Sabatos Crystal Ball

Louisiana 2019: Welcome to the Jungle (Primary)

J. Miles Coleman, Associate Editor, Sabato's Crystal Ball October 10th, 2019

Facebooktwittergoogle_plusredditpinterestmail

KEY POINTS FROM THIS ARTICLE

— Gov. John Bel Edwards (D-LA) defied the partisan lean of his state in 2015, but he will have to navigate an increasingly partisan electorate to win again. He’ll need Republican support, but he also must energize black voters.

— Louisiana’s unique jungle primary has shaped the contours of state elections for nearly 50 years and will be a key feature of the 2019 election.

— Regionalism has always been salient in Louisiana politics, and it should be a decisive factor in which Republican candidate makes a potential runoff with Edwards: Rep. Ralph Abraham (R, LA-5) or businessman Eddie Rispone (R).

Previewing Saturday’s Bayou battle

On the first day of class, students at Louisiana State University walked into Prof. Wayne Parent’s Louisiana Government course to kick off the spring 2014 semester. As the lecture progressed, Parent, a longtime authority on state elections, summed up his observations: “One thing I hope you’ll learn from this class is that Louisiana politics is volatile.” Sure enough, that November, the state would topple its senior senator, then go on to elevate a lowly state legislator to the governor’s mansion the next year. When the lecture turned to that 2015 gubernatorial race, this legislator came up, who was then the only announced Democratic candidate. “I hear this guy John Bel Edwards is supposed to have a hell of a background,” Parent said, “but I think the state may just be too red…”

Then state-Rep. John Bel Edwards fit his conservative state about as well as any Democrat could: a pro-life, pro-gun moderate whose training at West Point became a recurring theme of his campaign. The early frontrunner, Sen. David Vitter (R), was increasingly weighed down by incumbent GOP Gov. Bobby Jindal’s horrid image in the state, as well as a personal scandal that he could never fully put to rest. In a November 2015 runoff, Edwards’ unique profile as a candidate, coupled with the GOP’s inability to unite behind Vitter, culminated in a 56%-44% win for the Democrat; this made Louisiana the only state in the Deep South to gain a Democratic governor during President Obama’s tenure.

Long dismissed as an “accidental governor” by some, Edwards is now tasked with showing that his earlier win wasn’t simply a product of favorable circumstances. Edwards inherited a budget shortfall from Jindal; he’s generally kept good, but not great, approval ratings while putting the state’s finances in order. Still, statewide races are taking on an increasingly federal timbre, which gives the GOP an opening.

The Louisiana electorate owes much of its volatile character to its political history. Often grouped in with the rest of the south, Louisiana is, culturally, a conglomeration of three states.

One of the busiest ports in the country, New Orleans is situated at the mouth of the Mississippi River; it saw a significant influx of immigrants from places like Ireland, Italy, and Germany during the nineteenth century. The result was an urban political culture similar to that of New York City.

In the southwest, the bayous of Acadiana were settled by Francophone expats from Canada, who fled British rule in the eighteenth century. Despite their distinct lineages, and different styles of gumbo, metro New Orleans and Cajun country are bound by the Catholic tradition.

Northern Louisiana was settled by Protestants and has more in common with Mississippi than either southern region of the state. Additionally, African Americans make up roughly 1/3 of the state’s population; they’re dispersed throughout the three regions. Decades of factionalism have fostered a sense of distrust in the state’s body politic. As a result, successful statewide candidates have had to build broad coalitions.

During the twenties, an upstart politician from Winn Parish, Huey P. Long, was running for governor. A Protestant, Long needed Cajun votes. So while sweeping through Acadiana, Long recalled that as a boy, “I would get up at six o’clock in the morning on Sunday, and I would hitch our old horse up to the buggy and I would take my Catholic grandparents to mass. I would bring them home, and at ten o’clock I would hitch the old horse up again, and I would take my Baptist grandparents to church.”

After a campaign event, a surprised colleague approached him. “I didn’t know you had Catholic grandparents.”

Long replied, “Don’t be a damn fool. We didn’t even own a horse.”

T. Harry Williams leads his biography of Long with this anecdote. “The story seems too good to be true — but people who should know swear that it is true,” he wrote.

