2018 Midterm Projections
For better or worse, here are my projections for the midterm elections.
In just about twelve hours after this post goes live, voters will head to the polls across the country in what amounts to the first real electoral test for President Trump and the Republican Party after two years of one-party rule at both ends of Pennsylvania Avenue. For the better part of this year, we’ve seen the polls swing back and forth in both the Generic Congressional Ballot and the individual head-to-head Senate races that have been considered competitive. Initially, at least, the numbers seemed as though everything was breaking in favor of the Democrats, with polls indicating that they were headed for both pickup in the House of Representatives that would lead to a change in control in that body for the first time since 2010. In the Senate, the fact that voter enthusiasm seemed to be leaning in the Democrats direction suggested that Democrats would be able to hold on to many of the seats up for grabs in states that President Trump won in 2016 and that they would be able to flip a seat or two in states such as Nevada, Arizona, and Tennessee to gain control of the upper chamber as well. That began to change in the wake of the Kavanaugh hearings. At that point, Republicans began to become as enthusiastic about voting as Democrats were, and this caused polls in several states to start shifting in favor of Republicans. This has led to the probability, if not the likelihood, that we would end Election Night with a split decision in Congress, with Democrats controlling the House and Republicans hanging on to control of the Senate. Based on the late polling, that still seems to be the most likely outcome.
Initially, I was not going to post pre-election predictions. After getting burned along with pretty much every other pundit in 2016 by the outcome of the Presidential race, I’ve become far more circumspect about making predictions such as that. In the end, though, I decided to stick my neck out there just a little bit and let the chips fall where they may. As I noted on Sunday, the most likely outcome at this point appears to be one where the Republicans end up holding on to the Senate, and perhaps even picking up a seat or two in the upper chamber, while Democrats take control of the House of Representatives. By and large, that’s where my predictions for what things are likely to look at by the end of the day tomorrow land. However, given some of the polling trends we’ve been seeing at the last minute, there is at least some possibility that Democrats will do better in the Senate than expected, and that they might even be able to capture that chamber, albeit by a narrow margin.
In any case, on with the predictions.
House of Representatives Projection
There have been a few more polls handed down since yesterdays post and this has resulted in some changes in the polling averages and forecasts. So before we get to the prediction for the House, let’s review those final numbers:
- In the RealClearPolitics average, Democrats (49.7%) have a +7.3 point lead over Republicans (42.4%);
- In the Pollster average, Democrats (48.8%) have a +8.1 point lead over Republicans (40.7%);
- In the FiveThirtyEight average, Democrats (50.6%) have a +8.5 point lead over Republicans (42.1%); and,
- The FiveThirtyEight House Forecast gives Democrats an 87.5% chance at winning the House and gives Republicans a 12.5% chance of holding on to the House.
As the final RealClearPolitics chart shows, the Democratic advantage in the Generic Ballot has increased, slowly but steadily the closer we’ve gotten to Election Day:
in order to gain control of the House, Democrats need to pick up at least 25 seats, which would give them the 218 seats they need to form a slim majority that would, obviously, be vulnerable to change based on the retirement, illness, or death of a member in the middle of the upcoming new session of Congress that starts in January. Based on the polls and forecasts, the odds are that they will at least be able to pass this barrier. Beyond that, the projections lay out a serious of scenarios. Under the first scenario, Democrats pick up as few as 25-30 seats, which gives them a slim but workable majority in the lower chamber, Another scenario gives them something on a higher end such as a pickup of 30-40 seats, which would give them a much more comfortable majority. The final positive scenario for Democrats sees them picking up something closer to 50-60 seats, which would give them a strong majority that could theoretically stand through the 2020 elections. The final scenario is one in which Democrats pick up seats but that they fall short of the 23 they need to grab control. These last two scenarios, a Democratic wave akin to 2006 or 2010 or the GOP holding on to the House, strike me as being the least likely, although still possible, options.
