Stivers’ resignation sets up another House special election

Another prominent establishment-oriented Republican is leaving the U.S. House of Representatives. Rep. Steve Stivers (R, OH-15), a former chairman of the National Republican Congressional Committee, announced Monday that he is resigning in mid-May to become the president of the Ohio Chamber of Commerce. Stivers appeared to be gearing up to run for the state’s open U.S. Senate seat, so the news comes as something of a surprise.

Stivers’ eventual resignation sets up another House special election for later this cycle. OH-15 covers some western and southern Columbus suburbs as well as more rural counties both southwest and southeast of Ohio’s most populous city. Donald Trump carried the district 56%-42% in 2020, slightly less than his 2016 margin. This is a more Republican district at the presidential level than OH-12, which covers the northern Columbus suburbs and more rural areas north and east of Columbus and which featured a highly-competitive 2018 special election between now-Rep. Troy Balderson (R) and Franklin County Recorder Danny O’Connor (D). OH-12 voted for Trump by just six points in 2020, down from 11 in 2016. A key difference between OH-12 and OH-15 is that the former has a considerably above-average percentage of residents over 25 with a four-year college degree (a little over 40%), while OH-15 is right around average (a little over 30%). A higher percentage of degree-holders has been a key indicator of a trend toward Democrats in recent years across the country.

Additionally, and as we noted in our piece on House special elections last week, House specials are likelier to break against the president’s party. The 2018 OH-12 special happened with Donald Trump in the White House; now Joe Biden is president. That factor alone likely improves Republican chances of holding OH-15 without too much of a problem.

The OH-15 special likely won’t happen until later this cycle; the general election for another U.S. House special in Ohio, the heavily Democratic OH-11 that runs from Cleveland to Akron, isn’t happening until November.

We’re starting the special election race as Likely Republican, and it’s probably closer to being Safe Republican than it is to being Leans Republican. There is a deep bench of potential Republican candidates for the seat, although Democrats may also be able to produce a strong nominee given their strength in Columbus’ Franklin County. O’Connor is someone to watch: He announced recently that he was planning to run for Congress again in 2022, but he did not indicate the district in which he would be running, likely because the map in Ohio (and in every other state with more than one House district) is still up in the air. In all likelihood, there will continue to be a safe Democratic seat covering Columbus and central Franklin County (currently OH-3, held by Democratic Rep. Joyce Beatty), with Republican-leaning seats around it. Regardless, the eventual winner of the special election will have to run for reelection in a differently-configured district than the current OH-15.

Speaking of specials… a word on LBJ, 1937, and court-packing

In response to last week’s history of House special elections since 1957, reader Sammy Finkelman reminded us that President Lyndon B. Johnson got his electoral start in a 1937 U.S. House special election. Johnson ran in TX-10, which then covered Austin and what was then a surrounding constellation of tiny towns and rural areas. “Many of the district’s country people lived in towns too small to be marked on any map,” wrote Robert Caro in the first volume of his LBJ biography series, The Path to Power.

Johnson triumphed with just 28% of the vote, though his next-closest competitor only won 17%. All of the candidates were Democrats, and there was no runoff — unlike special elections in present-day Texas. In the upcoming TX-6 special election, there will be a runoff unless one of the nearly two-dozen candidates wins a majority of the vote in the first round of voting, which seems very unlikely.

Johnson ran as a strong supporter of then-President Franklin Delano Roosevelt, even backing FDR’s controversial (and ultimately unsuccessful) plan to expand the Supreme Court — or “pack” it, depending on one’s opinion of the plan — by adding members to it.

Expanding/packing the court is back in the news, with a proposal by House Judiciary Committee Chairman Jerry Nadler (D, NY-10) and Sen. Ed Markey (D-MA) to add four members to the court, presumably flipping the court’s 6-3 conservative majority into a 7-6 liberal majority with four additional nominations by President Biden. But Speaker of the House Nancy Pelosi (D, CA-12) said she did not intend to bring the legislation to the floor, and Biden recently created a commission to study the courts — which, like many commissions before it, probably won’t lead to massive, actual changes. This all comes in the wake of several conservative victories in the judicial wars that have rankled liberals: Senate Republicans’ blockade of President Obama’s nomination of Merrick Garland in advance of the 2016 election followed by their confirmations of two Donald Trump appointees, Brett Kavanaugh before the 2018 election and Amy Coney Barrett before the 2020 election.

North Carolina Senate primaries taking shape

On the other side of Capitol Hill, one Senate contest that has been heating up in the past weeks is North Carolina’s. The retirement of three-term Sen. Richard Burr (R-NC) has left an open seat in a state where Democrats have come close in many recent senatorial elections but haven’t actually sealed the deal since 2008.

Last week, former Gov. Pat McCrory (R-NC) got into the Republican primary. Former Rep. Mark Walker (R, NC-6) has already been in the race and has racked up some important endorsements, but his latest fundraising report was not impressive. McCrory was mayor of the state’s largest city, Charlotte, from 1995 to 2009, and self-identified as a pragmatic “Eisenhower Republican.” However, after winning the governorship in 2012, on his second try, he was pushed rightward by a conservative legislature and was defeated by now-Gov. Roy Cooper (D-NC) in 2016.

In his first two statewide runs, McCrory had considerable crossover appeal in the Charlotte area — he only barely lost Mecklenburg County in 2008 and carried it in 2012. But in 2016, much of that support collapsed, as he signed the state’s controversial “bathroom bill,” which prompted some businesses to boycott the state. (Map 1)

Map 1: Pat McCrory’s statewide races in Mecklenburg County

So if McCrory, though a transitional figure in the state’s politics, makes it to the general election, he likely won’t have much of an “electability” bonus in his home region — this is certainly something his primary opponents will claim. In south Charlotte, once his political base, Unaffiliated voters now dominate this formerly-Republican area. Given North Carolina’s semi-closed primary system, those Unaffiliateds could support him in the Republican primary, but they’d also have the option to vote in the Democratic primary.

Still, McCrory enters the primary with high name recognition, and could perform well, in the primary, in other parts of the Charlotte area.

Rep. Ted Budd (R, NC-13), who hails from the state’s very Republican Piedmont region, appears poised to enter the race. Lara Trump, the former President’s daughter-in-law who is originally from the coastal city of Wilmington, continues to be mentioned as a candidate. One other potential candidate, Lt. Gov. Mark Robinson (R-NC), said Monday night that he is not going to run. 

The most prominent Democratic candidate in the race is state Sen. Jeff Jackson. Jackson, who — coincidentally — represents the area McCrory is from, got in the race in January and is in the process of touring all the state’s 100 counties. Former state Supreme Court Justice Cheri Beasley is a likely candidate, as well. Beasley was very narrowly defeated last year, and she received more raw votes than Biden.

Overall, North Carolina probably isn’t Democrats’ best offensive opportunity, as the Crystal Ball rates it as Leans Republican. Still, by this time next month, the primaries there, on both sides, could be looking a bit more competitive.

