In last week’s maiden voyage of the Daily Kos Elections House Forecast, we noted that the current poll averages in individual races, if they held up on Election Night, would be enough to propel Democrats into the majority. But we also cautioned that a lot of races were on a knife’s edge, and that only a very small headwind for the Democrats could leave them excruciatingly close, but short of the majority.
Before anyone began to feel the tug of despair, however, we also offered a ray of hope (emphasis mine):
“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 (...) we will take a trip down (recent) memory lane, because there are legitimate parallels to be drawn between 2010 and 2018.”
In the coming weeks, we will delve into those parallels, starting with a key ingredient of any legitimate wave election: the ability of the party out of power to put races into an acceptable range of competitiveness. On that metric, the Democrats are putting nearly as many races into play as the Republicans did in 2010 en route to a massive 63-seat gain.
Before we take the deep dive into the data there, however, let’s kick off this week’s forecast by looking at the current state of play, in a week which saw a mixed bag of data for Democrats downballot. On balance, however, the Democrats appear to be in an incrementally better position than they were a week ago, a result that might defy conventional wisdom.
THE HOUSE FORECAST: FOUR WEEKS OUT
We will dispense with the “totem pole” graphic this week, if only because with so many races on the table, that kind of graphic gets a little cumbersome. We will just go with a simple listing of the relevant races. Starting with a reminder: there are three races that are totally unpolled, but are a given to fall to the opposing party. All of them are a result of the redistricting in Pennsylvania: two are flipping to the Democrats and one to the GOP.
Much of the public conversation about polling this week has been unusually pessimistic for the Democrats, as the pundit class relentlessly speculates that the Kavanaugh fracas has “awakened the sleeping giant” of Republican voters just in time for the midterms. And, indeed, there have been some polls at the House level to give Democrats pause. In the ongoing series of Siena/NYT polls, for example, several “reach races” where private polling had shown Democrats with cause for hope were released, and virtually all of them staked the incumbent Republicans to considerable advantages. But amid that dollop of bad news, there was plenty of good news for Democrats. In Virginia, a monster spending spree by Republicans to try to salvage the fortunes of Rep. Barbara Comstock seems to have been for naught, as Democratic challenger Jennifer Wexton remains staked to impressive leads in a trio of recent polls.
On balance, the Democrats are, despite the chattering classes, in a slightly better position on the House map than they were last week. From the most likely to flip to the least likely to flip, here are the 40 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 32 seats, which would put them at a 227-208 majority (last week’s forecast was a Democratic gain of 27 seats):
POTENTIAL DEMOCRATIC PICKUPS (36 SEATS)
PA-05 (--); PA-06 (--); NJ-02 (current average: D+23); IA-01 (D+13); PA-17 (D+12); AZ-02 (D+11); CA-49 (D+10); CO-06 (D+10); NJ-03 (D+8); MN-03 (D+7); PA-07 (D+7); VA-10 (D+7); CA-10 (D+6); KY-06 (D+5); MN-02 (D+5); CA-45 (D+4); IL-06 (D+4); KS-03 (D+4); NJ-11 (D+4); NY-22 (D+4); CA-25 (D+3); NC-13 (D+3); MI-11 (D+3); KS-02 (D+2); NC-07 (D+2); WV-03 (D+2); FL-26 (D+1); NY-19 (D+1); NC-02 (D+1); TX-07 (D+1); TX-32 (D+1); VA-07 (D+1); WA-08 (D+1); CA-39 (D+0); NJ-07 (D+0); WA-03 (D+0)
POTENTIAL REPUBLICAN PICKUPS (4 SEATS)
PA-14 (--); MN-01 (current average: R+14); AZ-01 (R+3); FL-07 (R+0)
It’s worth noting that some of these polling averages are based on, as is often the case in House campaigns, on a single poll. That is true for all three of the polled “GOP pickups”—all three of them are based off single polls, all of them from GOP outlets. To be fair, there are a couple of Democratic “pickups” at the moment based off similar data (NC-07, for example, comes to mind).
