On Monday, Daily Kos Elections rolled out its new 2018 elections portal. It’s a place with one-stop shopping for elections geeks, with interactive maps showing our race ratings and with polling data from our partners at Civiqs. Maybe most significantly, we have separate Senate, House, and gubernatorial pages, where you can look in more detail at each of those three battlegrounds. Click on any state on those pages and you can get a full list of polls released in those races, along with a trendline that charts the current polling average in each state.
Every week, my colleague David Jarman and I will be offering an update on the state of play in the Senate, the House, and with the various elections for Governor around the country. He got us started earlier in the week with his excellent analysis on the current forecast in the Senate, which is incredibly close and subject to meaningful changes depending on the slightest tailwind or headwind for the Democrats.
The House, on the other hand, does not get the “cliffhanger” moniker that it may well merit.
In the media, as well as in the public conversation, it has been a universal assumption for the majority of the cycle that the Democrats are the betting favorites to reclaim a majority in the House for the first time since the 2010 midterms. It is an assumption driven by audacious special election results, fundraising hauls, and reams of polling data.
So, how legitimate is the basis for the confidence in a forthcoming Democratic House? Well, it really depends on how you look at the math, but there is no doubt that the numbers are close enough to guard against any ridiculous levels of complacency on the Democratic side.
That said, the one thing we can say with absolute clarity is that, to paraphrase an old sports clichè, “We’d rather be us than them.”
We’ve been following polling here closely at Daily Kos Elections, and if you are reading this, you probably have, as well. There are two data points in the polling picture that paint a very convincing picture for the Democrats.
The first data point is the nature of the polling releases this cycle. As many election forecasters have noted, one leading indicator of partisan strength in elections is the ratio of sponsored polls released to the public by either party. Analysts (CNN’s Harry Enten among them) have spoken to the fact that it is an indicator of partisan strength when one party’s campaigns and institutions are more willing to make available their polling data than the other party.
This, of course, intuitively feels correct: a party that has a wind at their backs will see it in their data and be happy to share, whereas a party leaning into the gale probably would not like to publicly confirm that their polling numbers are ghastly.
In the 2010 midterms, a historically poor performance for House Democrats was capped by a net loss of 63 seats for the blue team plus the loss of what was a sizable majority. In that midterm, one of the harbingers of doom was the fact that a vast majority of the House polling released from partisan sources in 2010 came from Republican sources: two-thirds of it, in fact.
So, where are we in 2018? Of course, the cycle has not concluded yet, but with a little over a month to go, an eye-popping 81 percent of the partisan polling released so far has come from Democratic campaigns or allies. That is an indisputably good sign for the Democrats heading into the final stretch of the campaign, since we are five weeks until Election Day. At this late date, it is clear that one of two things is taking places: either (a) the GOP’s private polling is largely the shits, or (b) they are inexplicably holding it close to the vest.
The second data point, of course, is the polls themselves. In the spirit of David Jarman’s work earlier in the week, we will do a similar “totem pole” visualization.
Before we look at the graphic, let’s concede a few things. We could not do a “totem pole” visual of the entire field, because that’s 435 seats, and you would probably like to finish reading this before Election Day. Furthermore, we also cannot manage one for every race that has been polled thus far, because that is over a third of the seats in play. Believe it or not, 116 House races have been polled at least once. We could narrow it down to races where the poll average is within 5 points in either direction, but that would still be nearly fifty races!
So, here is how we’re going to handle it, we will stick to tossup races, or races that would presently be favored in our polling averages to flip, as long as they fall within 5 points of equilibrium. That pares it down to just over two dozen races.
To be fair, this excludes some races that are highly competitive (NM-02 comes to mind, as does IL-12). But the DKE crew did break down what are anticipated to be the closest races that have available polling. So if you want to see the whole list of races, you can check it out here.
Given these parameters, let’s look at the races that appear prepared to flip. These are races where the polling average is greater than 5 points, and is leaning to the non-incumbent party. This includes three unpolled races (PA-05, PA-06, and PA-14) which, after the late remap of the Keystone State, have essentially been conceded to the opposing party. PA-05 and PA-06 are the urban/suburban Philly seats abandoned by Republicans Pat Meehan and Ryan Costello which were turned deep blue in the remap, while PA-14 is the seat of newly-elected Democrat Conor Lamb after he decided to run in the reconfigured PA-17 against Republican Rep. Keith Rothfus (Lamb, for what it is worth, is a strong favorite in the 17th district).
As you can see, from this list alone, there are significant gains at the ready for the Democrats:
provisional democratic pickups (12 seats)
CA-49 (current average: D+7), CO-06 (D+11), IA-01 (D+13), KS-03 (D+7), MN-03 (D+10), NJ-02 (D+23), NJ-03 (D+8), PA-05 (--), PA-06 (--), PA-07 (D+7), PA-17 (D+12), VA-10 (D+10)
provisional republican pickups (2 seats)
MN-01 (R+14), PA-14 (--)
Right there, a well-trained eye for politics will see a problem. It is almost certain that Republican Jim Hagedorn is not up 14 points in MN-01. So, what is the problem there? A far-too-rosy GOP internal poll in southern Minnesota happens to be the only poll we have in the race, so it stands (for the moment) unanswered. The same, by the way, is true in PA-17, although no one thinks that Conor Lamb is any particular trouble.