Watch experts discuss voters of color in 2020; Center for Politics hosts Zambian ambassador Wednesday

On Thursday evening, the Center for Politics hosted the Brennan Center’s Theodore Johnson; Emory University’s Andra Gillespie; UCLA’s Natalie Masuoka; and the Pew Research Center’s Mark Hugo Lopez for the discussion “The Changing Face of America: Voters of Color in the 2020 Election.” They discussed several trends among voters of color, including the gender gap within certain minority groups, the importance of Black voters in Joe Biden’s victory, Republican growth among Hispanic voters, the value of “descriptive representation,” and much more.

To see the program, see this direct link or visit our YouTube channel, UVACFP.

On Wednesday from 4 p.m. to 5 p.m., the Center for Politics will be hosting His Excellency Lazarous Kapambwe, Ambassador of Zambia to the United States. Kapambwe will speak about the relationship between Zambia and the United States, current successes and challenges in the country, and the state of civil discourse in Zambia. The virtual event is free to attend, and more information is available on Eventbrite.

KEY POINTS FROM THIS ARTICLE

— With apparent victories in the Georgia Senate runoffs, Democrats have captured a narrow governing trifecta — control of the White House, Senate, and House.

— Republicans got a sizable Election Day turnout, but Democrats built a big enough lead in pre-Election Day voting to withstand their onslaught.

— Republicans will lick their wounds and look ahead to 2022, when the pendulum often swings against the party in power.

A narrow Democratic trifecta

With apparent victories in both Georgia Senate runoffs — our friends at vote-counting service Decision Desk HQ have called both races for Democrats Raphael Warnock and Jon Ossoff over Republicans Kelly Loeffler and David Perdue — Democrats have won their first trifecta in Washington since the 2008 election.

But whereas that election a dozen years ago was a landslide that elected huge Democratic legislative majorities, their victory this time was a collective squeaker, producing the barest of majorities.

This takes nothing away from the Democrats’ achievement, but it does illustrate the competitive era in which we find ourselves.

In 2008, Barack Obama won by seven points and captured 365 electoral votes. Democrats got as high as 60 Senate seats during the 2009-2010 Congress, or 60% of all the seats, and they won 59% of the House seats.

As this 2020 cycle wraps up, Democrats beat an incumbent president with 306 electoral votes and by four and a half points in the popular vote, and they netted three Senate seats while losing what appears to be a dozen seats in the House compared to what they won in 2018 (one race, NY-22, remains uncalled, and there is a lingering dispute over IA-2, which Republicans won by just six votes).

Joe Biden won 51% of the vote, and Democrats won 51% of the House seats. The Senate is 50-50, but with Vice President-elect Kamala Harris acting as effectively a 101st senator to break ties, the Democrats will have what amounts to a 51-50 Senate majority: 51 divided by 101 is 50.495%. If you’ll forgive us for slightly fudging the rounding, Democrats also have roughly 51% of the Senate seats.

Hence, the 51% trifecta.

This mirrors the governing alignment following the 2000 election, when Republicans held the White House along with 51% of the House and, for a few months, a 51-50 edge in the Senate (which evaporated when Republican Sen. Jim Jeffords of Vermont became an independent in mid-2001 and joined the Democratic caucus).

This new Democratic trifecta was made possible by Warnock and Ossoff’s narrow, apparent victories.

As of this writing (late Wednesday morning), Warnock was ahead by a little over a percentage point and Ossoff was up by four-tenths of a point. The remaining votes to be counted appear to be Democratic leaning: According to the New York Times’ “needle,” which is projecting the expected final votes, Warnock is on track to win by about two points and Ossoff by about one. This comes after Biden won the state by about a quarter of a point in November.

One surprise in Georgia was that turnout was very high, but that Democrats apparently were able to win anyway. As of the most recent tabulations, there were about 4.4. million votes cast, a number that will grow a little. That is roughly 88% of the total presidential vote — an incredible level of turnout for a non-presidential race, but one befitting the immense stakes involving two Senate runoffs to decide overall control of the Senate.

We wrote yesterday that Republicans needed a bigger share of the electorate to vote on Election Day than in the presidential election, and that they needed to perform better with that group than Donald Trump did in November. They checked both of those boxes, but still came up short.

Let’s focus on the Ossoff-Perdue race, which was closer than the Warnock-Loeffler contest.

Election Day voters as a share of the total electorate were 30% of the electorate, compared to 20% in November, and Perdue won these voters by 27 points, up from Donald Trump’s 23-point margin among these voters in November. The general belief among observers, including us, was that this kind of Election Day edge would be difficult for Democrats to overcome.

However, the Democrats were able to hold on because they built a more substantial edge in the pre-Election Day vote than they enjoyed in November.

Votes cast either early in-person or by mail made up 70% of the total votes, down from 80% in November, but Ossoff won these votes by 12 points, double Joe Biden’s six-point margin in November. This included winning two-thirds of the mail-in votes — a little up from Biden’s advantage in November — and narrowly winning the early, in-person vote by a little more than a percentage point after Trump had won the early in-person vote by six.

The pro-Democratic change in the early in-person vote is shown in Map 1.

Map 1: Early in-person Georgia vote, November vs. runoffs

Though both maps are generally similar, a few differences are apparent. In the Atlanta metro area, some of the suburban counties — such as Gwinnett and Henry counties — are darker shades of blue on the right map. In Georgia’s portion of the Black Belt, five counties supported Trump but flipped to Ossoff in the in-person early vote. Though these five are rural, they have relatively high Black populations — one of them is President Jimmy Carter’s home of Sumter County, which is majority Black by composition.

We mentioned yesterday that Black voters were performing better in the pre-Election Day vote for this election than they did in November. This turned out to be one of the biggest stories of the night.

Though the Democratic Party has, broadly, gained with college-educated whites in the Trump area, in the South, energizing the Black vote is still the most important ingredient to winning statewide races. In Louisiana, Gov. John Bel Edwards (D-LA) was reelected in 2019 in large part because of his 84% share in the state’s Black-majority 2nd Congressional District. Though he fared worse with presidential turnout dynamics a few months ago, now-former Sen. Doug Jones (D-AL) won with a similar playbook in 2017.

Yesterday, both Democrats fared worse than Biden in areas of the Atlanta suburbs. Specifically, northern Fulton County (which used to be known as “Milton County”), supported Biden 52%-47% in November, but preferred GOP Senate candidates. In the runoffs, Milton County was a rare area that, in unofficial returns, split its ticket: It was close both ways, but it appears to have voted for Perdue and Warnock — perhaps Loeffler’s having to, necessarily, trek rightward in her primary hurt her standing here.

From a historical perspective, Milton County being essentially tied represents a strong performance for Democrats — until recently, it routinely voted 2:1 Republican — but compared to the presidential topline, the result there for Democrats seems a bit underwhelming. That both Democrats will still likely beat Biden’s overall margin in the state despite their slippage with the white suburbanites there speaks to the strong support they inspired with Blacks. Indeed, on the other end of the Atlanta metro area, Clayton County, which is roughly 70% Black, swung several percentage points to Democrats from November.

Kabir Khanna of CBS News found that turnout ran a little lower in counties with higher percentages of white voters who do not hold a four-year college degree, counties that make up the GOP base. So, again, while the partisanship of the states are a lot different, there are some parallels between Democrats’ most recent victories in Alabama, Louisiana, and Georgia: They were all driven by strong Black turnout and not-so-strong white working class turnout. These sorts of patterns show up on Map 2.