Looking to Saturday’s primary, bridging the north-south divide has been a high priority for Rep. Ralph Abraham (R, LA-5). A family doctor and veterinarian by training, Abraham hails from the Monroe area and has represented the northeastern part of the state since 2015 in Congress. While the regional rift isn’t as stark today — nationalization has generated a more partisan-driven divide — most of the state’s population lives south of his 5th District.

In the 2015 gubernatorial primary, then-Public Service Commissioner Scott Angelle’s (R) Cajun base nearly put him in the runoff. He carried the southwestern 3rd District (see Map 1). Acadiana has no such “favorite son” this time. The Abraham campaign has moved to fill this void; Rep. Clay Higgins (R, LA-3) has stumped for his northern colleague. The last time Louisiana elected a governor from the northeast was in 1968; ironically, if Abraham can replicate that feat five decades later, it will likely be with Cajun help.

Map 1: 2015 gubernatorial primary by congressional district

While Abraham may make inroads outside his base region, his fundraising has been subpar throughout the campaign. This is problematic, considering his chief GOP rival, Baton Rouge businessman Eddie Rispone, has poured over $11 million of his own dollars into the race. A Republican donor with a tone reminiscent of the late Ross Perot, Rispone made millions in the construction industry. According to our sources, Rispone’s business resume may be better suited for Republicans in suburban Baton Rouge and New Orleans, as opposed to Abraham’s background as a country doctor.

Back to 2002

From a regional perspective, the 2002 Senate race may offer the best parallels to the current gubernatorial contest. That year, first-term Sen. Mary Landrieu (D) was seeking a second term. Her main opposition was from state Commissioner of Elections Suzanne Haik Terrell (R) and Rep. John Cooksey (R, LA-5) — although there were other candidates, including Tony Perkins (R), then a state legislator and now the president of the Family Research Council, a Christian conservative organization.

To retain crossover appeal in her red-trending state, Landrieu emphasized her work with then-President George W. Bush. While keeping a modicum of GOP support was important, this strategy also generated rumblings of neglect from black voters — the core Democratic constituency in the state. Edwards has likewise played up his relationship with President Trump; perhaps not coincidentally, the early vote totals suggest a less-than engaged black electorate.

On primary day, Landrieu took 46% and was forced into a runoff, which she eventually won 52%-48% (see Map 2). Edwards, likewise, has generally been polling around 45%. Cooksey, who represented the rural 5th district in north Louisiana, couldn’t find much support outside north Louisiana and was squeezed out of the runoff. Cooksey’s candidacy is a good cautionary example for Abraham, who currently holds the same seat.

Map 2: 2002 Louisiana Senate primary

Terrell, who had the backing of national Republicans, finished ahead of Cooksey in almost every southern parish, and outpolled him in the Shreveport metro. Polling from Republican pollster and state analyst John Couvillon shows that a similar scenario could take shape on Saturday. Couvillon puts Rispone ahead of Abraham 31% to 10% in the Baton Rouge area, and gives Rispone smaller leads in other southern metros.

The Law of the Jungle: A History

Abraham and Rispone have been battling for second place in recent polls — although the momentum seems to be with Rispone — and most public pollsters seem to agree that Edwards will fall short of 50% on Saturday. This would necessitate a Nov. 16 runoff under Louisiana’s jungle primary system. The jungle primary has been a feature of the state’s political landscape for nearly 50 years and has set the stage for some of its most memorable elections.

Until the 1970s, Louisiana used a more common partisan primary system. In his first term, Democratic Gov. Edwin Edwards (no relation to the current governor) proposed the jungle primary system. He framed it as a fiscally responsible move, with its potential to resolve elections after just one round of balloting, while also pitching it as a means to expand voter choice — voters would be free to select any candidate, regardless of party. A simpler explanation, though, could be found in the partisan realities of the day.

Louisiana was still overwhelmingly Democratic in the 1970s and major political conflicts fell on factional lines. Famously, in the decades following The Kingfish’s death in 1935, state politics was a tug-of-war between his loyal Longites and the anti-Long Democrats. Still, as with most formerly Confederate states, the Party of Lincoln was increasingly fielding local candidates. According to his biography, Edwards thought it was unfair that Republican candidates, who rarely faced intraparty competition, would get a “clean shot” in general elections. Democrats, by contrast, would often endure contentious and costly primaries. Edwards’ own electoral history bore this out.

