Heading into the 2018 election, Democrats have the wind at their backs. Republicans are stuck with an unpopular president, but are mostly afraid to distance themselves from him for fear of alienating “The Base.” All polling data, the evidence from recent special elections that Democrats have won in unlikely places, and the history of midterms (which show that the sitting president’s party usually loses ground in the midterms), all suggest that, if the election were held today, Democrats would be favored in a lot of races.
Without question, as of now, Democrats have a big lead in enthusiasm, which is evident not only in polling but in turnout for all the recent special elections, new registrations, appeal to young voters and others who have not always participated, and pretty much all the ways that political analysts attempt to discern the direction of the political winds.
If you aren’t cautious, you could convince yourself that Democrats are holding a pat hand. (That’s like getting dealt a straight or a flush in poker.)
But the Dems are not holding a pat hand. (In the fall of 2016, a lot of analysts thought Hillary Clinton was holding a pat hand. The excellent analyst Stu Rothenberg wrote in mid-October 2018 that Donald Trump’s path to victory was not “narrow,” but “nonexistent.”) I respect Rothenberg and often learn from his work, but in writing that particular column he went beyond mere analysis and convinced himself that he knew the future. He may be able to make better educated guesses about what will happen than I can, but neither he nor anyone else can see the future clearly.
Personally, in my little corner of the commentariat, I believe we should make clear that we don’t know the future. But if we did that, we probably wouldn’t get invited to be on the shows.
The lesson of 2016 is that nothing is in the bag until they count the votes. But, heading into campaign 2018, national generic ballot surveys (“if the election was held today would you vote for the Democrat or the Republican?”) look very good for the Dems (ranging from about a plus-five to a plus-11 points.)
At least three problems complicate the simple narrative of a great 2018 outcome for the Dems, leading to a blue takeover of Congress. The first I already covered: The future is unknowable. The second one is the U.S. House. The third one is the U.S. Senate.
The House: Dems need a net pickup of 24 seats
The problem in the House is that Democrats would need a net pickup of 24 seats (out of 435) to take control. (There are several open seats. For purposes of this calculation, I am treating those seats as if they are held by the party that last held them.)
It’s somewhere between a rule of thumb and an iron law that the party holding the presidency is likely to lose House seats in the midterm. It’s a very powerful tendency. In the last 20 midterms, the president’s party lost House seats 18 times. (OK, so it’s not an iron law. It’s a likelihood that occurs with about 90 percent frequency.
But the size of those losses for the president’s party is all over the map, from a minimum of four (in the 1962 midterm with John F. Kennedy in the White House) to a high of 63 (lost by the Democrats in 2010, the middle of Barack Obama’s first term, which led to Republican control of the House ever since).
Just based on the history of the recent similar cases, 24 is a gettable number but not an easy one. I looked just at post-World War II midterms occurring during the first term of a Republican president (which is what we have this year). There were six such examples, all coinciding with a Republican net loss in House seats in the middle of a Republican president’s first term, but in just two of those six cases was the net Republican loss in the midterm greater than 24.
I hear people who are hoping for that outcome talking about it with too much confidence. Given everything we know at the moment, it seems definitely possible, but hardly a cinch, that Democrats will accomplish a net pickup of 24 or more seats.
If you look at Real Clear Politics’ aggregation of eight recent polls that asked variations of, “If the election was held today, would you vote for a Democrat or a Republican for the U.S. House?” they all show an edge for the Democrats. But the size of the edge ranges from five to 11 percentage points.
Also, rather obviously, that question is not what voters will see on Election Day. They will face a choice between a specific Democrat and a specific Republican, and they haven’t all made up their minds about how (or even whether) they will vote.
The trouble with excessive certainty about the Dems’ chances of taking control of the U.S. House is that Democrats need a net gain of 24 specific seats, featuring not a generic Democrat versus a generic Republicans, but a specific Democrat and a specific Republican.
(An aside the improves the Dems’ chances: The new House district map in Pennsylvania, drawn by the courts after striking down the old map for excessive gerrymandering, could reasonably help the Democrats pick up three-to-six currently Republican-held seats, just in that state.) That could be a decent chunk of the 24 sets the Dems need to take over the House.