Map 2: Runoff turnout as a percentage of November turnout

Henry County, south of Atlanta, was only 90% reporting at the time of this writing, but generally, northwestern and southeastern Georgia — where some of the most heavily GOP counties in the state are located — saw the sharpest turnout drops from November. In the northwestern corner of the state, the 14th Congressional District, an overwhelmingly Republican seat held by far-right Rep. Marjorie Taylor Greene (R, GA-14), only cast about 85% of its November total, a few percentage points under the statewide 88%. With the contests as close as what we saw yesterday, those types of seemingly minute underperformances added up for the GOP. The relatively high turnout drop in GA-14 was also notable because Trump was in the district on Monday, where he held his final pre-election rally as the sitting president.

By contrast, turnout was relatively strong in several southwestern counties. Randolph County, which sits in the heart of the region, was the only county that turned out at over 95% of its November total. As a Black preacher, Warnock may have been a uniquely good fit for there: Randolph County is 60% Black, and in the House, the area is represented by Rep. Sanford Bishop (D, GA-2), a moderate member of the Congressional Black Caucus who often brings up his Baptist faith on the campaign trail.

It is not too surprising that Warnock performed a little bit better than Ossoff. Warnock faced Loeffler, an appointed incumbent, while Ossoff faced Perdue, an elected incumbent. We know from history that elected incumbents generally have more of an incumbency advantage than appointed incumbents, so perhaps this helped Perdue do a little better than Loeffler. Warnock probably also is a more compelling candidate than Ossoff, and the presence of a Black candidate on the ticket very likely contributed to the stellar Black turnout.

The differences between the two races were not huge, although under slightly different circumstances, we could have seen a split decision, with both Warnock and Perdue winning (thus saving the GOP Senate majority).

The results we saw in Georgia yesterday are more evidence that presidential and senatorial races are becoming more correlated. Both Ossoff and Warnock won by carrying the same 30 counties that Biden did in November. Looking nationally, of the 35 senators elected this cycle, Sen. Susan Collins (R-ME) was the only senator to win a state that voted for the other party’s presidential nominee.

Given the close outcomes in Georgia, any number of factors could have tipped the races the other way. It seems reasonable to suggest that if Trump had accepted his presidential loss, Republicans could have more easily made the Senate runoffs a referendum on unified Democratic control of Washington and perhaps generated a bit more crossover support to have held the seats, and the Senate. After all, Republicans did finish ahead of Democrats in the initial voting in both Senate races back in November. We doubt that the president’s late-breaking support of $2,000 stimulus checks helped the Republicans, either, as it immediately threw cold water on the $600 stimulus checks that Congress approved in advance of the election.

However, asking Trump to behave gracefully in the aftermath of the election would be like asking a vegan to eat a T-bone steak — it just wasn’t going to happen. So Perdue and Loeffler stuck with the president, and they almost squared the circle anyway.

The United States’ political pendulum never stops swinging. The Democrats have the most fragile of Senate majorities, which could be upended at any time by a surprising development (death, resignation, party switch, etc.).

Republicans can look ahead to the 2022 midterm, when they can run promising a check on unified Democratic control of Washington, a message that often appeals to a midterm electorate.

Warnock will be on the ballot again in two years; so too will Sen. Mark Kelly (D-AZ), one of the Democrats’ other 2020 majority-makers.

But for now, the Democratic victories in Georgia give President-elect Joe Biden more breathing room for getting his Cabinet and judicial appointments through the Senate and for pursuing his legislative agenda.

A 51% trifecta is not an overwhelming trifecta, but it’s a trifecta nonetheless.

KEY POINTS FROM THIS ARTICLE

— Joe Biden is on track to exceed Barack Obama’s 2012 popular vote margin, but his victory in the key states is even narrower than Donald Trump’s in 2016.

— Less than 2% of the national vote went to candidates other than Biden and Trump, a significant change from 2016.

— Assuming nothing changes, as many as 94 of 100 senators in the next Congress will share the same party as the state’s presidential winner.

— The ability to generate crossover support helped Republicans perform surprisingly well in both Senate and House races.

The presidential race: both not close and extremely close

Votes continue to be counted in the presidential race, and all indications are that Joe Biden’s lead in the national popular vote will continue to grow. Among those states that still appear to have a significant number of votes to count are California, Illinois, and New York. These big blue states will pad Biden’s national edge, which currently sits at 50.7%-47.4% in the national popular vote as of Wednesday morning. Biden’s national popular vote edge appears likely to exceed Barack Obama’s from 2012 (about four points), though it will fall short of Obama’s seven-point edge in 2008. Assuming Biden clears Obama’s 3.9-point 2012 margin, his will be the second-biggest popular vote win in the six elections this century (yes, we know, 2000 technically isn’t in this century, but we’re including it anyway).

Of course, the popular vote does not determine who wins the presidency.

We (and others) frequently noted the past four years that Donald Trump’s 2016 victory was built on the strength of a roughly 78,000-vote edge in three key states (Michigan, Pennsylvania, and Wisconsin). Flipping those states, which were Trump’s three-closest victories, to Hillary Clinton would have given her an Electoral College majority.

This time, Biden’s fate was in the hands of four states, Pennsylvania, Wisconsin, Arizona, and Georgia, that were collectively decided by about 97,000 votes (that number will change, and Biden’s edge at least in Pennsylvania should continue to expand while Arizona has gotten closer in later-counted returns). Give these four states to Trump, and Trump wins.

However, it’s actually more complicated than that, and Biden’s actual edge in the decisive states is really even narrower.

If one gave Biden all but his three closest states (Arizona, Georgia, and Wisconsin), he would have been stuck in a 269-269 Electoral College tie with Trump. That would be all of Clinton’s 2016 states — 232 electoral votes — plus Michigan, Pennsylvania, and Nebraska’s Second Congressional District.

As we noted several times before the election, a 269-269 tie broken by the House would likely have been broken in Trump’s favor because of GOP control of individual U.S. House delegations: In a House tiebreaker, each of the 50 states gets a single vote, and the Republicans went into the election controlling a bare majority of delegations, 26 of 50. They continue to hold 26, but Democrats fell from 23 to 20 after Republicans forged ties in Minnesota, Michigan, and Iowa (and Iowa may flip depending on what happens in the uncalled IA-2 race). So we can say with a bit more confidence that a 269-269 tie would have gone to Trump.

Biden’s victory therefore belongs to his narrow margins in just Arizona, Georgia, and Wisconsin — a combined 47,000 votes or so as of Wednesday morning. Flip these states to Trump, and there is a 269-269 tie that Trump likely wins in the House.

By that token, Biden’s victory in 2020 was even smaller than Trump’s in 2016, even though Biden will easily win the popular vote after Trump lost it by two points.

The situation is reminiscent of Harry Truman’s surprise victory over Thomas Dewey in 1948. In that election, Truman ended up winning the national popular vote by about 4.5 points over Dewey in a four-way race that also featured conservative Dixiecrat Strom Thurmond and progressive Henry Wallace (more on notable third party candidacies below).

