Daily Kos Elections Senate Forecast: Democratic pickup #5 takes shape in Iowa

At the big picture level, not a lot has changed in either the presidential or Senate races in the last week. Compared with a lot of other election years, that’s a good thing! When you’re in the lead in the fourth quarter, each minute that goes by without a screwup or without a change in momentum gets you that much closer to actually winning.

However, if you zoom in to a slightly smaller picture level in the battle over the Senate, there was enough movement in one race to move the needle from red to blue. That state was Iowa, where a trio of new polls pushed Democratic challenger Theresa Greenfield into an aggregate lead over Republican incumbent Joni Ernst.

As we discussed last week, the Democrats need to earn a net three pickups to take control of the Senate (also assuming a win in the presidential race), and that’s exactly what they’ve been doing, with significant and stable leads in pickup opportunities in Arizona, Colorado, Maine, and North Carolina, along with a likely loss of a Democratic-held seat in Alabama. That’s literally the bare minimum, though; a net gain of four would be insurance against, in a good-case scenario, letting Joe Manchin be the Senate’s pivotal vote, or in a very bad-case scenario, against Mike Pence still being Vice-President and able to cast tie-breaking votes. Adding Iowa as one more pickup does exactly that. 

Now, this move in Iowa isn’t exactly surprising. Iowa has been widely regarded as the Democrats’ fifth-best opportunity for a number of months. Greenfield actually led narrowly in the polling aggregates for much of the summer, but a Fox News poll in early September gave Ernst a 50-45 point lead. That was a big enough lead, in combination with a month-long drought of other polling in Iowa, for Ernst to take the lead overall for most of September.

Three polls in quick succession in the last week, however, pushed Greenfield back in front, by an aggregate of two points, 46-44. Probably the most significant of the three was the Selzer poll for the Des Moines Register, where Greenfield led 45-42. Pundits receive an automatic demerit if they do not refer to Selzer’s “gold standard” status in Iowa, where they have a long track record for accuracy.

While our model’s trendlines do not include a weighting factor for pollster accuracy, you might make an internal adjustment to your mental map of the race to reflect that. However, there’s no real need to do so; that poll was quickly confirmed by two other high-quality pollsters later in the week, Monmouth (who saw a tie) and Siena, who put Greenfield ahead 42-40. And more polls, of course, is very helpful. We can feel much more confident about a small lead when three polls in the field at the same time show approximately the same thing, as opposed to just one poll by itself.

In fact, the same thing happened in Maine this week, with a trio of polls bringing the race back more into line with expectations. You might remember that a week ago, we had Democratic challenger Sara Gideon sporting a gaudy 11-point lead against Republican incumbent Susan Collins in our average, thanks largely to a recent Quinnipiac poll giving Gideon a 12-point lead. While, of course, that’s what we like to see, that was tempered by the knowledge that, as much as Collins has set fire to her moderate credentials in the last few years, she probably wasn’t going to lose by double digits in the end and that one poll kind of stuck out like a sore thumb against previous polls.

Three new polls added to our averages this week, however, brought Gideon’s overall lead down to a more normal-looking four points, 47-43. That’s based on two high-quality polls, from Siena (with a 5-point lead) and Suffolk (a 4-point lead for Gideon), as well as a Republican internal poll from Moore Information showing the race tied. 

Now, you might be thinking “But that’s a huge drop!” or “But now that race is too close!” Well, both of those things are reasonable things to think, but, at the risk of sounding overly dispassionate, my responses would probably be more like “Reversion to the mean is better than setting your expectations based on one rosy outlier,” and “Actually, a four-point lead based on a wide variety of polls from reputable pollsters is pretty good in a pivotal race, especially against someone who has usually demolished her opponents!”

With all that in mind, let’s take a look at the overall picture, as seen in this week’s “totem pole”:

The good news is pervasive both above and below the red line. In North Carolina, Democratic challenger Cal Cunningham has increased his lead on Republican incumbent Thom Tillis to 7 points, currently 46-39. That actually pushes North Carolina up to the same level as Arizona and Colorado, two races that most prognosticators have on a more Democratic-friendly tier (including Daily Kos Elections; we rate Arizona and Colorado as “Lean Democratic,” but North Carolina, along with Iowa and Maine, as “Tossup”). That’s maybe a little surprising in that Cunningham wasn’t at the top of most people’s candidate wish lists here, but he has performed adeptly, routinely posting bigger leads against Tillis than Joe Biden over Donald Trump within the state.

And there is a wide array of races just hanging below the red line, to the extent that a strong last-minute breeze could give us up to three or four additional pickups. These are races where our Daily Kos Elections qualitative ratings tend to be a bit more skeptical, with a mix of “Lean Republican” and “Likely Republican” ratings. More than anything, that’s because (with the exception of Georgia, and maybe Texas if you define “swing state” expansively) these are races in presidential red states, which points to the difficulty of that final step of getting from, say, 46 to 50 without the aid of coattails and undecided voters are more likely to be used to Republicans as their default choice.

Even in those states, there’s still some room for hope. In Kansas and South Carolina, you have sizable portions of the state (the Kansas City and Charleston metro areas, respectively) where, through a mix of demographic change and suburban realignment, there’s been definite movement in the Democratic direction since 2016. In Alaska and Montana, those are “elastic” states (thanks largely to low religiosity and low racial polarization) that have a long tradition of splitting tickets downballot. Of course, we’ll continue to monitor these races closely, many of which may well receive a promotion if the polls continue apace.

Finally, there’s the curious case of Georgia’s second Senate race, a special election where appointed Kelly Loeffler is seeking to fill the remainder of the term. The November election here is (and this part may actually be news to one of the candidates here, Matt Lieberman) essentially a Top 2 primary, with the two candidates with the most votes, regardless of party, advancing to a January runoff.

Because of the weird format, we’re not including this race in the “totem pole,” but we are aggregating the frequent polls of this race. The good news is that in the last week we’ve actually seen a few polls with the leading Democratic candidate, Raphael Warnock, moving into second place (which would avoid the Democratic lockout), the best of which was from GBAO Strategies with Loeffler at 29, Warnock at 25, and Republican Doug Collins at 19. The aggregate still has Collins a tiny smidge ahead of Warnock: currently Loeffler 25, Collins 20, Warnock 19, and Lieberman (who seems to still believe he has a shot) at 13. One hopeful note is that, with all of the Democratic candidates little known, Warnock has some room to grow among undecided voters as he starts to advertise.