| Editor’s Note: Before we begin this week, we just wanted to acknowledge our friend Geoffrey Skelley, our long-time Crystal Ball associate editor. Geoff started a new position this week with FiveThirtyEight. For nearly seven years, Geoffrey was an exceptionally valuable member of our team, and we will miss him tremendously. However, we are also happy for him as he continues his career at an outlet whose work we deeply respect. We will have more to say in the coming weeks about how we plan to replace Geoff but, in the meantime, we wanted to thank him for his great work for us and to wish him success in his new position.
— Larry J. Sabato, Kyle Kondik, and the rest of the University of Virginia Center for Politics team. |
KEY POINTS FROM THIS ARTICLE
— There are lots of questions, and not many answers, about whether the Brett Kavanaugh confirmation saga might impact November.
— We have 11 House ratings changes, all in favor of Democrats.
— Five gubernatorial ratings changes go in different directions but are generally better for Democrats.
— Only one change in the Senate as the battle for that chamber remains in something of a stasis.
Table 1: Crystal Ball House ratings changes
Table 2: Crystal Ball gubernatorial ratings changes
Table 3: Crystal Ball Senate ratings change
The Kavanaugh curveball
As we enter the final month of Campaign 2018, the political world remains fixated on embattled U.S. Supreme Court nominee Brett Kavanaugh, who awaits the findings of an FBI investigation into whether he tried to sexually assault Dr. Christine Blasey Ford while both were in high school. The allegations, somewhat predictably given the tribal nature of our politics, divided the country along partisan lines. For instance, Quinnipiac University found an almost identical plurality who said they do not want Kavanaugh confirmed (48% against, 42% for) and who said they will vote Democratic in their local House race (49% Democratic, 42% Republican).
We find it difficult to assess the importance of the Kavanaugh situation while his nomination remains in limbo. It does make some intuitive sense to suggest that the outcome, whatever it is, could have a limited and possibly contradictory electoral effect depending on the race. Suburban, college-educated women are both enraged at the president and likelier to be sympathetic to Ford; these voters are very important in some of the key House races, many of which are in newly-swingy territory covering affluent, highly-educated suburbs. Meanwhile, many of the key Senate races are in dark red states, where voters probably are more sympathetic to Kavanaugh and/or outraged that Democrats are trying to submarine the president’s Supreme Court pick. There are some signs that the Kavanaugh fight has stirred the GOP base, at least for the moment.
Questions abound: How might the electorate react if the seat remains open? Would Kavanaugh losing in a floor vote actually help the GOP motivate base turnout because of the higher stakes of an open SCOTUS seat and rage, from the right, over the Democrats (and a handful of Republicans) blocking him? Will the issue fade in importance if Kavanaugh is confirmed? Or would Democrats get even angrier and more engaged if Kavanaugh is confirmed? These questions are important, but also impossible to answer in the midst of the process.
Here’s what we know: The environment remains treacherous for Republicans. The president’s approval rating has rebounded a bit from the dip it took in September — illustrating once again that the president may do better when others are dominating the news — but his disapproval still remains over 50% in polling averages and his overall approval is in the 42%–44% range. That’s not bad for this president, but historically it’s weak on the eve of a midterm, a point that Republican lobbyist Bruce Mehlman illustrates in vivid detail in his most recent, must-read quarterly presentation. The slight uptick for the president hasn’t changed the overall House generic ballot much: Democrats are up by about seven to eight points in averages. That’s suggestive of an environment where Democrats are favored in the House but not overwhelmingly so; that’s also where we’ve been at in terms of our own House handicapping.
The stratified political maps this year make it hard to make a sweeping generalization about the election’s direction. A Democratic House takeover would make this a wave year, in our view, and that designation would be reinforced if the Democrats also pick up a substantial number of governorships, which also seems likelier than not. And yet, the Senate could see very little net change or even a GOP gain, as we’ve repeatedly noted. That makes 2018 different, in all likelihood, from the last three big midterm wave elections (1994, 2006, and 2010). In those years, the presidential out party took the House from the presidential party each time and netted at least six Senate and six gubernatorial seats each year. Democrats could match or exceed those feats in the House and the governorships, but not in the Senate because of the map. (2014 was a wave too, but the GOP only made relatively modest gains in the House and governorships in part because they already held a majority in the House and a majority of the governorships, so we’re setting that election aside here.)
