After a very tiring two years, Election Day is almost upon us. As we’ve watched the polls over the last month, we’ve seen some different trends emerge in the various categories of races: for the House, governors, and the Senate. Maybe most importantly, from the perspective of immediately checking Republican power, the Democrats have put themselves into a position to take back control of the House of Representatives.
Generic ballot polling hasn’t really changed much in the last month (or the last year, really), but recent polling in individual House races, as well as watching which races the parties are spending their money on, seems to support more, not less, confidence in Democrats gaining the 23 seats necessary for flipping the House, and probably many more (though a gain of, say, 30 seats still seems likelier than 40).
From the longer-term perspective of rebuilding Democratic strength—through preventing gerrymandered House maps in the next decade, and signing legislation that preserves voting rights into law—the gubernatorial races may be just as if not more important, and that’s the area where the Democrats look poised to flip a historically notable number of seats. They’re currently on track, in our polling averages, to take around eight state houses from the GOP, with several more in striking distance. Keep in mind that the largest number of gubernatorial pickups in an election in the post-World War II era was 11, for Republicans in 1994.
The Senate races were always going to be tough sledding, even with this year’s dynamics working in Democrats’ favor, given that this year’s map is a rebound from the Democrats’ strong overperformance in 2012. This leaves them defending a number of incumbents in red states, with only a handful of Republican-held seats being on the table at all in states that are even remotely swingy.
In a way, it’s very encouraging that we aren’t looking at significant losses in the Senate, which is probably what would happen if a Democrat were in the White House; the likeliest outcome in the Senate may be no net change whatsoever (with the total staying at 49 Democrats plus independents, and 51 Republicans). A net gain of one or a net loss of one are also likely outcomes; what’s not especially likely is finding a way for Democrats to gain two seats, necessary for fielding a majority with a Republican vice president.
In our final forecast, we’re going to go into more detail about all three of these categories of races, starting first with the Senate.
SENATE
As we said above, the most likely outcome in the Senate, based on our polling averages, is no net change in the number of seats the parties hold: the Democrats (and friendly independents) enter with 49 seats, and leave with 49 seats. That doesn’t mean that no seats change hands; instead, what would happen would be that Democratic candidates would pick up one or two Republican-held open seats in swing states—Arizona and Nevada—while losing one or two incumbents in much redder states, most likely North Dakota and, not quite as likely, Missouri.
Here’s one last look at the Senate “totem pole”:
STATE |
D CAND. |
D AVG. |
R CAND. |
R. AVG. |
DIFF. |
FLIP? |
OHIO |
Brown (inc) |
48 |
Renacci |
37 |
+11 |
|
WISCONSIN |
Baldwin (inc) |
53 |
Vukmir |
42 |
+11 |
|
WEST VIRGINIA |
Manchin (inc) |
49 |
Morrisey |
38 |
+11 |
|
MINNESOTA (SP.) |
Smith (inc) |
48 |
Housley |
30 |
+9 |
|
NEW JERSEY |
Menendez (inc) |
48 |
Hugin |
40 |
+8 |
|
MONTANA |
Tester (inc.) |
48 |
Rosendale |
42 |
+6 |
|
INDIANA |
Donnelly (inc.) |
44 |
Braun |
42 |
+2 |
|
ARIZONA |
Sinema |
47 |
McSally |
46 |
+1 |
D FLIP |
FLORIDA |
Nelson (inc.) |
47 |
Scott |
47 |
0 |
|
MISSOURI |
McCaskill (inc.) |
46 |
Hawley |
46 |
0 |
|
NEVADA |
Rosen |
46 |
Heller (inc.) |
46 |
0 |
|
RED LINE |
RED LINE |
|
RED LINE |
|
|
|
TEXAS |
O’Rourke |
47 |
Cruz (inc.) |
48 |
-1 |
|
TENNESSEE |
Bredesen |
42 |
Blackburn |
50 |
-8 |
|
NORTH DAKOTA |
Heitkamp (inc.) |
41 |
Cramer |
51 |
-10 |
R FLIP |
MISSISSIPPI |
Espy |
27 |
Hyde-Smith + McDaniel |
49 |
-22 |
|
The only one of those seats that feels like a foregone conclusion is North Dakota, where Democratic incumbent Heidi Heitkamp hasn't led a poll since February and the polling average puts her down double digits. If she’s going to pull this one out, there would have to be either a significant polling error across pollsters or a late-breaking surge that hasn’t shown up yet in the polls. (Keep in mind that Heitkamp did win in 2012 despite trailing in polling averages, so there’s precedent for that happening. But she wasn’t this far behind in the averages that year.)