An ambitious congressman from Acadiana at the time, Edwards faced a daunting three-stage election to become governor in 1972 (see Map 3). In a 17-way Democratic primary, Edwards finished first with 24%. In the ensuing Democratic runoff, his statewide career was nearly halted in its infancy by Shreveport legislator J. Bennett Johnston, who held Edwards to just 50.2%; months later, Johnston was elected to the Senate. David Treen (R) carried the banner for a nascent GOP in the general election and took 43%, the best showing for a Republican since Reconstruction.

Map 3: Stages of 1972 Louisiana gubernatorial race

Edwards’ plan found support with legislators, many of whom were entrenched in their seats and had ubiquitous name recognition in their districts. Sure enough, Edwards was reelected with 62% in 1975 after one round of voting — a break from the gauntlet he ran in 1972.

Danger signs for Edwards in the early vote

While the younger Edwards seemed poised to win outright at various points during the campaign, given the nature of his opposition, history is against him. Of the 11 gubernatorial races that have taken place since the advent of the jungle primary, five have resulted in candidates winning outright. Table 1 shows the years gubernatorial candidates have won outright, along with the percentage margin and margin over the next finisher. All but one of the five scenarios saw the winner clear 60%; Bobby Jindal took 54% in 2007, but still finished 36% ahead of the next candidate. Notably, Edwin Edwards was the last Democrat to win without a runoff, back in 1983 — a less polarized time. Given Louisiana’s red hue, it always seemed unlikely that Edwards would win in such a cakewalk in 2019.

Table 1: Louisiana governors who won outright

If past results didn’t suggest a single-round Democratic victory on Saturday, future trends certainly don’t if the early vote in any indicator. Early voting isn’t as popular in Louisiana as it is in other states, but roughly 20% of the state electorate typically casts a ballot before Election Day.

Over the week-long period of early voting that concluded last Saturday, over 374,000 votes were cast. This was the highest tally ever for a non-presidential election (Table 2), according to John Couvillon (the figures listed are as of Sunday, so there are a small number of absentee voters that have trickled in since then that are not included in the 2019 tally).

Table 2: Recent Louisiana early vote tallies

The early vote is concerning for Democrats on three fronts.

First, while it’s rare that Republicans outvote Democrats in early voting in Louisiana, they came close last week, holding Democrats to just a 1.9% turnout advantage. Partisan composition isn’t a perfect electoral indicator in Louisiana, as many older conservatives are still registered with the Party of Jackson, but such a small advantage is hardly encouraging for Democrats. Compared to the 2015 gubernatorial primary, where Democrats had a more comfortable 51-36 lead, the GOP gained in nearly every parish (Map 4). The Edwards campaign maintains that Republicans will split their ballots for him, a theme of one his closing ads, but the extent to which GOP partisans will ultimately jump ship is unclear.

Map 4: Partisan composition of the Louisiana early vote

A second danger sign for Edwards is that early returns suggest a lack of enthusiasm with black voters. Conventional wisdom is that for Democrats to win in Louisiana, they need 30% of the white vote coupled with an electorate that’s as close as possible to 30% black; the latter part of that equation is typically easier to achieve. In 2015, Edwards had a 28% black electorate in the primary; that inched up to 30% in the runoff. In 2014, Landrieu generated similarly healthy numbers with blacks, but lost because she faced an implacable white electorate. Ironically, Edwards is still competitive with whites — Couvillon pegs him at 31% — but if blacks continue to make up just 25% of the vote, as they did in the early vote, the governor will be in a perilous position.

Finally, one of the hallmarks of Edwards’ 2015 runoff victory was low turnout. He won with 647,000 votes, which was good for 56% of the vote. In the last competitive gubernatorial runoff, 2003, the late Gov. Kathleen Blanco (D) earned 731,000 votes but just 52% of the vote. With higher turnout, states tend to behave more like they would in a presidential election; in Louisiana, this movement would favor Republicans. The robust early turnout numbers could well be a cannibalization of the Election Day vote, but at the very least, they suggest an engaged Republican base.

Right on cue, as he’s done with many elections during his time in office, President Trump recently weighed in on the Louisiana race. The president took to Twitter over the weekend to announce a pre-primary rally intended to hold Edwards under 50%.