In other words, we don’t envy the headline writers who have to pithily sum up the results of an election when, hypothetically, Democrats could net 35 or so House seats and a half a dozen governorships or more but lose, say, one net Senate seat. That would be an outcome unsurprising to those who closely follow elections, but we suspect many who are just tuning in on Election Night will need a crash course in the particularities of this cycle’s Senate map to understand what happened.
What follows is an explanation of ratings changes in all three categories (Senate, House, governors), along with our latest intel and observations:
House: Assessing “triage”
A telltale sign of House peril is when national third-party groups begin to pull money out of certain districts. Known as “triage,” these moves by outside groups to cut off incumbents who might be behind can be suggestive of what might happen in the fall. Both the National Republican Congressional Committee and Congressional Leadership Fund, the two most important outside groups on the GOP side, have been engaging in triage. However, the two entities cannot coordinate, and they sometimes have different opinions about races.
For instance, CLF recently pulled advertising money from the races of Reps. Mike Coffman (R, CO-6) and Mike Bishop (R, MI-8). Both occupy suburban districts with above-average numbers of residents with four-year college degrees, a demographic suggestive of hostility to President Trump. Both face Democratic opponents, veteran Jason Crow (CO-6) and former Obama administration Defense Department official Elissa Slotkin (MI-8), who significantly outraised the incumbents in 2018’s second quarter and very well may have again in the third quarter (while we do not yet know what Crow and Slotkin raised, similar Democrats in similar races have begun to post eye-poppingly huge fundraising quarters). In response, the NRCC said they would boost their ad buy in CO-6, a race we moved to Leans Democratic a couple of weeks ago. The NRCC also remains engaged in MI-8. But the NRCC is engaging in triage of their own: The committee recently cut off Rep. Kevin Yoder (R, KS-3), who faces attorney Sharice Davids (D) in a suburban Kansas City district that Hillary Clinton carried (CLF remains engaged there, at least at the moment). So, on one hand, the committees may be shifting around money to cover a wider playing field with the knowledge that one or the other will be covering most of the competitive territory. But for incumbents like Bishop, Coffman, and Yoder, they now are getting less outside support even as their Democratic opponents likely will have tons of money to spend. For them, this is a problem.
As analysts, what do we do about “triage?” On one hand, there is the old cliché that applies here: “money talks, and BS walks.” Meaning that, in the political game, how the committees spend their money is the best sign we have about how they really feel about a race. The professionals working these campaigns on both sides have access to the most information, more than analysts do, and their actions are often telling. On the other hand, committees are not always correct in their assessments. For instance, in 2016, the National Republican Senatorial Committee cut a million dollars in support of Sen. Ron Johnson (R-WI) around this time two years ago, and he ended up winning. So the committees aren’t always right, although they also often are.
Of course, now that there are several different big outside groups on each side, different committees have different opinions about different races. One place where CLF and the NRCC are at odds is in Northern Virginia, where Rep. Barbara Comstock (R, VA-10) appears to be an underdog. CLF has never booked any time there, while the NRCC has a giant reservation (so do Democratic groups, although one of them, House Majority PAC, just cut some of its reservation, likely a move made out of confidence). Comstock is very likely trailing but is still in the game: Monmouth University found her opponent, state Sen. Jennifer Wexton (D), up six points, an improvement for Comstock since Monmouth’s last poll a few months ago.
We have several ratings changes this week, noted above. Some of them are made with triage in mind.
Those include moving one of the aforementioned “victims” of triage, Rep. Kevin Yoder (R, KS-3), from Toss-up to Leans Democratic. Davids, Yoder’s opponent, announced that she raised $2.7 million in the third quarter (which ended Sunday), outraising Yoder (who raised less than half of that, $1.1 million). We have heard that Yoder is at best tied and in all likelihood is trailing; the ongoing New York Times/Siena College polling project found Davids leading 51%-43%.