Missouri, by contrast, is not that clear at all. Democratic incumbent Claire McCaskill has crept back into a tie with Josh Hawley after several weeks of trailing by several points in the averages; this is a race where Republican-affiliated pollsters have been “flooding the zone” but most polls from nonpartisan pollsters have shown the race tied or given McCaskill a small lead. She’d still need undecided voters to break pretty significantly in her way, in order to win; a Democratic incumbent in a red state polling at only 46 in the days before the election should, if nothing else, not be putting down a deposit on next year’s drapes for her office.
One other cause for concern, even if McCaskill pulls it off, is that there are more races where there’s the potential for things to go wrong for the Democrats than there are for the Republicans. In other words, there are four races with a narrow Democratic lead or a tie (Arizona, Florida, Indiana, and Nevada), compared with one GOP race that is similarly close (Texas).
The circumstances in each case are a little different, though. In Indiana, the race isn’t polled that frequently, and there was a cluster of GOP pollsters flooding the zone in that race two weeks ago, as well; however, several national nonpartisan pollsters found Democratic incumbent Joe Donnelly with a tolerable lead last week. Indiana’s a red-enough state, though, that undecided voters may yet break against Donnelly.
By contrast, Florida is polled more frequently than any other state, and there, most pollsters are finding incumbent Bill Nelson with a one- to two-point lead (and often at 48 or 49 percent, closer to the 50 percent mark); a few polls still bounce up with Rick Scott having a one- or two-point lead, though, so it’s currently tied in our average. And while pollsters have given Democratic candidate Jacky Rosen only very small leads in Nevada (and for several weeks in October, she was outright trailing in the averages), Nevada’s detailed reporting of the early vote points to the possibility that Democratic voters have baked in an early voting lead that will be difficult for Republicans to overcome on Election Day.
Finally, there’s the matter of the two other “Lean Republican” seats, Tennessee and Texas, which are important because if the Democrats are going to run the table and find that 51st seat after saving McCaskill and winning all the other closer races, it’s going to be one of them (unless North Dakota truly surprises us). Tennessee, which took a wrong turn during the Kavanaugh confirmation saga but where a few polls have seen a tie in the last week, seems to be moving back in Phil Bredesen’s direction—but he may have gotten in too deep a hole in October to climb out in time.
Meanwhile, in Texas, the trends have been a little flatter, though with Beto O’Rourke gaining some significant ground in the last week, mostly due to one poll from Change Research. Texas currently looks like a better pickup opportunity than Tennessee; either way, a win in either one would be the final move in what would already be a pretty unlikely triple bank shot for the ultimate victory.
Finally, we’d be remiss if we didn’t mention the Mississippi special election (to replace Republican Thad Cochran, who resigned for health reasons). The election on Tuesday here will be something of a Louisiana-style jungle primary, where, assuming no one hits 50 percent, the top two finishers will advance to a runoff on November 27. So if the Democrats emerge with a net gain of one (taking them up to 50-49) on Tuesday (and assuming Democratic ex-Rep. Mike Espy is one of the participants in the runoff, presumably against appointed incumbent Cindy Hyde-Smith), look for everyone’s attention (and money) to shift to this previously sleepy race.