Taking all of that into account, here’s my prediction for the House:
Current makeup of the House:
Republicans — 235 seats
Democrats — 193 seats
Vacant — 7 seats
Makeup of the House after Election Day:
Democrats — 230 seats
Republicans — 205 seatsNet Democratic gain +37 seats
This projection is, quite obviously, toward the upper end of the second scenario, but based upon where the polling is going I believe that it is a fairly good projection of where things will end up at the end of the night on Tuesday. While there are far too many battleground seats for me to make projections for each of them, I feel fairly confident in projecting that Democrats will succeed in winning at least two seats in Virginia, and stand to make significant pickups in New Jersey, New York, California, and Pennsylvania. These gains, combined with gains elsewhere around the country, will be a strong indicator of where the night is headed. For example, if you see Democrats picking up a majority of the four Congressional seats in Virginia that are deemed to be vulnerable in Virginia, and doing the same thing in New Jersey and Pennsylvania, then you can start to anticipate that it will very good night for Democrats. If Republicans manage to hold on to these seats, then the Democratic wave could be blunted and their majority, if it occurs, will end up being smaller than what I’m projecting.
U.S. Senate Projection:
As things stand right now, the odds are favoring the Republicans holding on to the Senate, thanks in no small part to both a map that has been favorable to them from the beginning and the fact that the Brett Kavanaugh hearings in September helped to boost the campaigns of several Republican candidates who were looking quite vulnerable over the summer. Looking at the projections, the FiveThirtyEight Senate Forecast gives Republicans an 83.2% chance of holding on to the Senate and Democrats just a 16.8% chance of capturing the Senate. The projection gives the GOP a 17.8% chance of keeping the 51-49 majority they have now, a 16.7% chance of increasing their majority to 52-48, and a 13.5% chance of getting a 53-47 seat majority, and a 15.3% chance of a 50-50 split that would result in the GOP maintaining control due to Vice-President Pence’s tie-breaking vote. The probabilities of other scenarios where the GOP increases its majority beyond 53-47 are all given less than a 10% chance as of today. The RealClearPolitics forecast meanwhile, puts 50 seats in the GOP column and 43 in the Democratic column, with 7 seats listed as “toss-ups.” Without toss-ups, the RealClearPolitics forecast puts the GOP at 51 seats and the Democrats at 49 seats.
Most of the Senate seats, of course, are basically already decided and I won’t be commenting on those, but here’s a look at where I think the races that have been garnering the most attention are likely to fall:
- Nevada: Dean Heller has long been seen as the most vulnerable of all of the Republican Senators up for reelection this year, and the polling in the Silver State has reflected that. As has been the case in much of the country, the state of the race changed in the wake of the Kavanaugh hearings and Heller actually enjoyed a period where he had a slight lead in the polls. That has changed in the closing weeks of the race, though, and Heller’s Democratic opponent, Congresswoman Jacky Rosen has moved back ahead to the point where the final RealClearPolitics average gives her a slim one-point lead. Based on this late shift of momentum, I’m going to predict that Nevada goes the way most analysts have assumed it would since the start of the year, Rozen 51% Heller 49% for a Democratic pickup;
- Arizona: The battle to the south in Arizona for the seat being vacated by Jeff Flake is similar to what we’ve seen in Arizona. Previously, Democratic nominee Kyrsten Sinema had been leading Republican Congresswoman Martha McSally. In the wake of the Kavanaugh hearings, though, the polls began to shift toward Congresswoman Martha McSally, the Republican nominee. Over the past week, though, the polling has taken yet another turn, and Sinema (47.8%) now has a slim +0.1 point lead over McSally (47.7%) in the RealClearPolitics polling average, This makes projecting the race basically a coin flip. Based on the fact that the post-Kavanaugh momentum for McSally has faded, though, I’m going to predict this seat flips, Sinema 51% McSally 49% for a Democratic pickup;