One other House Republican we’re moving to Leans Democratic, from Toss-up, is Rep. Jason Lewis (R, MN-2). A first-term member, Lewis barely won in 2016 over former medical device executive Angie Craig (D), and he was aided by the presence of a third party candidate who probably hurt Craig. It’s just Craig and Lewis on the ballot this time in a worse environment for Republicans; a New York Times/Siena poll found Lewis down double digits. We don’t think it’s necessarily that bad, but Lewis is probably behind.
Bishop, mentioned above, remains in a Toss-up race, but he is the next Republican incumbent in danger of being moved to the Leans Democratic column.
Moving from Leans Republican to Toss-up are two other House Republicans: Reps. Ted Budd (R, NC-13) and Pete Sessions (R, TX-32). This puts three Texas House Republicans in the Toss-up column: Sessions, who is a former chairman of the NRCC, along with Reps. Will Hurd (R, TX-23), who represents a perennially swingy district that extends from San Antonio almost all the way to El Paso, and Rep. John Culberson (R, TX-7), who holds a traditionally Republican seat in the Houston suburbs that (you guessed it) is highly affluent and educated. Most seem to believe, perhaps counterintuitively, that Hurd is in the best position of the three, in part because TX-23 is majority Hispanic and Republicans are holding up better in lower-turnout, less affluent Hispanic-heavy districts than in whiter, more educated suburban districts like TX-7 and TX-32. But even though Hurd is probably leading, we’re having a hard time moving any House races toward the Republicans right now. That said, if you’re looking at our Toss-up column and trying to find Republicans who are in decent shape, Hurd would definitely be one. Meanwhile, we have previously noted the potential dangers to Republicans in North Carolina this cycle, and NC-13 joins the open NC-9 as another good Democratic target.
If you’re looking for a state where the Democrats might make the biggest net improvement from 2016, Pennsylvania might be the best bet. Democrats won only five of the state’s 18 seats in 2016, but a new House map imposed by a Democratic-controlled state Supreme Court that replaced a GOP-drawn map eased their path to gains. We currently favor the Democrats to win at least nine of the state’s 18 seats. And that tally does not include PA-1, a traditionally swingy Bucks County seat where Rep. Brian Fitzpatrick (R, PA-1) and philanthropist Scott Wallace (D) remain locked in a tight race. Monmouth just found Fitzpatrick up 50%-46%. Nor does it include two longer-shot Democratic targets, Rep. Mike Kelly (R, PA-16) in an Erie-based, Western Pennsylvania seat or Rep. Scott Perry (R, PA-10) in a substantially redrawn seat centered on Harrisburg. The DCCC started spending in both races, perhaps a sign that the battlefield in the Keystone State is expanding. Both PA-10 and PA-16 move from Likely Republican to Leans Republican.
Speaking of Pennsylvania, we’re also moving two seats covered by the Philadelphia media market — the open PA-6 and NJ-2 — from Likely Democratic to Safe Democratic, solidifying these two seats’ status as two of the likeliest Democratic pickups. Democrats have strong recruits in both districts and Republicans have made little or no effort to back up their weak nominees in either. Also moving from Likely Democratic to Safe Democratic is the open CT-5, a traditionally competitive seat that seemed like a possible GOP target after Rep. Elizabeth Esty (D, CT-5) opted against running for reelection in the midst of a scandal in which she seemed to cover up for an abusive former staff member, but former National Teacher of the Year Jahana Hayes (D) should be fine.
Finally, we’re moving a couple of highly educated, typically Republican suburban districts to more competitive categories. In suburban Atlanta’s GA-7, Rep. Rob Woodall (R) faces a growing threat from ex-Georgia Senate budget office director Carolyn Bourdeaux (D) as we move that race from Likely Republican to Leans Republican; and, in an open seat based in the Dallas-Ft. Worth Metroplex, TX-6 enters the ratings as Likely Republican out of caution; we list several Texas districts as Likely Republican as a hedge against what could be a suburban revolt in an otherwise blood-red state. That said, in all likelihood if the Democrats pick up a seat or more from Texas, it’ll be from the three toss-ups described above.