GOVERNORS
If the Democrats are likely to fall narrowly short in the Senate, the category where they’re poised to overperform the most is probably the governor’s races. There’s no one target number in the governor’s races where there’s a definitive sense of having “won” or “lost” the whole thing, but Democratic candidates are on track to pick up around ten seats, which would be a historically large number exceeding even other relatively recent wave elections (like 2006 or 2010).
And most importantly, these aren’t just dinky little states; they include some of the most populous ones in the nation that are currently under Republican trifectas—like Florida, Michigan, and Ohio—where a Democratic governor would not only be able to stop the Republican legislative policy agenda but also force a compromise or court-drawn House map for the coming decade.
Here’s one last look at the “totem pole”:
STATE |
D CAND. |
D AVG. |
R CAND. |
R AVG. |
DIFF. |
FLIP? |
ILLINOIS |
Pritzker |
48 |
Rauner (inc.) |
32 |
+16 |
D FLIP |
MAINE |
Mills |
51 |
Moody |
38 |
+13 |
D FLIP |
RHODE ISLAND |
Raimondo (inc.) |
43 |
Fung |
33 |
+10 |
|
MICHIGAN |
Whitmer |
47 |
Schuette |
38 |
+9 |
D FLIP |
NEW MEXICO |
Lujan Grisham |
52 |
Pearce |
43 |
+9 |
D FLIP |
COLORADO |
Polis |
48 |
Stapleton |
40 |
+8 |
|
MINNESOTA |
Walz |
47 |
Johnson |
40 |
+6 |
|
CONNECTICUT |
Lamont |
44 |
Stefanowski |
39 |
+5 |
|
OREGON |
Brown (inc.) |
45 |
Buehler |
41 |
+4 |
|
OHIO |
Cordray |
44 |
De Wine |
41 |
+3 |
D FLIP |
FLORIDA |
Gillum |
47 |
De Santis |
45 |
+2 |
D FLIP |
WISCONSIN |
Evers |
48 |
Walker (inc.) |
46 |
+2 |
D FLIP |
KANSAS |
Kelly |
42 |
Kobach |
41 |
+1 |
D FLIP |
IOWA |
Hubbell |
44 |
Reynolds (inc.) |
44 |
0 |
|
NEVADA |
Sisolak |
45 |
Laxalt |
45 |
0 |
|
SOUTH DAKOTA |
Sutton |
44 |
Noem |
46 |
-2 |
|
ALASKA |
Begich |
44 |
Dunleavy |
47 |
-3 |
|
NEW HAMPSHIRE |
Kelly |
43 |
Sununu (inc.) |
47 |
-4 |
|
OKLAHOMA |
Edmondson |
43 |
Stitt |
47 |
-4 |
|
GEORGIA |
Abrams |
43 |
Kemp |
51 |
-8 |
|
MARYLAND |
Jealous |
35 |
Hogan (inc.) |
52 |
-17 |
|
The race that’s the closest to being a foregone conclusion is Illinois, where Republican incumbent Bruce Rauner won only narrowly against an unpopular Democratic incumbent amidst a GOP-friendly year in general, in 2014. Illinois is clearly reverting to its blue-state form, with Democratic challenger J.B. Pritzker having consistently led Rauner by 15 to 20 points throughout the year. New Mexico is another blue state whose gubernatorial race has gotten little attention throughout the year, with Democratic Rep. Michelle Lujan Grisham leading consistently by significant margins in what few polls we’ve seen.
Political observers were expecting Michigan to be one of the most competitive gubernatorial races this year, but it hasn’t really turned out to be: since consolidating Democratic voters after the primary, Democratic candidate Gretchen Whitmer has consistently led Bill Schuette, who’s been running a decidedly lackluster campaign. Some polls have seen Schuette closing within single digits recently, after being down more than 10 before that, but he’s still at a distinct disadvantage with no time left on the clock.