Truman’s victory in the Electoral College was markedly tighter than his popular vote margin would indicate: He won California by a shade under half a point, and Ohio by just a quarter of a point. Had Dewey won both of those states, the election would have been thrown to the House — and if Dewey had won those two plus Illinois, decided by a little less than a point, Dewey would have won.

Voters focus on the major parties

We suggested several months ago that third party candidates did not seem likely to attract as much support as they did in 2016. Donald Trump, as an incumbent president, seemed to focus the minds of both his supporters and his opponents. As it stands now, Trump and Joe Biden are attracting 98.2% of all votes cast, with just the remaining 1.8% going to other candidates and write-ins. That is up markedly from 2016, when Trump and Hillary Clinton split 94% of all the votes cast, with a larger 6% going to other candidates and write-ins.

Jo Jorgensen, the Libertarian nominee, currently is winning close to 1.2% of the vote — so about two-thirds of the total non-major party vote. This is the second-best Libertarian share ever, though well behind Gary Johnson’s 3.3% in 2016. The Libertarian tally is bigger than Biden’s margin of victory in Arizona, Georgia, Pennsylvania (at least for now), and Wisconsin. So, in a reversal of 2016, Republicans are the ones wondering “what if” about conservative third party defectors (just as Democrats were about Jill Stein Green Party voters four years ago).

As it stands now, 2020 appears likely to feature a relatively low share of third party votes compared to the 38 other post-Civil War presidential elections. Table 1 shows the third party vote in this timeframe — both the combined vote share in each election and the top third-party vote-getter in each election (we only included the top non-major party performer each year in the table for space reasons, so the table omits some notable third party candidacies, such as Eugene Debs in 1912 and the aforementioned Wallace in 1948).

Note that 2016 is in the top 10 for third party performance, while 2020 is currently in the lower half. We bolded 2016 and 2020 so you can see for yourselves.

Table 1: Third party presidential performance, 1868-2020

Note: *In 1968, one faithless elector in North Carolina cast his vote for George Wallace rather than Richard Nixon. In 1972, one faithless elector in Virginia cast his vote for Libertarian John Hospers rather than Nixon. The “Oth. EVs” column only includes electoral votes cast for third party candidates who competed in the election.

Source: Dave Leip’s Atlas of U.S. Presidential Elections

Crossover state senators decline

Heading into the election, 11 senators represented states won by the other party’s presidential candidate in 2016. The results of both this year’s Senate and presidential contests have thinned that group.

There are now only six senators, three from each party, who represent states that their party did not win in the 2020 presidential race. (This classifies the two independents who caucus with the Democrats, Angus King of Maine and Bernie Sanders of Vermont, as Democrats for the purposes of this analysis.)

Let’s go through what happened and why the number of crossover state senators declined.

First of all, Sens. Doug Jones (D-AL) and Cory Gardner (R-CO) lost in their states, both of which voted Republican (Alabama) and Democratic (Colorado) for president by double-digit margins. They were always the most vulnerable senators this election cycle and, in the end, their races didn’t feature much drama.

Michigan flipped from Trump to Biden, and Sen. Gary Peters (D-MI) won a narrow victory. Michigan’s other senator, Debbie Stabenow, is also a Democrat, so that’s two more senators (Peters and Stabenow) whose party is aligned with the presidential winner in their state.

Arizona flipping from Trump to Biden also meant that Sen. Kyrsten Sinema (D-AZ) no longer represents crossover turf, nor does Sen.-elect Mark Kelly (D-AZ). Kelly will have to defend his narrowly-won new seat in 2022 as he seeks a full term in what has become a very competitive state. Biden’s margin in the state was of course narrow.

Georgia, which it appears voted for Biden, could add to the crossover group if one or both of Sens. David Perdue (R) and Kelly Loeffler (R) hold their seats in the looming Jan. 5, 2021 Senate runoffs that will determine control of the Senate. But let’s set them aside for now.

Sen. Susan Collins (R-ME), who won an impressive victory last week with a great deal of crossover support, is joined as a Biden-state Republican by Sens. Pat Toomey (R-PA) and Ron Johnson (R-WI). Both of those seats are on the ballot in 2022: Toomey is not running for a third term, and Johnson may or may not.

Meanwhile, three Trump-state Democratic senators remain: Sens. Jon Tester (D-MT), Sherrod Brown (D-OH), and Joe Manchin (D-WV). Those senators are next up in 2024, when the Republican presidential tilt of those states will make all three attractive GOP targets.

Map 1: 2021 party control of senators by state

One other note: Only six states have a split Senate delegation; that number could increase with the Georgia runoffs if voters render a split verdict in those races (typically, though, when two Senate races are on the ballot at the same time, the same party sweeps both races).

Republicans benefit from crossover support

Republicans had an impressive night in the Senate and House: They currently hold a 50-48 edge in the Senate, with two runoffs in Georgia looming. In the House, and with several uncalled races, Republicans have significantly cut into the Democratic House majority, pushing the 235-200 advantage the Democrats won in 2018 down to perhaps 225-210 or even smaller (it may be a little while before we have a complete House tally).

The GOP accomplished their House gains — gains that came as a surprise to us and other handicappers — by both knocking off Trump-district Democratic incumbents and generating crossover support in districts that were more competitive for president.

The number of Trump-district Democrats will be markedly smaller than the 30 that Democrats held going into the election, both because Republicans knocked off many House Democrats in Trump-won districts and because Biden flipped some districts with Democratic House incumbents that Trump carried in 2016. For instance, Reps. Elaine Luria (D, VA-2) and Abigail Spanberger (D, VA-7) held on for reelection as Biden carried their Trump-won districts.

There also are going to be several Biden-district Republicans — one of them is Rep. Don Bacon (R, NE-2). The voting in that district merits a closer look.

Ever since Obama narrowly carried the Omaha-based NE-2 in 2008, it’s been a target for Democrats in the Electoral College, but Republicans retained it in 2012 and 2016. This year, though, the district seemed primed to shift blue. According to the census, just over 42% of its residents over 25 years old have a bachelor’s degree or higher, a number comparable to a state like Massachusetts. While voters don’t cast ballots based on educational attainment alone, it’s become an increasingly salient factor in elections.

Biden ultimately carried NE-2 by a 52%-46% spread, about in line with what polling suggested. But down the ballot, the district summarized House Democrats’ predicament fairly well, at least in suburbs. Despite Biden’s healthy margin there, two-term Rep. Don Bacon (R, NE-2) won a third term 51%-46% (Map 2).

Map 2: NE-2 in 2020

While the NE-2 result lines up cleanly with the national narrative that emerged last week, there were certainly some factors in this race that pointed to a Bacon win (the Crystal Ball’s final rating for the race was Leans Republican). In the closing weeks of the campaign, former Rep. Brad Ashford (D, NE-2), the district’s most recent Democratic congressman, endorsed Bacon — though for some context, Democratic nominee Kara Eastman beat Ashford in the primary when he tried to stage a comeback in 2018 and then decisively defeated his wife in May’s primary. Speaking of that May primary, at the time we noticed that Bacon received more raw votes than Trump — something fairly rare for members of Congress in down-ballot races, so perhaps that was an early indicator of his appeal. Last week, Bacon again received more votes than the president. On the third image in Map 2, Bacon tended to run further ahead of Trump in the Sarpy County portion of this district; though it sits just outside of the district, the area is home to Offutt Air Force Base, so it’s easy to see Bacon’s biography as a retired Air Force brigadier general playing well there.

