beyond just basic observation. A close look at advertising, speeches, and social media usage would be needed to draw firm conclusions.

Wednesday morning will bring some clarity. Then, may God be with all of us, the presidential campaign will begin in earnest.

*Mississippi has two Senate elections on Nov. 6. One is a special primary to fill Thad Cochran’s seat, which was vacated for health reasons. If no candidate gains a majority, a run-off will happen on Nov. 27. The other seat, which is normally up this cycle, belongs to Republican incumbent Wicker. This is considered a safe Republican seat.