- Montana: Notwithstanding the fact that President Trump won here by more than ten points in 2016, Senator Jon Tester has managed to hold on to a lead here even in the wake of the Kavanaugh nomination, which he voted against. In the final RealClearPolitics average, Tester (48.5%) has a +4.5 point lead over Republican Matt Rosendale (44.0%). Based on this, and on the presence of a Libertarian on the ballot who will likely pull at least some support from Rosendale, I’m going to project that Tester pulls out a victory for a Democratic hold;
- North Dakota: Heidi Heitkamp has long been seen as the most vulnerable Democrat in the country, and that is justified by the current RealClearPolitics average which gives Republican Kevin Cramer (53.7%) has led Senator Heidi Heitkamp by double digits in some recent polling and he now has a +11.4 point average lead over Senator Heidi Heitkamp (42.3%). Count this as a Republican pickup;
- Texas: This is a race that many national Democrats have hoped would go their way even though Democrats have not won statewide in Texas since 1994 and have not elected a Democrat to the Senate since 1988. Admittedly, Beto O’Rourke is a surprisingly strong candidate for a Texas Democrat and has given Senator Ted Cruz a run for his money. Nonetheless, the polling in the race has shown Cruz maintaining a strong lead over his Democratic opponent, with the final RealClearPolitics average showing Cruz (51,0%) with a +6.0 point lead over O’Rourke (45.0%). Added into this is the fact that other statewide races show Republicans doing very well, which will probably give Cruz the stability he needs to pull off a win. Count this as a Republican hold;
- Missouri: Claire McCaskill has faced challenges before and managed to survive, but this time she has been facing a tough challenge from state Attorney General Josh Hawley. While some last-minute polls suggest that McCaskill could pull off a win, the last minute trend seems to be moving in Hawley’s direction. The final RealClearPolitics average here shows Hawley at 46.8% and McCaskill at 46.3%, giving Hawley a slim 0.5% point lead. This is a race that could go either way, but I’m going to predict a narrow win for Josh Hawley for a Republican pickup
- Indiana:, This race has seen a last-minute surge in the polling to incumbent Democratic Senator Joe Donnelly (44.0%), who now has a narrow +0.7 point lead over Indiana State Representative Mike Braun (43.3%) in a race that also include a Libertarian candidate who is standing at 4.5% in the RealClearPolitics average. Because of that, I’m going to predict that Donnelly holds on here for a Democratic hold;
- Tennessee, the Democratic nominee, and former Governor, Phil Breseden has lost the lead he once had over Congresswoman Marsha Blackburn and Blackburn (49.2%) now has a +5.2 point lead in the polling average over Breseden (44.0%). As a result, I’m going to go with Blackburn 52% Breseden 48% for a Republican hold; and,
- Florida, the picture is still looking good for Democratic Senator Bill Nelson (49.3%) has an average lead of +3.2 points over Florida Governor Rick Scott (46.1%). One interesting factor here is that Nelson seems to be being helped by a surge toward Andrew Gillum in the Governor’s race. As a result, I’m going to predict Nelson survives for a Democratic hold;
- Finally, I’m going to predict that Democrats will hold onto their seats in West Virginia, Wisconsin, Michigan, Ohio, and Pennsylvania, all of which were states that the President won in 2016 but where the Republican candidates never caught on for one reason or another.
All of this leaves us with this projection:
Current Senate Makeup:
Republicans — 51 seats
Democrats — 47 seats (plus two Independents)New Senate Makeup:
Republicans — 51 seats
Democrats — 47 seats (plus two Independents)Net Change: None
This projection includes the assumption that the Special Election in Mississippi to fill the remainder of Thad Cochran’s term will end up with a runoff later in the month between Senator Cindy Hyde-Smith and Democratic nominee Mike Espy, a race I expect Hyde-Smith to win easily.