This week’s ratings changes get the Democrats closer to the magic number of 218 in our ratings. We now list 210 seats as Safe, Likely, or Leaning Democratic, 197 Safe, Likely, or Leaning Republican, and 28 Toss-ups. Democrats need to win just eight of the 28 Toss-ups to win the House, although we suspect if Republicans hold on to the majority, they will have clawed back at least some seats that we currently rate as Leans Democratic.
Our best guess in the House right now remains a Democratic net gain in the low-to-mid 30s, with enough uncertainty that we would not rule out the Democratic gains sputtering out short of the 23-seat net gain they need.
Table 4: Crystal Ball House ratings
The Governors: A handful of changes, going in both directions
This week’s gubernatorial ratings includes shifts in both directions, but the overall trend is toward the Democrats.
Let’s start in Connecticut, where outgoing Gov. Dannel Malloy (D) is dreadfully unpopular. We’ve long looked at this open-seat race as a Toss-up, and yet based on what we’ve heard in Connecticut, about the rosiest possible interpretation for Republicans is that the race is tied; more realistically, 2006 Senate nominee and businessman Ned Lamont (D) seems to be leading businessman Bob Stefanowski (R) by some amount, as several public polls have indicated (as well as some private ones we’ve heard about). This race is not over, but Republicans couldn’t win Connecticut even in the optimal national conditions of 2010 and 2014, so it’s reasonable to question their ability to win it now, especially with Lamont seemingly already ahead. We’re moving Connecticut from Toss-up to Leans Democratic.
We’ve noted before how Republicans seem to be holding up well in New England. That remains the case in Massachusetts and Vermont, where Govs. Charlie Baker (R-MA) and Phil Scott (R-VT) both appear to be safe. Yet Democrats do have a shot in the other four. Gov. Gina Raimondo (D-RI), though weak, benefits from split opposition in Rhode Island and remains a small favorite. Lamont is now a small favorite as well, and Maine is a Toss-up. Meanwhile, New Hampshire, where Gov. Chris Sununu (R) has solid approval numbers, could nonetheless be a real race. Former state Sen. Molly Kelly (D) is Sununu’s opponent in a state where one typically does not want to be on the wrong side of a potential wave environment, although we still lean the race to Sununu.
Better news for Republicans comes in Arizona and Tennessee. Both states are home to highly competitive Senate contests, but the gubernatorial races both seem to be moving in the GOP’s direction. In the Grand Canyon State, Gov. Doug Ducey’s (R-AZ) race seems to have stabilized, and he’s opened up a lead in what seems to be the mid-to-high single digits. And in the Volunteer State, businessman Bill Lee (R) is up double digits on former Nashville Mayor Karl Dean (D). Arizona goes from Leans Republican to Likely Republican, and Tennessee from Likely Republican to Safe Republican.
As Tennessee moves toward the Republicans, Georgia may be becoming even more competitive. The battle between former state House Minority Leader Stacey Abrams (D) and Georgia Secretary of State Brian Kemp (R) appears to be something of a tie, and the presence of a Libertarian, Ted Metz, means the race very well could go to a runoff on Dec. 4. We’re dusting off a seldom-used Crystal Ball rating category for the time being, moving Georgia from Leans Republican to Toss-up/Leans Runoff, meaning that at the moment we see the race going to overtime. Historically, one would think that a lower-turnout runoff would benefit Republicans, and that very well might be the case if there is indeed a runoff. However, in this environment, where Democrats have performed well in irregularly-scheduled special elections across the country, and where Abrams as an African-American candidate could do a good job of generating black turnout even in a runoff, one can’t rule out the possibility of Abrams having an edge.
Finally, we mentioned in the House section the difficulties the Republicans are having in Pennsylvania. One factor that compounds their troubles is that both Sen. Bob Casey (D-PA) and Gov. Tom Wolf (D-PA) appear to be running away with their reelection bids. We’re moving Wolf from Likely Democratic to Safe Democratic.
Overall, this week’s changes leave seven Toss-ups (including Georgia, in the Toss-up/Leans Runoff category), and all of those Toss-ups are held by Republicans. So while we have the Democrats favored to win three GOP-held governorships already, they could win substantially more than that if the Toss-ups break their way. Map 1 shows the gubernatorial ratings.