Moving one state to the southeast, Democrats were excited about their prospects against Rep. Ann Wagner (R, MO-2). In 2018, then-Sen. Claire McCaskill (D-MO) carried this suburban St. Louis district by about 2.5%, and Wagner held on by four percentage points. Democrats landed a quality recruit in state Sen. Jill Schupp, who raised serious money. Democrats were hoping Biden would carry this district — instead, it looks about tied, and Wagner ended up expanding her margin from 2018, from four points to seven.

After the 2018 elections, California’s Orange County was one of the focal points for Democratic gains in the House. This populous county contains all or parts of seven congressional districts and in the California GOP’s heyday, Orange County would sometimes give Golden State Republicans — like then-Gov. Ronald Reagan — around 70% of the vote in state elections. Fueled by changing demographics and anti-Trump sentiments, Democrats swept the entire Orange County congressional delegation in 2018. As of Wednesday morning, first-term Reps. Gil Cisneros (D, CA-39) and Harley Rouda (D, CA-48) were either trailing or had lost, and both are running behind Biden in their districts and faced Asian-American women: Cisneros is in a rematch with Young Kim (R), while Rouda lost to Orange County Supervisor Michelle Steel (R).

One of the surprises of Election Night was that not only did Sen. Susan Collins (R-ME) win, but she got over 50% and avoided having to win through Maine’s ranked-choice voting system. The level of crossover voting in her race was significant: As of right now, Biden is winning the state by 8.7 points, while Collins us up by 8.9 points.

The crossover vote in Maine may be even more impressive when you factor in the House picture. In the northern ME-2, Trump held the district’s electoral vote by about 6% and Collins won it by nearly 25% in her senatorial reelection. However, the district reelected first-term Rep. Jared Golden (D, ME-2). In the Portland-based ME-1, Collins held her opponent, state House Speaker Sara Gideon (D), to just a 2.5% margin, while neither Trump nor the GOP nominee for Congress broke 40% there. (Map 3)

Map 3: Maine in 2020

In North Carolina, Sen. Thom Tillis (R) prevailed by a 49%-47% vote, defying the preponderance of polling (just like Collins did). Though this result lined up much closer with the presidential result than Maine did, Tillis outperformed Trump in the wealthier suburbs.

Let’s consider an old congressional district: the 2002 to 2010 version of NC-9. This was based primarily in the heavily white, economically affluent south Charlotte suburbs, and took in parts of Gaston and Union counties, which are more exurban. In 2008, Barack Obama would have lost that iteration of NC-9 54%-45%, and it was essentially a no man’s land for down-ballot Democrats. If this district were still in place, our unofficial calculations have Biden carrying it by about 100 votes. But Tillis’ 4.5% margin there represents something of a reversion — in Map 4, note that there are fewer blue precincts in Mecklenburg County on Tillis’ map.

Map 4: Last decade’s NC-9 in 2020

Once results are finalized, we’ll have much more to say about what happened down the ballot and how many Trump-district Democrats and Biden-district Republicans were elected. But just as in 2016, it appears that with Trump on the ballot, Republicans running in House and Senate races often ran ahead of the president.

KEY POINTS FROM THIS ARTICLE

— We are moving the Iowa Senate race from Toss-up to Leans Democratic. Iowa remains a Toss-up for president.

— Iowa may have replaced North Carolina as the race likeliest to decide Senate control, although North Carolina remains a very viable Democratic target despite Cal Cunningham’s (D) troubles.

— We have a couple of House rating changes, mostly out of an abundance of caution. We also update the House state of play.

Table 1: Senate rating change

Table 2: House rating changes

Map 1: Crystal Ball Senate ratings

Iowa and the race for the Senate

In Iowa, Sen. Joni Ernst (R-IA) is looking increasingly like an underdog against her Democratic challenger, Theresa Greenfield. At the presidential level, Iowa is looking like a true toss-up, and Greenfield seems to be running at least even with — and often better than — Joe Biden in the state.

For political junkies, perhaps one of the most anti-climactic features of the 2016 election was that, across the 34 states that saw contests, the senatorial results lined up perfectly with the presidential picture. That year, no Donald Trump-won states sent a Democrat to the Senate, or vice versa. But Iowans may be in a more split-ticket mood now — or, they may even deliver the Democrats a sweep in the topline races.

As of Monday, Greenfield’s average lead in the RealClearPolitics average was 4.8%, while Biden was up by a more meager 1.2%. Perhaps tellingly, most of that difference comes from Ernst underperforming Trump — her average polling share stands at 42.5%, compared to Trump’s 46.3%. So it makes sense that, despite his sagging national fortunes, Ernst generally is still trying to frame herself as a firm ally of the president.

For Senate Democrats, last quarter’s fundraising numbers were truly historic. For the period spanning from the start of July to the end of September, Democrats in key Senate contests across the country raised a combined $265 million. In Iowa, Greenfield set a state fundraising record — her $29 million quarterly haul was roughly four times what Ernst raised. Candidate fundraising isn’t everything — outside groups are also spending heavily in the state, and Senate candidates in past cycles have lost to lesser-funded opponents — but going into the final stretch of the campaign, Greenfield seems to have the momentum.

In 2014, Ernst had the luxury of running against a gaffe-prone candidate in then-Rep. Bruce Braley (D, IA-1). Though Braley was initially seen as a strong recruit, he fumbled early in the campaign when he was caught on tape speaking derisively about the state’s other senator, Chuck Grassley, at a fundraiser. If Republicans took the Senate majority that year, Grassley was in line to chair the Judiciary Committee — Braley dismissed the veteran senator as a “farmer from Iowa” who didn’t go to law school. Fairly or not, Braley’s campaign never really recovered from that episode, and 2014 ended up being a disastrously bad year for Democrats across the country.

In last week’s debate, Ernst had something of a Braley-esque gaffe herself. Though her down-to-home image plays prominently into her campaign, she didn’t know the break-even price of soybeans, an important cash crop for the state. Greenfield, by contrast, nailed the price of corn when asked. Senate debates across the country have taken on increasingly national tones, but we’d expect Democrats to use Ernst’s debate answer on such a local issue to cast her as out of touch with the state.

The soybeans gaffe seems symptomatic of Ernst’s reelection effort: in 2014, playing up her rural persona and vowing to “make ‘em squeal,” she rode her motorcycle onto the national scene as a political outsider. After a term in the Senate, and running in a totally different environment, Ernst’s campaign seems to have lost some of that 2014 zeal. Though Iowa hasn’t ousted an incumbent senator since 1984, we see Greenfield with the upper hand. The race is moving from Toss-up to Leans Democratic.