Other Races
In addition to the races for the House and Senate, tomorrow also will decide the outcome of Gubernatorial and state legislative races in 36 states. I have admittedly not paid much attention to most of these races outside of the high profile races that have garnered national attention, including the following:
- Florida: In what may be the most high-profile race in the country, Tallahassee Mayor Andrew Gillum is facing off against Congressman Ron DeSantis, and the recent polling has all been pointing toward Gillum becoming the first African-American Governor of Florida. Currently, Gillum (49.4%) is leading DeSantis (45.4%) by +4.0 points in the RealClearPolitics average. Based on that I’m projecting that Gillum will win this race giving Democrats a pickup;
- Georgia: In this race, Stacey Abrams, who currently serves as Minority Leader in the state House of Representatives, is facing off against Republican Secretary of State Brian Kemp. In the current RealClearPolitics average, Kemp (48.5%) has a +2.8 point lead over Abrams (45.7%), However, there is a Libertarian candidate in the race drawing sufficient support to keep both candidates under 50%, meaning that this race is headed for a runoff;
- Wisconsin: Governor Scott Walker is facing yet another uphill battle in his bid for a third term, but given his electoral history and the fact that the most recent Marquette University poll shows the race as a tie, I’m going to give Walker the benefit of the doubt and predict a Republican hold
As for the rest of the nation, RealClear Politics projects that the balance of Governors nationwide will be 26 Republicans and 24 Democrats, meaning that Democrats would pick up as many as eight Governor’s Mansions.
Anyway, that’s where I stand. We’ll see after the dust settles if I’m right or not.
Thanks much Doug, this is a great compilation.
Frankly, I worry about any forecast Democratic advantage in toss-up states or districts that is less than 10% – primarily because Republicans have built a formidable structural advantage in states where they control the governor’s mansion and the legislature.
I hope those numbers hold up, but seeing is believing.
I’m convinced that pollsters are missing on the very real visceral reactions to Trump. I believe we will see a systemic polling error.
I predict the Dems win upwards of 40 seats in the House and that we end up with a 50-50 tie in the Senate or 51-49 Dems, with Heitkamp losing, but O’Rourke, McCaskill, and Bresden all winning.
Good for you Doug. Nothing ventured, nothing gained.
For my ownself… McCaskill wins by a squeeker. She’s tuff, Hawley has run a really weak campaign, and everybody hates him. Oh yeah, House goes DEM, Senate remains GOP. And Governerships go mostly D by 1 or 2, maybe even 3. State Legislatures even out, but Rs hold on.
I’m at least a little doubtful that this sort of middle of the road consensus forecast is what’s actually going to happen tomorrow night. I think we’ll end up on one end of the tail or the other … with either the GOP holding both chambers, or (more likely) a surprise (possibly big) on the upside for the Democrats.
I also think this will finally give us more of an answer to the “what really happened in 2016” question. If the GOP has a better than expected night, sadly it will be time to admit that we really do live in a country where Donald Trump’s nationalism resonates with a larger plurality of people than many of us want to believe. On the other hand, if Democrats have a great night, I think it will be an indication that negative feelings about Hillary Clinton had more of an impact on the 2016 election than anything Donald Trump may have done right in his own campaign (since he’s basically working out of the same playbook right now).
I hope voters will contemplate what an unholy union this is before they cast ballots.
Romeo and Juliet
Samson and Delilah…..
Pud and Kim YIKES!
@Todd:
That’s because it almost never does. Midterm elections almost always seem to defy polling forecasts to some degree (unlike presidential elections, where the polls are usually at least somewhat accurate). I think it has to do with the ephemeral nature of the national effects combined with the relatively small sample sizes in state and local races.
The projections mentioned are probably as good as can be made based on polls and earlier special elections. I have some doubts that conventional and time-tested polling is going to be worth it in this election. A TeeVee show today claimed that in Texas the early votes from the 19 to 34yr old group exceeded the same cohort’s voting in ’14 by 447%. Almost 5 times more younger voters cared enough to vote early!
I think this election is going to have a very different electorate than anything we’ve seen. I’m hopeful and fearful and completely agnostic about the outcome.
I am hoping for a few “wow, didn’t see that coming” races. Admittedly, at the top of that list is Beto over Cruz, mostly because I’ve found Cruz to be irritatingly smug and would love to see that ego shut down.
Aside from that, I’ll be too nervous to watch. It will likely be an early evening for me, especially considering some of these races could hinge on absentee and provisional votes, which could take days to finalize and certify. Ignotum per ignotius.