If Trump ultimately torpedoes Edwards’ reelection prospects, it wouldn’t be the first time presidents have influenced state dynamics. Though he eventually rose to become U.S. House majority leader, the late Rep. Hale Boggs (D, LA-2) was cut out of the Democratic gubernatorial runoff in 1952. Boggs’ biggest liability was his association with a deeply unpopular Harry Truman. Forty years later, ahead of the 1991 gubernatorial election, the George H.W. Bush White House was credited with enticing then-Democratic Gov. Buddy Roemer to switch parties. Ironically, Bush would ultimately have to back a Democrat that year anyway. Roemer finished third in the primary; in the famous runoff, dubbed “the Lizard versus the Wizard,” Edwin Edwards once again won the governorship by defeating former Ku Klux Klan leader David Duke. Edwards would later serve a prison sentence for corruption.

All things considered, we’ll keep our rating for Louisiana at Leans Democratic for now in advance of Saturday’s jungle primary. An outright Edwards win is seemingly less likely, but soft Republicans could conceivably save him; even a 48%-49% finish on Saturday would justify this rating. If Edwards is closer to 45%, though, the race may be more like a Toss-up.


The Electoral College: It Doesn’t Always Lean Toward Republicans

Reform efforts face long odds, and such reforms might not necessarily help Democrats in the long run

Hunter Brown, Guest Columnist, Sabato's Crystal Ball October 10th, 2019

Facebooktwittergoogle_plusredditpinterestmail

KEY POINTS FROM THIS ARTICLE

— Proposals to abolish the Electoral College, whether through a constitutional amendment or through the National Popular Vote Interstate Compact, face an uphill climb.

— Historically, the Electoral College has not given the Republicans a consistent advantage in presidential elections, but one may develop in the near future.

The long odds of Electoral College reform

Throughout our nation’s history, the only constant has been change. Our founders foresaw that when they set up the amendment system, which we’ve used to end slavery, give women the right to vote, and both enact and repeal the prohibition of alcohol. Recently, several Democrats have called for a constitutional amendment to abolish the Electoral College and replace it with a popular vote system, or failing that, to pass the National Popular Vote Interstate Compact (NPVIC). Both proposals face an incredibly uphill battle, and ending or going around the Electoral College might not necessarily benefit the Democrats in the long term. For despite the sting of losing two narrowly-decided presidential elections over the last two decades where Democrats won more votes than Republicans, it’s not clear that the Electoral College has systematically benefited Republicans over the past several decades.

The Electoral College dates back to our nation’s founding, resulting from a compromise between large states and small states. It grants each state a number of votes equal to the size of its congressional delegation. A state’s legislature then decides how to allocate said votes. Currently, all but two give all of their votes to their state’s popular vote winner. The remaining two, Maine and Nebraska, opted to give one to the winner of each congressional district and grant the remaining two to their state’s popular vote winner. In recent years, both states have had split outcomes: Barack Obama won NE-2 in 2008 while he was losing statewide in Nebraska, and Donald Trump won ME-2 in 2016 while losing Maine overall.

State-level methods of awarding electoral votes weren’t always so democratic. In fact, up until the Civil War, the South Carolina legislature simply appointed their electors. Nothing bars legislatures from doing the same today. This consideration will become important later when we discuss the NPVIC.

The Constitution sets a high bar to pass a constitutional amendment. First, an amendment must pass either two-thirds of both chambers of Congress or have two-thirds of state legislatures call for a constitutional convention, the latter of which has never happened. Then, three-fourths of the states (38) must ratify the amendment before it becomes a part of the Constitution. This option doesn’t seem very realistic. Whether the Electoral College actually benefits the Republicans systematically or not, clearly at the present moment, Democrats dislike the Electoral College much more than Republicans. Given that eliminating the Electoral College would almost certainly be a partisan exercise endorsed by Democrats and opposed by Republicans, it’s hard to see how such an effort could muster two-thirds majorities in either chamber of Congress. Additionally, there are at least 20 strongly Republican-leaning states, and it’s hard to imagine any of these states approving such a constitutional amendment. So an amendment to abolish the Electoral College would scarcely be able to gain more than 30 backers, much less the requisite 38. Hence, a more attractive option for popular vote supporters may be the NPVIC.