Moving Iowa to the Leans Democratic category before North Carolina wasn’t something we’d have expected for much of this year. In April, the Crystal Ball identified a quartet of Senate races that we identified as the Democrats’ “Core Four” states to flip. Back then, it seemed that Democrats’ most feasible path to a bare 50-seat-plus-VP Senate majority ran through Arizona, Colorado, Maine, and North Carolina — in fact, at the time, we cited Montana as the Democrats’ next best pick-up prospect. We now see Democrats as at least modest favorites in those first three states, while Sen. Steve Daines (R-MT) seems to have a slight edge over his challenger, Gov. Steve Bullock (D-MT), in Big Sky Country.

With today’s update, North Carolina is the only Toss-up race left on our senatorial board. Since news broke of his extramarital activities a few weeks ago, former state Sen. Cal Cunningham (D-NC) continues to lead first-term Sen. Thom Tillis (R-NC). In some surveys that have been conducted since the scandal broke, Cunningham has, perhaps counterintuitively (or maybe completely expectedly?), expanded his lead in some polls — although both his overall share of the vote and his lead over Tillis in polls are a little smaller than before the scandal broke, according to the RealClearPolitics average.

As we discussed recently, amidst all the dizzying news cycles of the Trump era, Cunningham’s affair may seem outright vanilla, and voters may not end up caring. But Republicans are adamant that the affair has, at minimum, hurt Cunningham’s image. Perhaps the key factor is whether the scandal remains in the headlines and what new developments emerge. Over the weekend, the joint editorial boards of the Charlotte Observer and the Raleigh News and Observer issued an unusual dis-dorsement in the race, noting that they would have endorsed Cunningham, but because of the scandal and how he has handled it, they declined to back either candidate. Given the uncertainty there, plus the state’s purple hue, we’re hesitant to write off Tillis.

ABC News reported Monday that the Trump campaign is “increasingly worried” that its chances of winning North Carolina have “all but evaporated” after categorizing the state as “super-safe” just a few weeks ago (both comments are head-scratchers to us). We can’t help but think back to 2016, though, when some of our GOP sources were despondent about the Tar Heel State, and then Trump ended up winning it by a little over 3.5 points. That said, the president arguably is behind in North Carolina, which could have major ramifications for the Senate race.

All in all, the Republicans’ defensive playing field remains large. We are curious to see if Biden makes a late, gigantic advertising play in Texas, a state that Leans Republican in our Electoral College ratings but appears close. That sort of late cash infusion could end up hurting Sen. John Cornyn (R-TX), who, at a Likely Republican rating, we have seen as more secure than many of his Senate colleagues. Texas is already nearing half of its 2016 vote total in early votes cast, although the huge, fast-growing state may blow past its raw vote tally from four years ago.

Today’s change in Iowa puts the Democrats at 50 Senate seats, with the Republicans at 49 and just a single Toss-up, the aforementioned seat in North Carolina. Combined with our Electoral College ratings, which show Biden at least narrowly favored in states and districts containing 290 electoral votes, this change formalizes the Democrats as a small favorite to win the Senate, if only thanks to a potential tie-breaking vote from a Democratic vice president.

A key question we are pondering down the stretch: Do the Democrats flip one or more of the states we currently rate as Leans Republican? If Democrats want to get beyond 51 Senate seats, they will have to, so watch states like Alaska, Georgia, Kansas, Montana, and South Carolina to see whether the Republican floodgates can hold. If any of those states flip, it might involve a state splitting its vote for president and for Senate — giving us something different than we saw in 2016.

The House

We are adding a couple of additional seats to our Likely Republican column: Reps. Doug LaMalfa (R, CA-1) and Brian Mast (R, FL-18). In the former seat, LaMalfa’s 2018 challenger, Audrey Denney (D), more than tripled the incumbent’s fundraising tally in the third quarter. LaMalfa beat Denney by 10 points in 2018. That was a smaller margin than Trump in 2016 and the GOP gubernatorial candidate in 2018, who each carried the district by roughly 20 points apiece. So this is clearly right-of-center territory. Meanwhile, Mast should be OK against Pam Keith (D), and a late September St. Pete Polls finding had him up 50%-42%, similar to his 2016 and 2018 margins of victory. But there are some signs that Trump’s nine-point victory in this district may be smaller than 2016 (the aforementioned poll had the district tied for president). Both of these seats are very much dark horse Democratic targets.

Our overall House assessment remains largely unchanged by the flood of money reported in the third quarter fundraising reports.

Our ratings continue to show 232 seats at least leaning Democratic, 185 at least leaning Republican, and 18 Toss-ups. Were we to pick the Toss-ups today — we’re not, though we will by Election Day — we’d probably predict a roughly two-to-one break in the Toss-ups toward the Democrats. That would give the Democrats a nine-seat gain over the 235 seats they won in 2018.

Given that we are going to pick all the Toss-ups before the election, we try to keep the Toss-up column as short as reasonably possible. But that means there are several shaky seats in both the Leans Republican and Leans Democratic columns.

On the Republican side, those include open seats in CO-3, MI-3, and NY-2, along with Reps. Mike Garcia (R, CA-25) and Rodney Davis (R, IL-13).

On the Democratic side, that includes Reps. Xochitl Torres Small (D, NM-2) and Ben McAdams (D, UT-4), along with open seats in GA-7, IA-2, TX-23, and TX-24.

As we shuffle the ratings in the final couple of weeks, it may be that we end up picking the current underdog in some of these seats. And, of course, we’re not going to pick every seat correctly — we just hope we’re close to the eventual net change in seats.

KEY POINTS FROM THIS ARTICLE

— Recent rosy polling for Joe Biden in the presidential race may represent an artificial sugar high for the challenger.

— But at this point, Donald Trump needs to be making up ground — not treading water or falling further behind.

— 11 rating changes across four categories of races (president, Senate, House, and governor) almost exclusively benefit Democrats.

Table 1: Crystal Ball Electoral College rating changes

Table 2: Crystal Ball Senate rating changes

Table 3: Crystal Ball gubernatorial rating change

Table 4: Crystal Ball House rating changes

The state of the presidential race

We could recap, at great length, all of the crazy events that have happened just in the week since we unveiled last week’s rating changes. But we suspect Crystal Ball readers don’t need any reminders.

Last week, we wrote that Donald Trump needed to help himself at last week’s debate more than Joe Biden did, because Trump is trailing and Biden is leading.

We did not think Trump did help himself, and the numbers we’ve seen since the debate have backed up that initial reaction. It also does not appear that the president and many top officials contracting coronavirus has made the president look better or improved his chances.

Ever since Trump took that ride down the Trump Tower escalator more than five years ago, a common question has been whether this episode or that episode would hurt his standing. In many instances, the answer has appeared to be no. But with just weeks to go until the election, asking whether something hurts Trump is the wrong question. Rather, the key question is whether something helps.

In recent days, Joe Biden’s significant lead nationally has widened. Biden’s leads in the RealClearPolitics and FiveThirtyEight polling averages are now close to 10 points apiece, and a couple of respected national polls, CNN/SSRS and NBC News/Wall Street Journal, have shown Biden leading by 16 and 14 points, respectively. The state-level numbers generally have been bad for the president, too: for instance, Monmouth University pegged Biden’s lead in Pennsylvania earlier this week at around 10 points; the pollster’s previous Keystone State survey had Biden up only a few points based on different turnout models. In other words, one of the better state-level polls for Trump in a key state was reversed in fresher polling.