Another light-blue state that took rather the opposite direction is Maine, where for many months the race looked like a tie, though that was largely a factor of how few polls we’d seen; in recent weeks, though, Democratic candidate Janet Mills seems to have built a substantial lead.
Two more races where the Democratic advantage isn’t quite as wide are Florida—where Democratic candidate Andrew Gillum has had a pretty consistent lead since surprisingly winning the primary, though the very frequent polls disagree whether he’s up only 1 or 2, or whether he’s outpacing ticket-mate Bill Nelson by a few points more—and Ohio, where, unlike Florida, we haven’t see as many polls, but where most of them in the last few weeks agree that Democratic candidate Rich Cordray is edging out Mike De Wine.
And finally, there are several more races where the Democratic lead is very thin: Kansas and Wisconsin. (And some of those races may be at the top of your priority list—what with the opportunity to knock off odious opponents like Kris Kobach and Scott Walker—more so than some of the less memorable ones where there’s a larger lead.)
That adds up to “around” eight; it’s not clear the Democratic candidate is likely to win in all of those, but there are several more tied or close, poorly polled races that could compensate for a loss in any of those, like Alaska, Iowa, South Dakota, and Nevada (where, in fact, the early vote numbers could presage a win by Democratic candidate Steve Sisolak even though he has trailed most polls and is now tied in this week’s polling average). In fact, nine or ten seem likelier than, say, six or seven.
Keep in mind that Georgia’s unique law might force us into overtime here, as well: if no one tops 50 percent, the top two finishers will be forced into a runoff on December 4. With a Libertarian candidate on the ballot, and with Democratic candidate Stacey Abrams neck-and-neck with Brian Kemp in the polling averages, we might need to do it all over again next month. (Also, keep in mind that the Georgia race was tied in our averages the previous week; a GOP internal poll from Trafalgar with ludicrous numbers pushed our averages down significantly.) The good news is, with national attention focused purely on Stacey Abrams, if that happens, she certainly won’t lack for the money needed to fight the final battle.
HOUSE OF REPRESENTATIVES
The battle for the House has received the most ink, arguably, if only because it has long been the body of government in this election that seemed most susceptible to major changes. As we head into the final hours of the election cycle, that dynamic has not changed very much.
For two weeks in mid-October, on a race-by-race basis, the Democratic standing had grown a bit weaker. Then there was a bit of a bounce back last week for the blue team. That dynamic continued this week, as the Democrats ended the month-plus of our forecasts in an incrementally better position than when they began. If the goal of any campaign is to “peak at the right time,” the data tells us that the blue team managed to accomplish that feat:
WEEK # |
DATE |
DEMOCRATIC NET GAIN |
1 |
10/3 |
Democrats +27 |
2 |
10/10 |
Democrats +32 |
3 |
10/17 |
Democrats +27 |
4 |
10/24 |
Democrats +25 |
5 |
10/31 |
Democrats +28 |
6 |
11/5 |
Democrats +30 |
From the most likely to flip to the least likely to flip, here are the 33 races currently favored to change parties according to the polling averages on our House race pages. If these held according to form, the Democrats would net a gain of 30 seats, which would put them at a 225-210 majority (last week’s forecast was a Democratic gain of 28 seats):
POTENTIAL DEMOCRATIC PICKUPS (33 SEATS)
PA-05 (--); PA-06 (--); NJ-02 (current average: D+18); AZ-02 (D+12); PA-17 (D+12); KS-03 (D+10); MN-02 (D+10); PA-07 (D+10); VA-10 (D+10); CA-49 (D+9); NJ-11 (D+8); CO-06 (D+7); FL-27 (D+7); MN-03 (D+6); NJ-07 (D+6); IA-01 (D+5); MI-08 (D+5); IL-14 (D+4); UT-04 (D+4); IL-06 (D+3); MI-11 (D+3); TX-32 (D+3); CA-10 (D+2), PA-16 (D+2); WA-08 (D+2); CA-39 (D+1); GA-06 (D+1); IA-03 (D+1); ME-02 (D+1); NY-19 (D+1); PA-01 (D+1); KS-02 (D+0); KY-06 (D+0)
POTENTIAL REPUBLICAN PICKUPS (3 SEATS)
PA-14 (--); MN-08 (current average: R+11); NV-04 (R+1)
For those reading for the first time, please note that we have an assumption in place that three seats in Pennsylvania will “flip” because they were dramatically altered in the court-ordered redistricting. Two seats, whose essential territory once were repped by Republicans Ryan Costello and Pat Meehan, are near-certain to fall to the Democrats. Meanwhile, one seat, whose turf was primarily represented since the Spring by Democrat Conor Lamb, will likely fall to the GOP after Lamb elected to run in his new home district against Republican Rep. Keith Rothfus.