The plan, first conceived in 2001 by Professor Robert Bennett at Northwestern University, calls for a group of states that collectively command at least 270 electoral votes to all allocate their votes to the national popular vote winner. Because the state legislatures have the power to allocate electoral votes in whatever way they see fit, they do not have to follow the will of the people of their state, and hence can give their votes to the national popular vote winner. The compact, officially rolled out in 2006, would only go into effect once the requisite number of states have joined. It gained its first backer, Maryland, in 2007, and currently boasts a total of 196 votes. However, so far only Democratic-controlled states have passed it. Also, swing states that benefit from additional attention and outsized influence every four years may lack an incentive to pass the accord. In fact, Gov. Steve Sisolak (D-NV) vetoed legislation joining the pact earlier this year for that very reason. Only one swing state has passed the resolution to date: Colorado. However, state Republicans are pushing a ballot initiative to pull the state out of the compact, so even its inclusion is not secure. Thus, the NPVIC faces an uphill battle to passage.

Compounding its problems, the legality of the NPVIC is questionable, as it may violate the Compact Clause of the Constitution, which prevents states from entering into interstate compacts. The point of contention is whether the NPVIC legally counts as an interstate compact. There have been several conflicting legal arguments by both sides. Whether the NPVIC violates the Constitution is anybody’s guess and probably would be subject to the whims of whatever court would hear an inevitable lawsuit if the pact ever came into effect.

Furthermore, it is not clear that the Electoral College benefits Republicans consistently. To demonstrate this, I will endeavor to determine the advantage the Electoral College gave to candidates in each presidential election since 1972. I documented my findings in Table 1 below. First, for each year, the table includes the election winner, his electoral vote total, and the two-party popular vote margin against the opponent of the other major party. Then, I shifted all the states towards the national popular vote loser by this margin to see who would have won the election had the popular vote been 50-50. I also included this candidate’s electoral vote total under these circumstances. Finally, I calculated the national popular vote margin necessary for the Electoral College’s preferred candidate to lose along with the tipping point state, which I define as the state necessary for the underdog, the candidate disfavored by the Electoral College, to reach 270 electoral votes.

Note: For the sake of simplicity and to more easily demonstrate purely the effect of the Electoral College itself, I have ignored both faithless electors and the electoral vote allocation system used in Maine and Nebraska.

Table 1: Electoral College advantage, 1972-2016

Note: * is the margin or state necessary for the underdog candidate to cross the 270-vote threshold but without it, the vote would have been 269-269.

Source: Calculated by the author from results provided by Dave Leip’s Atlas of U.S. Presidential Elections.

Since 1972, the Electoral College has favored each party six times, showing no long-term Republican bias. In fact, the Electoral College has benefited Democrats more than Republicans recently, with the Democrat favored in four of the last six match-ups. It just so happens that the two that benefited Republicans occurred in years when the overall election was close enough that the split made a difference in the outcome.

Over the last 12 elections, the average Republican earning 50% of the two-party popular vote would be expected to get 276.42 electoral votes, suggesting a pro-Republican bias (average number of Republican electoral votes minus the expected 269) of 7.42 electoral votes. When only considering the last six elections, said bias drops slightly to 5.33. Since the dawn of the modern two-party system in 1856, only two elections (1876 and 2000) have been decided by less than 10 electoral votes, showing this bias to be insignificant in the long run. The average national popular vote margin needed for a Republican to win over the last 12 elections is 0.12%. Averaging over the last six increases it slightly to 0.25%. Since 1856, only two elections (1880 and 1960) have had a national popular vote margin less than 0.25%, and only one (1880) less than 0.12%. Thus, one measure favors the Republicans and another favors the Democrats, both by an insignificant margin.

However, a consistent Republican bias may develop in the near future. The populous Democratic-leaning states of California, New York, Illinois, New Jersey, and Washington are becoming increasingly more Democratic, effectively wasting extra Democratic votes in already strongly Democratic-leaning states. By contrast, the populous Republican-leaning states of Texas, Georgia, and North Carolina are moving towards the center, shedding inefficiently-distributed Republican voters. Thus, the Electoral College may begin to favor the GOP consistently in the future, although it does not seem to do so currently, and future events and electoral alignments can (and potentially will) upend current political trends.

Fighting to overturn the Electoral College may excite Democrats, but any reform faces myriad obstacles, and it may not pay the dividends many in the party would hope for.

Hunter Brown is an intern at the Center for Politics who recently graduated from the University of Virginia with a degree in mathematics. He is currently pursuing a master’s degree at UVA’s McIntire School of Commerce. Contact him at hab3zb@virginia.edu.