At least some of the numbers we’re seeing for Biden now likely represent a sugar high for the challenger, which can happen sometimes when one candidate performs poorly or is on the wrong side of a bad story. Hillary Clinton’s national lead against Trump hit double digits in some polls following the revelation of the Access Hollywood recording in early October 2016, although Clinton’s aggregate lead over Trump in national polling as measured by RealClearPolitics hit only a high point of seven that October, and her share of the vote in the average never surpassed 49% in the fall (she ended up getting 48%).

Biden, meanwhile, has not dipped under 49% in RealClearPolitics since early August, and he was approaching 52% Wednesday evening.

To borrow a phrase from Biden, here’s the deal: The president needs the election to get closer to have a reasonable chance of once again pulling off an upset. That can still happen, but it is not happening now. The clock keeps ticking to Election Day, and voters are already voting.

Last week, we pushed some Midwestern state ratings in favor of Biden, citing his apparent gains with white voters. This week, we are moving another northern, largely white state, New Hampshire, in his favor.

The Granite State, so close in 2016, does not appear to be a major focus of the campaign, and the notoriously fickle state may be moving away from the president. Several recent polls of New Hampshire have shown Biden leading by roughly 10 points in aggregate — a shift that is similar to the kinds of improvements Biden appears to be making over Clinton’s showing across the competitive states of the Northeast and Midwest (Clinton only won New Hampshire by 0.4 points in 2016). New Hampshire is now Likely Democratic.

Of the states Clinton won, all are now rated Likely or Safe Democratic, with the exception of Nevada (Leans Democratic).

Beyond the Midwest, we now have seen enough to make two other shifts in the emerging battlegrounds of the Sun Belt: Arizona and Georgia. The former moves from Toss-up to Leans Democratic, and the latter from Leans Republican to Toss-up.

Biden’s lead in Arizona has been small but steady, generally in the neighborhood of 3-5 points in poll averages. His path to victory in the state almost certainly involves flipping Maricopa County (Phoenix) from red to blue. Maricopa, which casts about 60% of the state’s votes, was one of only a handful of the nation’s most populous counties to back Trump in 2016, but it seems primed to shift to Biden. One recent data point: the state pollster OH Predictive Insights released a poll Wednesday morning of AZ-6, a relatively affluent and highly-educated suburban district that covers Scottsdale and other parts of Greater Phoenix (AZ-6 is entirely contained within Maricopa). The district shifted from giving Mitt Romney a 21-point margin in 2012 to a 10-point margin for Trump, and Trump was only up one point in the poll.

This kind of erosion threatens the GOP’s statewide margin.

While Georgia and Arizona are a lot different — the former’s diversity is driven by Black voters, while the latter’s is driven by Hispanic voters — one commonality is that both states have a huge metro area where Biden seems poised to improve on Clinton. In Georgia’s case, that is metro Atlanta, with the suburban congressional districts GA-6 and GA-7 representing prime candidates to switch from Trump to Biden.

Metro Atlanta, however, doesn’t dominate the statewide vote the same way that metro Phoenix does: Fulton County (Atlanta) and its three biggest suburban/exurban satellite counties (Cobb, DeKalb, and Gwinnett) only cast about a third of the statewide vote, and Biden has to make up more ground statewide there than in Arizona (Trump won Arizona by 3.5 points and Georgia by 5.1 points). But Georgia is very close in polling, with Biden and Trump exchanging small leads. The state’s more of a Toss-up now.

These changes, shown in Map 1, push Biden a little further over 270 electoral votes while making Trump’s deficit a little bigger.

Map 1: Crystal Ball Electoral College ratings

The Senate

While we have rating changes in a trio of states, we’ll start with one race we’re not moving. In North Carolina — which we’d seen as arguably the Democrats’ best pickup opportunity, aside from the races that the Crystal Ball already has them favored in — the Democratic nominee, former state Sen. Cal Cunningham, was put on the defensive last week when text messages emerged indicating that he had engaged in an extramarital affair.

For now, we’re leaving the race as a Toss-up. A handful of polls conducted during and/or after the story emerged have not shown much difference from previous iterations of the same polls, and Cunningham generally still leads first-term Sen. Thom Tillis (R-NC). So this could be a case of the electorate being too polarized and/or too desensitized to care about a candidate’s personal scandals.

As of Friday, more than 300,000 early votes had already been cast in North Carolina, so both sides are starting to bank votes (for reference, in 2016, the state cast 4.7 million ballots).

Still, with the Army Reserve looking into this — Cunningham is a lieutenant colonel in the reserve — the situation seems fluid. Though the news doesn’t seem to have hampered Cunningham’s prospects yet, if it keeps escalating, at some point it may just become too much to sustain. Republicans are already working to make the story stick.

The revelations about Cunningham’s texts came on Friday evening — for state political observers, it was a development that concluded an already-eventful week in the race. On Thursday, before the two candidates squared off in their final debate, the Cunningham campaign announced it raised $28 million in the third quarter, shattering previous records. Then on Friday, Tillis announced he was diagnosed with COVID-19.

Ironically, after nominating women candidates exclusively in the four Senate races from 2008 to 2016, North Carolina Democrats are once again running a male candidate. After news of the texts broke, Cunningham was quickly compared to the state’s last male Democratic senator — John Edwards, whose electoral career was also rocked by an extramarital affair.

The bottom line in North Carolina is that we had been very tempted to move the race to Leans Democratic. The emergence of this scandal makes us glad we held off, as it is the sort of thing that could eventually undo Cunningham’s campaign (though it is not guaranteed to).

Staying in the South, we’re feeling less confident about Republicans’ chances in the special Senate election in Georgia. This election will be held under Louisiana-style jungle primary rules. Because the race features about 20 candidates, it’s unlikely that anyone will claim the majority needed to win outright in November, so a Jan. 5, 2021 runoff seems certain.

Some polls over the last few months pointed to an intraparty GOP runoff, featuring appointed Sen. Kelly Loeffler (R-GA) and Rep. Doug Collins (R, GA-9). But, with an endorsement from President Obama and by running a more active campaign, Democrats’ preferred candidate, the Rev. Raphael Warnock, seems better-positioned to at least land a spot in the runoff. Polling from Quinnipiac University puts Warnock ahead of the field, with 31% — more importantly, he claims almost 60% of the Black vote (Warnock is Black). We expect him to further consolidate that support.

Loeffler may have a slight edge for the second runoff berth, though she’s trying to get there by running to the right of Collins. This might be a good short-term strategy, but in a runoff, it may give her less room to win independents. Either way, we’re moving the race from Likely Republican to Leans Republican. The state’s regular Senate race, between Sen. David Perdue (R-GA) and 2017 GA-6 candidate Jon Ossoff (D), could also end up in a Jan. 5 runoff.

Another state we’re moving from Likely Republican to Leans Republican is Kansas. Republicans seemed to catch a break in the August primary when their voters nominated Rep. Roger Marshall (R, KS-1) over former Kansas Secretary of State Kris Kobach, a controversial Trump acolyte who lost the 2018 gubernatorial race. While Marshall may have been the safer pick, he hasn’t been able to put the race away. In polling from his allies, Marshall leads his opponent, state Sen. Barbara Bollier (D), 43%-39% — tellingly, his share is 10 percentage points less than the 53% Trump earns in the state.