@JohnMcC:
Hopeful and fearful yes, but for me agnostic might not be the best word. I think a good night for Democrats probably tells us less about how “good” the future may play out, than a better than expected night for Republicans tells us about how much worse things are likely to get. For instance, if the GOP holds the House, Sessions, Rosenstein and probably Mueller himself will be fired before Christmas (and maybe even Thanksgiving) … of that I have almost no doubt.
Thanks, Doug. I largely agree with you.
@Mister Bluster: Worse than that–
https://www.washingtonpost.com/world/asia_pacific/north-korea-threatens-to-restart-nuclear-program-unless-us-lifts-sanctions
@Jen: I don’t see any particular rationale/advantage/utility in being nervous about the outcome of this race. As Todd noted, as a nation, we are either the nation that those in the center and moving leftward hope we are, or we are people who deliberately and with forethought elected Trump believing that what the GOP stands for is the correct direction. I’m afraid it really is that simple, and if we are Trump’s nation, those who want a better nation will have to ride this out until the GOP destroys itself.
Or go to Plan B if they think the GOP is durable.
I don’t see any reason to change my prediction from way back (when I first realized that there were three times as many Dem Senators up for re-election as Republicans): the Dems will end up with a 12 vote majority in the house and will lose 2-3 Senate seats.
@MarkedMan:
Wow, that Senate result (while possible) would actually be a pretty bad night for Democrats. Most plausible scenario that gets to -3 is they lose ND, MO & IN and also fail (just barely) to pick up the GOP seats in AZ and NV. Given that Democrats appear to have whatever late momentum there is, this seems unlikely. But I guess we’ll all have a much better idea in about 24 hours.
I expect the unexpected.
I don’t think the likely voter models are going to be accurate this year, and figure anything could happen.
I will give two predictions: GA Gov to Kemp, because of effective vote suppression; KS Gov to Kobach with a wider margin. And lots of pissed people and protests, but no significant violence.
@Jen: Beto is working hard to get people out to actually vote but I don’t think it’ll be enough. He’s going to lose by 2 percentage points.
I predict that the Democratic party barely picks up a majority in the house while losing a seat in the senate.
Sen 51 Ds House 250 Ds
@Matt:
https://www.vox.com/policy-and-politics/2018/9/21/17852260/beto-orourke-cruz-down-ballot-texas-senate
@Kylopod: Ah yes I have pointed that out in the past. That just getting close is a major victory considering.
There will be only one Republican this year that will get my vote here in Texas this year. A county judge that I saw in action.
These last two scenarios, a Democratic wave akin to 2006 or 2010 or the GOP holding on to the House, strike me as being the least likely
The Democrats only picked up 31 seats in 2006. You are predicting 37, which is actually a bigger wave.
They picked up another 30-odd in 2008, so 2006 and 2008 together was a massive wave. But wait until 2020…!
@Just nutha ignint cracker: Oh, I agree. Zero utility in being nervous…but I am an anxious sort anyway, it’s sort of baked in to my personality. 🙂
@Todd:
If that happens, it will be time to write America’s obituary.
As usual, I predict all the close senate races end up breaking the same way, so either Dems pick up 2 or Repubs pick up 2, but we don’t end up with the same balance we gave now.
There is far too much in the wind for anyone to predict anything with any amount of certainty, so I’ll just go with this;
Dems win
Dennison claims victory
On to 2020!!!
@Kathy:
Sadly, I have to agree with this.
An emboldened Republican party will not be able to contain themselves.
What they haven’t destroyed already, they will.
@Todd: @Kathy:
There is in todays NYTimes (I know, I know) an OpEd (“The Beginning of the End for Poland’s Populists”) by some foreign gent (Slawomir Sierakowski) detailing the very recent election there that has effectively shown the power of the vote in turning these Bannonesque neo-fascists out of gov’t.
I’m still ‘agnostic’ (grasping uncertainty with a death grip) but find a few encouraging signs of approaching better days.
