This is part of an ongoing series, updated periodically, about the 2020 election, with separate diaries on the Presidential and Senate elections (for the most recent Presidential summary, on June 12, see here). My first summary of Senate forecasts was on May 31 (here).
Three weeks can make a lot of difference! The Black Lives Matter protests that have swept across the country for nearly those three full weeks have brought with them significant attitudinal shifts among the American people, including whites, in support of BLM and recognizing police brutality and systemic racism as a problem. That positive change in awareness is in my opinion starting to be visible in the political arena, with the election looking more favorable to the Democrats. This is seen in an expanding Senate battlefield, where more GOP-held seats are starting to look shaky. Keep in mind the current Senate balance is D 47, R 53. With a Biden win, the Democrats have to pick up 3 seats to gain Senate control (with the Democratic Veep holding the tiebreaker vote); without a Biden win, 4 pickups are necessary (and let us all work to make sure it doesn’t come to that).
This week we add one new Senate model, from CNN. You’ll see that a number of the other models have not been updated in 1-2 months. The polling averages for many of the races are, however, more up to date. We’ll look at forecasts from each model, then consider the polling data provided by RealClearPolitics and 538.
CNN: This analysis is actually from May 2, so it’s out-of-date. CNN rates Democrats as clear favorites in AZ, CO, and ME, while forecasting the Rs to take back the AL seat. This would give us 49 D senators; the next most likely D pickup is NC, the closest to a true toss-up in CNN’s list. After that, Ds have a shot in IA, KS, and MT (CNN puts their chances at 25-30% in each of those states). Named longshots (at least a 5% chance) are both GA races, KY, MS(!), SC(!), and TX. However, MI is also a longshot for the Rs (all other D seats are safe). All in all, CNN states the Ds are a slight favorite to gain Senate control (multiple races with low odds still give a good chance of winning at least one of them).
Electoral-vote.com: D 53, R 45, ties 2 (updated daily with new polls). This is a 3-seat shift towards the Ds, compared to 3 weeks ago. I expect that once other models start updating, we’ll see shifts towards a bigger D gain in those as well. Ds pick up AZ, CO, 1 GA seat (Perdue’s), IA, KY, ME, and MT; Rs pick up AL. Ties are NC and SC.
Rachel Bitecofer: D 50, R 47, toss-up 3 (no update; this forecast is from April 28). Ds pick up AZ, CO, ME, and NC; Rs pick up AL. Toss-ups are IA, KS, and MT.
Inside Elections: D 47, R 50, toss-up 3 (no update; this forecast is from April 3). Ds pick up CO, Rs pick up AL. Toss-ups are AZ, ME, NC. In the next tier of states, MI only leans D, while IA, KS and MT only lean R.
Sabato's Crystal Ball: D 48, R 50, toss-ups 2 (no update; this forecast is from April 30). Ds pick up AZ and CO; Rs pick up AL. Toss-ups are ME and NC. In the next tier of states, MI only leans D, while GA-special, IA, and MT only lean R.
Cook Political Report: D 46, R 50, toss-ups 4 (updated June 18). Rs pick up AL; toss-ups are AZ, CO, ME, MT, and NC. The next tier of states includes MI as lean D and GA (both seats), IA, and KS as lean R.
RealClearPolitics: D 46, R 48, toss-ups 6 (CO, IA, ME, MI, MT, NC). Ds pick up AZ, Rs pick up AL. As mentioned above, RCP provides polling averages as well as a Senate map. Some of their criteria for poll inclusion must be a bit...off, as we shall see.
In the 6 models listed, there are 15 different seats named as toss-ups or as “close” races (tilt/lean towards one of the 2 parties): AL, AZ, CO, GA (2), IA, KS, KY, ME, MI, MS, MT, NC, SC, TX. We’ll look at which way the consensus goes on each, along with the RCP averages and polls listed by fivethirtyeight.
First, the only 2 Democratic-held seats on the list:
AL (Jones vs ?): This seat appears on everyone’s radar, and everyone assumes the Rs win it back. They’re almost certainly correct; the most recent polling shows any of the likely R candidates winning by spreads of 8-13 points.
MI (Peters vs : Multiple sites list the MI seat as only leans D, while RCP rates it a toss-up. This seems optimistic (if you’re a Republican). The RCP average has D Gary Peters winning by 9 points; most May and June polls listed by 538 put Peter’s lead in the double digits. This seat is not currently in jeopardy.
So, losing AL would put the Ds at 46 seats. How to get to 50? Well, in military terms, the 2020 Senate elections are a target-rich environment for Democrats. We’ll order the remaining states by RCP polling average.
CO (Hickenlooper vs Gardner): 5 of our sites list this as a D pickup; two (Cook and RCP) call it a toss-up. This is delusional. The only poll listed by RCP, from last August (!), shows Hickenlooper winning by 13. And this is where we start to see problems with the RCP average; 538 shows three poll of Gardner vs. Hickenlooper since then, and the two most recent have Hickenlooper up by 17 and 18 points. D win.