 


The Senate: Ratings Changes and the Shadow of Impeachment

If there’s a trial in the upper chamber, who might feel the heat?

Kyle Kondik, Managing Editor, Sabato's Crystal Ball October 3rd, 2019

Facebooktwittergoogle_plusredditpinterestmail

KEY POINTS FROM THIS ARTICLE

— Nationalization is an increasingly important trend in American election outcomes. It’s hard to think of a more nationalizing issue than a presidential impeachment.

— Vulnerable members on both sides in the Senate will have a lot to consider if and when they have to cast a vote on convicting President Trump in a potential Senate impeachment trial.

— There are two Senate ratings changes this week, one benefiting each side. The most vulnerable senator, Sen. Doug Jones (D-AL), moves from Toss-up to Leans Republican, while Sen. Thom Tillis (R-NC) moves from Leans Republican to Toss-up.

Table 1: Crystal Ball Senate ratings changes

Senator Old Rating New Rating
Doug Jones (D-AL) Toss-up Leans Republican
Thom Tillis (R-NC) Leans Republican Toss-up

Map 1: Crystal Ball Senate ratings

Ratings changes and impeachment

A major overall theme in American political life is the nationalization of politics. How people feel about the president is bleeding down the ballot to an extreme degree, to the point where congressional expert Gary Jacobson observed that the 2018 midterm was “the most sweeping national referendum on any administration at least since the Great Depression.” This helps explain the 2018 results, when Democrats swept the House but lost ground in the Senate because they held too many seats in states that had trended so far to the right in presidential races.

This trend showed up in 2016, too. Every state with a Senate race voted for the same party for Senate and for president for the first time in the history of Senate popular elections.

Nationalization seems very germane when assessing the Democrats’ drive for impeaching President Trump in the House and the possibility of an impeachment trial in the Senate.

In what is already a nationalized political time, impeachment may be the ultimate nationalizing event for members of the House and the Senate. An up-or-down vote on the continuation of the Trump presidency in advance of a presidential general election would be an unprecedented event in our lifetimes, and it would be occurring in a time when down-ballot politics is more defined by opinions of presidents than in any time in most, if not all, of our lifetimes as well.

It would make sense, therefore, that the impeachment vote would most potentially threaten members of Congress who occupy seats in enemy partisan territory.

Let’s take a look at the Senate, where the president would be tried if the House impeaches him.

Only a little more than a third of the Senate is up in 2020 (35 out of the 100 seats), so while certain members of the Senate not up for reelection next year might be unpredictable votes on impeachment, like Sens. Mitt Romney (R-UT) or Joe Manchin (D-WV), let’s just focus on the seats up next year. If there is a major political fallout over impeachment one way or the other, one would expect it to happen in this cycle as opposed to future ones, when this impeachment battle may be, politically speaking, ancient history (and it may be ancient history by November 2020, given how quickly the news cycle moves these days).

The senator on the ballot next year who most clearly represents the enemy territory we referred to above is Sen. Doug Jones (D-AL). He most clearly needs to attract a significant amount of crossover support next year, given that Trump won Alabama by 28 points in 2016. Jones can’t be looking forward to a possible impeachment vote, because he’d either let down his party base if he votes to acquit, or burn some already rickety bridges with Trump voters if he votes to convict.

In truth, our Toss-up rating in Alabama was being generous to Jones; it should be more like Leans Republican, particularly in a hyper-nationalized political environment. While it’s true Jones may again face Roy Moore, his disastrously bad GOP opponent from the December 2017 special election, in all likelihood another Republican should win the nomination, and even Moore vs. Jones would probably be a Toss-up, at best, for the Democrat. So we’re moving this race to Leans Republican, formalizing Jones as the most vulnerable incumbent of either party next year. The one possible exception might be Sen. Ed Markey (D-MA), who faces a stiff primary challenge from Rep. Joe Kennedy (D, MA-4), but that seat is safe for the Democrats either way.

Other Democrats up next year who may have trouble with a Senate vote represent Trump’s smallest win by margin, Sen. Gary Peters (D-MI), and his smallest loss, Sen. Jeanne Shaheen (D-NH). We have both rated as Leans Democratic, matching our Electoral College ratings in those states. Neither appears to be leading the Democratic drumbeat on impeachment.