@Kylopod :
Texas is slowly and inevitably turning purple and that’s *terrifying* the GOP. Politically, Texas is their big stronghold, their reservoir of electoral votes and fall back position. Culturally, it’s their touchstone and place that will *always* be theirs by nature of its people and ideology. They may lose AL or ND for an election but TX will stay red by god! It would be the equivalent of watching liberal California slowly turn to Redneckistan for Dems – a huge blow to self-image and power.
The problem the GOP has is Texas is big and it’s filling up with people. More people traditionally means more like to go Dem – liberals live in cities and higher populous states, conservatives tend to like rural and less crowded areas. Beto has incredible energy because enough people are living there to resonate with what he’s saying and still view themselves as Texan. Republicans that can see the writing on the wall know that if he doesn’t win today, he’s going to win next time. The thing about waves is that there isn’t just one – it’s a series of impacts big or small that continually wears away the shore over time. Being Texan no longer means automatically being Republican…. and that’s a death knell for them.
There may come a time in the near future that Texas is in play more often then not.
@Daryl and his brother Darryl:
Nothing lasts forever.
A broad generalization would be that states promise something to their constituents, and they thrive as long as they deliver. The promises may be implicit or explicit, or partly both.
America promises, broadly, legal equality and economic opportunity. The former wasn’t available to large portions of the population for a very long time. The latter is becoming scarcer now, and some in power are dialing back on equality, or trying to, as a means to appease their favored groups for lack of economic opportunity.
This will not end well.
There is a lot of things making this a tough election to predict:
– Bigotry usually under polls, sometimes very substantially. People may be reluctant to state that they support racists and cheats, but may end up pulling the lever anyway.
– The increasing spread of early voting may be having some kind of effect, but what is that, exactly?
– Rain across much of the country on Thursday
But one thing that is easy to dispense with: that Trump is doubling down on bigotry, fear and misogyny because he is uniquely insightful into the American mind. He may be right in this case or he may not, but the proof that he is not uniquely insightful on anything lies in the wreckages of his businesses strewn across the past four decades. Trump, never, ever, knows when to quit. He always doubles down on losing enterprises. Heck, his TV show is a good example. It had a couple of good years, and it was kept around for quite a while because, as a reality TV show, it was very cheap to make. But as it steadily slid down the ratings ladder year after year, Trump’s genius idea every season was: “More Trump!”. I would be willing to bet the only reason he agreed to do the “Celebrity Apprentice” was because he would have been cancelled otherwise. At some point even a low cost show isn’t worth keeping on. But it must have killed Trump to think he would have to share camera time with people more famous than him. No worry there! The celebrities ended up being washed up hacks or one shot wonders. Gary Busey, anyone?
Casinos, airlines, bottled water, urine tests (Yes. Urine tests), and the list goes on. Trump either stays in too long or gets kicked out when his partners realize he is a deadweight and incredibly annoying to boot. The idea that Trump has some great insight is ludicrous. After he burned through his father’s money, he found himself a nice little gig helping Russians launder money. Being born to a rich father and becoming the Russian Mob’s tool doesn’t say anything about insight.
@KM: “Texas is slowly and inevitably turning purple…” Indeed. There is a theme from the Obama years that demographics will prevail in the end and that the proof will be Texas and Georgia moving into states Dems regularly win because Hispanics, blacks and urbanization.
We sure are closer to that moment than any of us have ever seen. I just hope it’s true.
@One American:
Not as much of a fatalist as you are. I tend toward human fallibility as the determiner rather than some sort of cosmic intent. But thanks for the good wishes.
@KM: Remember it wasn’t that far back when California regularly flipped.
And there are enough pissed-off ex-Puerto Ricans in Florida that we might start seeing FL as reliably blue.
@Todd: well shit, it looks like @MarkedMan: predictions might not have gone far enough when it comes to GOP Senate pickups.
@Matt: Beto lost by 2.6 percentage points so my guess was quite close.
I guess I wasn’t pessimistic enough on the senate side of things.
So whoever downvoted me… neener neener!