AZ (Kelly vs McSally): 5 sites list this a D pickup, 2 as a tossup. The RCP average has Kelly up over McSally by 9.8; the most recent polls in 538 give him leads of 9, 10, and 13. Once again, the two sites don’t include all the same polls,but it doesn’t matter because they all tell the same story. D win.
ME (Gideon vs Collins): 3 sites have this as a D pickup, 4 as a toss-up. The RCP average puts D Sara Gideon up by 2.5 points over Collins. A more recent poll listed by 538 (but not RCP) has Gideon’s lead as 9. Gideon has steadily increased her lead in the available polling for months. If the trend continues, ME will definitely be in the top tier of likely D pickups.
If Ds sweep CO, AZ, and ME, that still puts them at 49 seats, assuming a loss in AL. Fortunately, the number of races that appear to be true toss-ups is now greater than I think anyone imagined even one or two months ago.
NC (Cunningham vs Tillis): 1 model rates this a D pickup, the other 6 have it as a toss-up. RCP has D Cunningham with a 0.2 point average lead, while the polls listed by 538 show the two parties trading 2-pt polling leads back and forth. Right now, this seat may be the best example of a true tossup.
MT (Bullock vs Daines): While MT is typically a red state — and is, as the Presidential level — Ds routinely win statewide office here. 1 site has this as a D pickup, 3 as a toss-up, and 3 as leaning RRCP lists no polls here, but 538 shows two: a March poll with the race a tie, and a May poll with D Bullock 7 points ahead. Worth watching, and clearly at least a tossup with a D candidate who has won prior statewide election. More polls with a similar Bullock advantage will push this one towards a likely D pickup.
GA (Ossoff vs Perdue): Again, RCP lists no polls here. Polls listed by 538 for May and June show Ossoff with 1-2 point leads. Another genuine toss-up.
GA-Special (Warnock? vs ?): This may be the hardest to call, with uncertainty about exactly who the final candidates will be. But May polling of possible head-to-head matchup reported by 538 shows Warnock with a 1-pt lead over Collins and a 13-pt (!) lead over incumbent Loeffler. In the absence of other information this has to be considered a toss-up as well.
IA (Greenfield vs Ernst): This one is all over the place. Electoral-vote (which updates to the most recent poll) has IA as a D pickup; 2 other sites rate it a tossup, while 4 consider it lean/tilt R. RCP lists 2 recent polls, one with Ernst up 1, the other and more recent (by the Des Moines Register) with Greenfield up by 3. 538 shows 3 different polls in June, all of them with Greenfield up 2 or 3 points. Right now, IA has to be considered a genuine toss-up. If the shift towards Greenfield keeps up, it’ll start to look like tilt-D.
KS (Bollier vs ?): We probably didn’t expect KS on this list, but they did recently elect a D governor, thanks to the extremism of R Kobach. And Kobach may just turn out to be the gift that keeps on giving, as he may win the R primary. However, a poll listed by 538 shows Bollier within 1 point of all 3 possible R candidates (RCP shows no polls). Unless this turns out to be a fluke, KS is another genuine toss-up.
With 6 genuine toss-ups, Democrats just need to win 1 for Senate control (if Biden wins) or 2 for outright control. If we consider each toss-up a true 50/50 chance, then odds are the Ds end up with 52 seats. So I believe we have room for cautious optimism now.
And now we come to a handful of states that we can consider wild cards, stretches, and overreach: these remain unclear or else wildly optimistic for Ds.
KY (McGrath? vs McConnell): I have no idea what’s going on here. Head-to-head polls from February and June on 538 have McGrath within 1-3 points of McConnell; a June poll has McConnell with a massive 20-pt lead. More polling will be needed to clarify if this is a bizarre fluke. For now, it’s too soon to say KY is a toss-up. I certainly hope it is, because next to Trump the most despicable figure in Republican politics today is Mitch McConnell.
SC (Harrison vs. Graham): Oddly, RCP lists no polls at all for this race. 538, on the other hand, has 4 from this year so far. They tell an interesting story. In February, Graham had leads of 13 and 17 pts; in March, 4 pts; and in May, the race polled as even. That’s a very clear trend, even if fewer data points than I would like. One or two confirmatory polls, and I’m prepared to say this is a genuine toss-up. Please let it be true, because I would love to see the odious Lindsey Graham booted from the Senate.
TX (Hegar vs Cornyn): Only 1 site flagged this race. The single poll listed by RCP, from February, has an 8-pt Cornyn lead. 538 adds one from April, but it’s no better; that has Cornyn up 13 pts. Right now, this is clearly out of reach for the Ds.
MS (Espy vs Hyde-Smith): Only 1 site flagged this race as less than solid R. This is almost certainly out of reach for the Ds; the most recent polls (one each on RCP and 538) give Hyde-Smith 8 and 10-pt leads. MS is not where I’d spend my money, if I were the DSCC.
For our next update (in 2 weeks), I’ll leave out MI, MS, and TX as almost certainly out of reach to flip. If AZ and CO continue to be double-digit D leads, we’ll drop those too, because they’re almost certainly D gains. In short, we’re in a great position to win the Senate right now, although we must fight hard right up to election day.