Bollier, who was a Republican until 2018, seems to have a background tailor-made for Kansas’ suburban Johnson County, where she’ll need to win by double-digits. In her ads and on the campaign trail, she has pointed to her endorsements from some prominent state Republicans, most notably former Sen. Nancy Kassebaum (R-KS).

The Kansas race has recently attracted some significant investments from both sides. Last week, the Republican-aligned Senate Leadership Fund announced a $7 million buy there. Days later, the Democrats’ Senate Majority PAC pumped $7.5 million into the race. With this type of spending, it just doesn’t have the feel of a Likely Republican race anymore, even though a Democratic Senate win in Kansas would be a huge upset (Democrats have not won a Senate race there since 1932).

Finishing back up in the South, we’re moving Mississippi onto the board. The contest there features a rematch between Sen. Cindy Hyde-Smith (R) and former Rep. Mike Espy (D, MS-2), who served in President Clinton’s administration as Secretary of Agriculture. Though their 2018 contest went to a special runoff, which Hyde-Smith won by less than 8%, Espy’s margin represented Democrats’ best showing since 1988 in a Mississippi Senate race. Like many Democratic candidates, Espy’s campaign saw an influx of cash in the days following the death of the late Justice Ruth Bader Ginsburg. The anti-Trump Lincoln Project has also waded into the race.

Mississippi is a highly inelastic state: Blacks make up roughly 40% of the electorate and give Democrats near-unanimous support, while whites vote Republican with a similar intensity. But during his time in the House, Espy worked to build multi-racial coalitions, and seems to be taking a similar approach this year. We still see Hyde-Smith as a clear favorite, but we don’t think the seat is completely secure. We’re rating the race as Likely Republican now.

Overall, the battle for the Senate remains close, and Cunningham’s missteps in North Carolina could be crucial to the overall outcome. However, the playing field also appears to be expanding, and we now list six GOP-held seats in the Leans Republican column, in addition to three others where Democrats are at least narrowly favored (Arizona, Colorado, and Maine) and two that are Toss-ups (Iowa and North Carolina). Republicans remain on track to defeat Sen. Doug Jones (D-AL), and they are pushing Sen. Gary Peters (D-MI) in a race that is still Leans Democratic.

Republicans remain in the running to save their Senate majority, but there’s also a world in which the dam breaks against them.

Map 2: Crystal Ball Senate ratings

The Governors

In Vermont, two-term Gov. Phil Scott (R-VT) fits his deep blue state as well as any Republican can. One of the party’s most ideologically moderate members currently in office, Scott has received especially strong reviews for his handling of the COVID-19 pandemic. In September, Dr. Anthony Fauci, the nation’s top infectious disease expert, called Vermont’s response a “model” for the rest of the county.

Scott’s opponent is Lt. Gov. David Zuckerman — though he’s a member of the Vermont Progressive Party, he’s running with the support of state Democrats. A similar formula has worked out well for Sen. Bernie Sanders (I-VT), an independent who caucuses with Democrats. But with the COVID-19 pandemic still dominating the news cycle, it’s hard to defeat an incumbent governor, especially one who has earned bipartisan praise.

A poll released jointly by Vermont Public Radio and Vermont PBS gives Scott a 55%-24% lead. So Scott seems well-positioned to, again, transcend the state’s presidential partisanship. We’re upgrading his race to Safe Republican.

Elsewhere in New England, Democrats are hoping that they can push Gov. Chris Sununu (R-NH), who retains a big polling edge but who also may be easier to tie to national Republicans than Scott. That race remains Likely Republican.

We’re also keeping the gubernatorial race in Montana as a Toss-up — it’s the only Toss-up we have in the gubernatorial races this year. Republicans are trying to win this race for the first time since 2000, and there are some signs that their losing streak may end this year. In their June primary, Republicans passed over state Attorney General Tim Fox — who was the best-performing statewide Republican there in 2016 — for, arguably, a weaker candidate in Rep. Greg Gianforte (R, MT-AL). Still, Gianforte has led Lt. Gov. Mike Cooney (D-MT) in every poll that’s been released since July, about a handful in all.

The Democrats’ best offensive target appears to be Missouri, where Gov. Mike Parson (R) is trying to fend off a challenge from state Auditor Nicole Galloway (D). Parson appears to retain a lead of somewhere in the mid-single digits, but this is one to watch.

Overall, Missouri (Leans Republican) and Montana (Toss-up) appear to be the two races where the outcome is least certain. Trump seems very unlikely to replicate his roughly 20-point wins in each state, and he may lag significantly behind those kinds of margins. That said, Republicans probably are still better-positioned in each race, albeit not clearly enough in Montana to push the race out of Toss-up.

Democrats do appear to retain an edge in the marquee gubernatorial race of this cycle, North Carolina: Gov. Roy Cooper (D) consistently runs ahead of Biden in polls and still appears to be on track to win a second term, even if he likely will win by a smaller margin than some polls earlier this year suggested. In other words, do not expect a double-digit Cooper victory even as we remain confident enough in his chances to keep his race at Likely Democratic.

Map 3: Crystal Ball gubernatorial ratings

The House

We have just a handful of House rating changes this week, all benefiting Democrats.

Three of the four changes upgrade Reps. Chris Pappas (D, NH-1), Susan Wild (D, PA-7), and Matt Cartwright (D, PA-8) from Leans Democratic to Likely Democratic. While all three have competent GOP challengers, key Republican outside groups do not appear likely to spend big outside dollars in any of the three races, which indicates to us that the incumbents are in decent shape for the stretch run.

The two Pennsylvania seats merit special mention. Map 4 shows how the two Democratic House members performed in 2018 versus the 2016 presidential results. As the map makes clear, the president performed well in both, but Democrats did well the following cycle (these district lines were drawn in advance of the 2018 election).

Map 4: Recent results in PA-7/PA-8

Morning Call/Muhlenberg College showed Wild up 52%-39% in her race recently, with Biden up by seven. While there is no public polling of the neighboring and Trumpier PA-8, our understanding is that Cartwright is doing well, and Biden could even do well enough to flip the district, which covers his boyhood home of Scranton. In a sign of confidence, the Democratic Congressional Campaign Committee recently cut ad spending in the district. These are the kinds of districts where Trump needs to perform well — arguably, even better than he performed in 2016 — to once again win Pennsylvania, and that Republicans really need to compete for in order to win a House majority. But the region may not be as hospitable to Trump as it was four years ago.

Moving north from eastern Pennsylvania to upstate New York, Rep. John Katko (R, NY-24) appears to be in an increasingly competitive race against former college professor Dana Balter (D). Katko beat Balter by six points in 2018, making him one of the few Clinton-district Republicans to survive the Blue Wave. For much of the cycle, we thought Katko would be fine, but an increasing amount of outside spending has prompted us to reconsider. So too has a Siena College poll that showed Balter up 45%-42%. This race moves from Leans Republican to Toss-up.

Our overall House outlook — Democrats are better-positioned than Republicans to net seats — remains unchanged.