So, if these results hold, we are looking at a Democratic gain of ten seats before we even get to the truly close races. So, the Democrats, just on races that seem to be leaning precipitously one way, are almost halfway home.
How do those critical two dozen races near the midline perform? The polling averages are below.
(A couple bits of housekeeping—(1) the letters “inc” indicates the incumbent in the race, while an asterisk indicates the incumbent party in an open seat race; (2) where the difference is zero, you can go to our House pages and see that the “leader” was determined by who is on top in the graph, which implies a fractional advantage. Of course, this should not be interpreted as a “lead”, but we have to draw the thin “red line” somewhere.)
| DISTRICT |
D CAND. |
D AVG. |
R CAND. |
R. AVG. |
DIFF. |
FLIP? |
| NC-09 |
McCready |
43 |
Harris* |
38 |
+5 |
D FLIP |
| CA-45 |
Porter |
48 |
Walters (inc) |
44 |
+4 |
D FLIP |
| NC-13 |
Manning |
44 |
Budd (inc) |
40 |
+4 |
D FLIP |
| MN-02 |
Craig |
48 |
Lewis (inc) |
45 |
+3 |
D FLIP |
| MN-08 |
Radinovich* |
45 |
Stauber |
42 |
+3 |
|
| NJ-11 |
Sherrill |
44 |
Webber* |
41 |
+3 |
D FLIP |
| KS-02 |
Davis |
41 |
Watkins* |
39 |
+2 |
D FLIP |
| CA-10 |
Harder |
49 |
Denham (inc) |
48 |
+1 |
D FLIP |
| WA-08 |
Schrier |
47 |
Rossi* |
46 |
+1 |
D FLIP |
| NY-22 |
Brindisi |
45 |
Tenney (inc) |
44 |
+1 |
D FLIP |
| NY-19 |
Delgado |
45 |
Faso (inc) |
44 |
+1 |
D FLIP |
| NJ-07 |
Malinowski |
45 |
Lance (inc) |
44 |
+1 |
D FLIP |
| MI-08 |
Slotkin |
45 |
Bishop (inc) |
44 |
+1 |
D FLIP |
| IL-06 |
Casten |
44 |
Roskam (inc) |
43 |
+1 |
D FLIP |
| WV-03 |
Ojeda |
43 |
Miller* |
42 |
+1 |
D FLIP |
| NC-02 |
Coleman |
44 |
Holding (inc) |
44 |
0 |
D FLIP |
| CA-48 |
Rouda |
45 |
Rohrabacher (inc) |
45 |
0 |
D FLIP |
| TX-32 |
Allred |
46 |
Sessions (inc) |
46 |
0 |
D FLIP |
| RED LINE |
RED LINE |
|
RED LINE |
|
|
|
| IA-03 |
Axne |
43 |
Young (inc) |
43 |
0 |
|
| ME-02 |
Golden |
48 |
Poliquin (inc) |
48 |
0 |
|
| FL-26 |
Mucarsel Powell |
44 |
Curbelo (inc) |
45 |
-1 |
|
| KY-06 |
McGrath |
46 |
Barr (inc) |
47 |
-1 |
|
| TX-07 |
Panill Fletcher |
46 |
Culberson (inc) |
47 |
-1 |
|
| VA-07 |
Spanberger |
46 |
Brat (inc) |
47 |
-1 |
|
| CA-25 |
Hill |
44 |
Knight (inc) |
47 |
-3 |
|
So, first, the good news: if you take the presumed ten-seat advantage outlined prior to the “totem pole” visual, and then add the seventeen Democratic flips on the favorable side of the “red line”, you get a net of 27 pickups. If the Democrats net 27 pickups, they would have the majority with 223 seats to just 211 for the GOP.
But before anyone starts shopping for champagne, allow us to offer a word of caution. It would take only the slightest alteration in the current climate to leave the Democrats tantalizingly close, but short, of the majority. If you move the needle just two points in the direction of the GOP, Democrats would net a gain of just 16 seats. That would leave the Republicans with a 224-211 majority.
Of course, the converse is also true: if the winds blow a little stronger at the back of the Democrats (let’s assume the same two-point uniform swing), they move to a more secure 232-203 majority, for a net pickup of 37 seats (in addition to the tossups in the diagram, a two-point flip to the Democrats would also snare some current lean/likely GOP seats like TX-31, NM-02, and OH-12.)
Also worth noting: there are a couple of races that are certain to be competitive for which there is not, as of this exact moment, a completed poll. Ironically, that will change in the next few days, as Siena is currently in the field in one of those races (the open Republican seat in MI-11). The other one, currently rated as Lean Republican, is NY-11, where there has not been a whisper of a poll.
That’s the current state of play, and Democrats would probably consider that just a bit too close for comfort. That said, as we will explore next week, wave elections do have their own dynamics, and those dynamics often intensify as we speed towards Election Day. We are not far removed from a major-league wave election, one that Democrats would love to forget. Alas, next week we will take a trip down (recent) memory lane, because there are legitimate parallels to be drawn between 2010 and 2018.