My advice: Between now and November, you may see many stories speculating on whether the Dems are likely to pick up a net 24 House seats. Keep your shirt on, cut the cards, and bear in mind that no one really knows. (Reminds me of an old crack from a Republican operative in Arkansas in my early days as a reporter down there. I asked him what was going to happen in a particularly close election. He replied: “No one knows that but me. I could tell you, but then I’d have to kill you.” A little googling suggests that this wisecrack goes at least as far back as “The Hound of the Baskervilles.”)
The Senate: There’s a catch
Now onto the U.S. Senate. The good news for Democrats is that (counting as Democrats the two independents who caucus with the Dems) the current division of that chamber, after the recent upset in Alabama, is just 51-49 in favor of the Republicans.
As you know, senators serve six-year terms but federal elections occur every two years. That means that in every federal election year, just one third of the Senate seats are on the ballot.
(Then you have to account for special cases, such as we have in Minnesota, where the appointment of interim Democratic Sens. Tina Smith to temporarily hold the seat vacated by the resignation of Democrat Al Franken generates an additional special election to fill that seat for the unexpired portion of the Franken term). Counting the two special elections (the other one is in Mississippi), there will be 35 Senate seats on the November ballot.
Still, a net gain of a just two Senate seats by the Dems in November would result in a 51-49 Democratic majority. And, as with the House, the party controlling the White House tends to do poorly in midterm elections for the Senate.
A net gain of two seats sounds a lot easier than the challenge facing Dems in the House. But there’s a catch, a big and obvious one, and I’m surprised at how seldom the pundits bring it up. It’s this: Because Democrats had a very good year in 2012 Senate races — winning an overwhelming share of the seats up that year, when Obama was also re-elected — there are just nine seats currently held by Republicans on the November ballot, compared with 26 seats on the ballot that are currently held by Democrats.
That’s why, although it’s true that the Dems need a net gain of just two, the New York Times recently run an article headlined: “Democrats Need to Win 28 Seats to Control the Senate. Republicans Need Only 9.”
That’s the same as saying that, because the Dems already control 26 of the Senate seats that will be on the ballot, they make no gains at all until they win a 27th race out of the 35 races on the 2018 ballot. Whereas Republicans will suffer no net loss at all if they win just nine of the 35 seats on the ballot. If they win just eight, that will create a 50-50 Senate, and Republicans will control a tenuous control because the vice president has the power to vote in case of ties.
Lastly, if you have digested the pure math that one could say favors the Republicans to retain control of the Senate, there’s the question of the lineup itself of the races that are on the ballot. I won’t try to capsulize 35 races. But consider this fact:
Only one of the incumbent Republican senators who are up in November represents a state that Hillary Clinton carried in 2016. That would be Sen. Dean Heller, Republican of Nevada. Clinton beat Donald Trump by just 47.9 to 45.5 percent in Nevada, so it’s not exactly a solid blue state. It’s purple. The pundits who rate all the Senate races generally believe the Nevada Senate rate is “toss-up” at present.
So if the winds blow blue in November, the Nevada seat might be one of the two current Republican-held seats that Democrats need to flip to take over the Senate. Arizona is a state that Trump carried narrowly in 2016. And its Republican senator, Jeff Flake, is retiring. The pundits also rate that one a “toss-up.”
After that, the Democratic pick-up opportunities get slim. The seat in Tennessee, where incumbent Republican Bob Corker is retiring, seems the next most likely pick-up opportunity for Democrats. But Trump carried Tennessee by 61-35 percent, which illustrates how slim the pick-up pickings are for Democrats.
By contrast, there are seven – yes, seven – states that Trump carried in the presidential race where Democratic senators are up for re-election this year. The senators are:
- Claire McCaskill of Missouri (who is generally rated a “toss-up” bet for re-election);
- Jon Tester of Montana (generally favored by the raters, but far from safe);
- Debbie Stabenow of Michigan (rated safe to likely for re-electon);
- Joe Donnelly of Indiana (rated a “toss-up” by all the raters, Trump carried Indiana by 57-38 percent);
- Heidi Heitkamp of North Dakota (Trump by 63-27 percent in North Dakota, Heitkamp is rated a “toss-up” by all);
- Sherrod Brown of Ohio (Trump won Ohio by 52-44; Brown is favored, but only slightly, by the raters);
- Bill Nelson of Florida (Trump won Florida by 52-44; the raters call Nelson’s re-election either a toss-up or leaning toward Nelson).