Map 1: Crystal Ball gubernatorial ratings
The Senate remains in stasis
We have less to say this week about the Senate. That’s in part because a Kavanaugh effect, if one is present, could have more of an impact in the Senate. The senators are the ones actually deciding on him, and voters in the key Senate races (particularly in the red states with Democratic incumbents) might be animated by the outcome, perhaps particularly if the seat remains open. Think about it this way: The five dark red state Senate Democratic incumbents on the ballot have to persuade a considerable share of voters to take off their Trump/Republican jerseys and vote Democratic for Senate. Doesn’t that task get a little harder if the election becomes defined by the battle for the court? Again, we’re just thinking out loud here: It’s possible, truly, that the court battle won’t have much of an effect, even in the Senate.
Our one change comes in Michigan, where Sen. Debbie Stabenow (D-MI) moves from Likely Democratic to Safe Democratic. Sens. Sherrod Brown (D-OH) and Tammy Baldwin (D-WI), who remain in the Likely column, might not be far behind in making that move as well unless their Republican opponents show more signs of life. Of the current Likely Democratic seats, the most vulnerable one seems to be in New Jersey, where scandal-plagued Sen. Bob Menendez (D) appears to be leading but not by that much. Menendez is fortunate to be running in a Democratic state in a Democratic year and in a partisan, polarized era; remove any one of those elements, and he’d be in clearer danger.
A few additional observations:
— Sens. Jon Tester (D-MT) and Joe Manchin (D-WV) remain in the Leans Democratic column, but both seem shaky to us. The worst polling we’ve seen for either, from Republicans, show them tied, which probably means both are leading (and Manchin has consistently led public polls by close to double-digit margins; Tester’s leads have been more modest). But both of these seats remain live GOP targets.
— By all rights, the open Tennessee race probably should be a Toss-up. We’re not entirely sure whether former Gov. Phil Bredesen (D) or Rep. Marsha Blackburn (R, TN-7) is leading, which means it’s probably something of a tie. And yet, we continue to give a slight edge to Blackburn because of Tennessee’s red hue. But even some top Republicans who know the state well tell us they are worried sick about this race, and our stubbornness about the race’s GOP lean may be misplaced. A series of Fox News polls in several key Senate states, released Wednesday evening, showed Blackburn up five points. One of the other notable findings was that Sen. Heidi Heitkamp (D-ND) was behind Rep. Kevin Cramer (R, ND-AL) by a whopping 12 points, which seems high to us but it is not the only poll we’ve seen showing Cramer up double digits. It’s getting harder and harder to keep Heitkamp in the Toss-up category, and she remains the most clearly endangered Senate Democrat.
— While we’re not ready to move either out of the Toss-up column, Democratic odds seem to be brightening for Sens. Bill Nelson (D-FL) and Joe Donnelly (D-IN). After trailing Gov. Rick Scott (R-FL) slightly, Nelson seems to be back at parity or even leading, and Donnelly may have a small edge on former state Rep. Mike Braun (R), who Republicans worry is not attacking his race aggressively enough. Donnelly led by two points in the new Fox News poll; he was behind by two in an earlier Fox News survey.
As we noted last week, Democrats need to win 80% of all the Senate races this year (28 of 35) to win the Senate, something a party has accomplished only twice before in more than a century of Senate popular elections. We remain skeptical of their ability to do so but we won’t rule it out either.
Map 2: Crystal Ball Senate ratings
Conclusion
We’re now just a little over a month from the midterm. The usual trend in these off-year elections is for the numbers to generally move in favor of the presidential out-party over the course of the final month. That was certainly the case in the last three midterms, anyway. Will the same thing happen? Quite possibly, but the ingredients of this particular election are odd. Trump, and people’s reaction to him, remains a wild card, even if he’s doing more to animate his opposition than his supporters right now. Turnout seems likely to be high: Pew Research Center, mirroring the findings of others, finds high enthusiasm for voting among both Democrats and Republicans, with Democrats a little bit more engaged. And there’s also the Kavanaugh situation, the resolution of which may — or may not — be electorally impactful.