Despite the positive numbers we see above in terms of potential pickups, prudence dictates that we note that the same cautions and caveats from last week still apply. If there is just little bit of a red-tinted headwind, the ascendant Democratic majority gets imperiled. Of the 36 potential pickups listed above, the lead is two points or less in 13 of them.
Of course, the converse also remains true: Democrats could also be cruising toward a net of 50 seats gained if only a couple of points of tailwind get added to the tally.
anatomy of a wave, part one: seats on the table
A fundamental element of a wave election is the desire to generate as expansive a playing field as possible. In the GOP wave election of 2010, the Republicans netted a gain of 63 seats. In other words, they were able to defeat candidates defending over a quarter of the Democratic seats in the House (recall that the Democrats actually picked up a couple of seats—low hanging fruit in HI-01 and LA-02, notably). You simply cannot do that unless you have candidates that can put races on the table in a wide variety of seats, from the most vulnerable locales to places where the incumbent party appears impenetrable. Eight years ago, the Republicans managed to do just that, putting so many races on the table that a monster gain became plausible.
In 2018, an enormous key for the potential success of House Democrats, paralleling the GOP in 2010, is the fact that they have had the good fortune to feature an unusually heavy amount of well-funded and excellent candidates. As a result, Democrats find themselves with a potential battlefield that actually rivals 2010’s field of play for the GOP.
Consider the data: as of October 1 of this year, there were 74 GOP-held seats that fell under one of four categories:
- At least one poll where the Democratic candidate had a lead of 6 or more points.
- At least one poll where the Democratic candidate had a lead of any kind.
- At least one poll where the Democrat and the Republican were in a tie.
- At least one poll where the Republican candidate led by 5 points or less.
That’s right—over six dozen Republican seats have seen a poll in this cycle where the Democrat was either leading or within 5 points of the Republican. While the majority rests on a necessary gain of 23 seats, Democrats would do well to have a target list which is substantially larger than that. because you don’t usually hit every race you aim for.
That mission, one month out from Election Day, has been accomplished by the blue team. And you can’t win a majority without that kind of playing field.
Some might be wondering--how does this compare with the 2010 wave election? Therein lies some very good news, albeit with a critical caveat. In 2010, the Republicans by the same point had managed to put 82 races in one of those four tiers described above. In other words, the Democrats by one month out had managed to put roughly as many races in play in 2018 as the Republicans managed to do when they seized a net of 63 seats in 2010.
Now, for the caveat: even though the “master list” of close races in 2010 and 2018 were of similar size, it is worth noting that the nature of the competitiveness in the two cycles is quite different. In short, Republicans had managed to put over twice as many races in the most vulnerable column (at least one poll with a substantial lead for the out-party). The Democrats in 2018, on the other hand, have stacked the lowest tier of races where the Democrats have not polled in the lead in a race, but are extremely close. So while the Democrats have put a lot of seats at the table, so to speak, there is still more work to be done.
Worth noting, though—the list of “competitive races” based on available polling is almost certainly incomplete. There are a number of races that have the real prospect of becoming competitive, but remain unpolled. For example, when this list was culled (as of October 1st), there had been no polling of the competitive open seat in MI-11. Since the list was created, not only was a poll conducted in this GOP-held open seat, but it showed Democrat Haley Stevens with a seven-point advantage. There are some other potentially competitive races that are unpolled, including one Lean R race (NY-11), and nearly a dozen Likely R races.
What’s more—in wave elections, some races on the periphery of competitiveness will break late. It always happens in wave elections. In fact, five of the GOP’s pickups in 2010 occurred in races that did not see a competitive poll until within a month of the election. These were the dark-horse upsets that we remember with no small amount of pain, like Blake Farenthold’s monster upset of Solomon Ortiz in TX-27, or Chip Cravaack knocking off a dean of the House in Jim Oberstar in Minnesota.
The good news: there is at least a modest chance that a few of those lurk out here for the Democrats today. Could it be a Tracy Mitrano in New York? An Audrey Denney in California? One of the true joys of the election season will be trying to figure out which late-breaking upset comes across the finish line on Election Night.
Next week we will know the (semi) final results of the money chase, and we will look at that as well as the recent polling to see if the wave is ebbing, building, or preparing to break right on top of the GOP majority.