When we began conducting these House forecasts, we did so by viewing this election through the lens of a wave election, given that optimistic Democrats (and a handful of despondent Republicans) had offered up those words as a designation of the 2018 cycle. Over the course of the past month, we have looked at the essential ingredients of the wave. Those conditions have all been met, at least to an extent.
However, a word of caution—while the conditions have been met, the indicators are not quite as strong in some cases as they were in 2010.
For example, a credible argument can be made that the single-most important ingredient in building a political wave is creating an expansive playing field. The Democratic field of targets, while truly expansive, is not as bonkers as we saw in 2010. Unbelievably, by the end of the 2010 cycle, there were 77 Democratic-held seats where at least one poll showed a lead for the GOP candidate, and there were 54 where there was at least one poll that either showed a lead for the GOP candidate of six points or more, or where the GOP candidate was at or above 50 percent.
The Democrats did damned well, but they are behind those numbers (58 GOP seats with at least one poll showing a Democratic lead, and 33 GOP seats where that lead was six-plus points or the Democrat was at or above 50 percent). Of course, it’s also worth remembering that the GOP picked up north of 60 seats in this cycle, and expectations here are considerably more tempered. What’s more: one place where the Democrats in 2018 have a considerable edge over the GOP in 2010 is the number of GOP-held seats where the Democrats have not led, but are within striking distance. There are 28 GOP-held seats where the Democrat has polled within 5 points or less, almost double what the GOP enjoyed in 2010.
Finally, to truly understand the range of outcomes tomorrow, let’s close by looking at what we have taken to calling the “headwind/tailwind” model. In our polling averages, there are a number of races within 3 points of the equilibrium. So, if either party gets a little gust at their backs at the last, it could have dramatic differences. For example, in last week’s Forecast, we noted that a breath of wind in the form a three-point uniform swing could be the difference between a hugely disappointing Democratic gain of just 10 seats (remaining in a 230-205 minority), or a hugely energizing Democratic gain of 50 seats (resulting in a 245-190 Democratic majority).
On this metric, the numbers are quite a bit better for the Democrats this week. Their “worst-case scenario” (a late three-point swing to the GOP) would keep them out of the majority, but the net gain would still be pretty substantial: a net gain of 16 seats that would put the state of play at 224-211 in favor of the GOP. However, the “best-case” scenario for the Democrats got even rosier. They would reclaim one of their “lost seats” in our averages—NV-04, which, if you believe the guru of Nevada politics (Jon Ralston), we are winning anyway, and they would also pick up an additional 21 seats. That would mean a net gain, if Democrats were to reap the tailwind, of 52 seats, and a yawning 247-188 majority.
Note that the data we are playing with here was everything that came into our hands prior to the close of the weekend, which means that any late data did not make it into our forecasts (hey—we have to turn off the spigot as some point). It’s worth noting that the data that has trickled in on Monday morning has done nothing to dispel the notion of a very good Democratic day on Election Day. How good a day it will be, of course, remains to be seen.