Caution is also evident so far in the cases of Sens. Cory Gardner (R-CO) and Susan Collins (R-ME), who are the only two Republicans up for reelection next year from states Hillary Clinton won in 2016 — in fact, they are the only two Clinton-state Republican senators, period. But their states are not quite as blue as commonly believed. Colorado’s turn toward the Democrats is only recent, and Clinton carried it by five points — a respectable victory margin, for sure, but not an overwhelming one. Maine, meanwhile, only backed Clinton by three, very similar to her national margin of victory. Gardner and Collins are keeping their opinions close to their vests, and the nationalization trend could hurt both of them: Collins has lost some of her once-overwhelming crossover appeal as part of the fallout from the last great nationalizing moment in American politics, the confirmation battle over Supreme Court Justice Brett Kavanaugh last year. A vote to acquit Trump might nationalize Collins even further, although at the moment we’re keeping her race rated Leans Republican. Gardner, meanwhile, has generally positioned himself as pro-Trump following the president’s election (although he doesn’t yell it from the rooftops, understandably), and his state could very well get bluer in 2020, which puts him in real trouble. Gardner’s seat is the second-likeliest to flip parties next fall, behind only Jones, although we’re not moving him out of the Toss-up category yet.

Two other vulnerable Senate Republicans, Sens. Martha McSally (R-AZ) and Thom Tillis (R-NC), seem to be more clearly supportive of the president in the impeachment battle in the early going, and part of the reason may be that they both have at least somewhat serious primary challengers. Tillis, in particular, appears to be sweating the challenge from businessman Garland Tucker (R), as the incumbent has dropped $2 million on a pre-primary ad buy and is highlighting his backing from Trump. McSally and Tillis should both be OK in their primaries, but pressure from the right likely will keep them in line with their party on an impeachment vote, which could deprive them of crossover support in states that backed Trump in 2016 by about 3.5 points apiece but appear to be live targets for the eventual Democratic presidential nominee next year. McSally was already firmly in the Toss-up column, and we’re moving Tillis to that rating too. His personal favorability numbers are not good, and his primary challenge is pushing him to embrace Trump strongly, which may or may not be the right move in the long term. The Democratic field is uncertain, but the preferred pick for D.C. Democrats appears to be Cal Cunningham (D), a veteran and former one-term state senator who lost a primary for U.S. Senate in 2010. Cunningham does not have the profile of an obvious top-tier U.S. Senate recruit, although the last two winners of this seat, Tillis and former Sen. Kay Hagan (D), didn’t have elected experience beyond the state legislative level either.

One final point: While there are potentially vulnerable senators here we’ve left out, such as Sens. David Perdue (R-GA), Joni Ernst (R-IA), Tina Smith (D-MN), and John Cornyn (R-TX), most of the others are only hypothetically vulnerable in primaries. We have no clue when a Senate trial and vote on removing the president might occur, but the timing may be important depending on whether it comes before or after a state’s filing deadline. If there are senators on either side of the aisle who decide they want to buck their party on impeachment, it might be an easier decision if they could cast this vote after major party candidate filing deadlines have passed. Some of these deadlines — helpfully compiled by Daily Kos Elections and listed in Table 2 along with the Senate seats on the ballot next year — are coming up as early as November, although others don’t come until the middle of next year. In fact, only eight of the 34 states with Senate elections next year — Georgia has two — have filing deadlines that come before the end of January.

So if the impeachment wheels turn quickly, many members on the ballot next year may have to cast their votes prior to their states’ filing deadlines, which means that if there are political consequences from the decisions they make, that could lead to the emergence of new general election or primary challengers.

This may have the effect of hardening party discipline in the Senate on an impeachment vote, although given the nationalization and polarization of our politics, strong party discipline should be the baseline expectation anyway.

Table 2: Filing deadlines in 2020 Senate states

Notes: * indicates member is retiring at the end of next year. ** indicates a special election for 2020; Martha McSally (R-AZ) is running for reelection to the Senate after taking over for the late John McCain (R-AZ) last year, while Johnny Isakson (R-GA) is resigning at the end of this year. A previous version of this table had the incorrect date for Kentucky, which has been moved up to Jan. 10, 2020.

Source: Filing deadlines are from Daily Kos Elections’ compilation.

Conclusion: Impeachment’s uncertain directions

We would be lying if we said we had a great sense as to how the Democrats’ drive to impeach Trump will impact the elections next year. There are just too many variables and moving pieces to feel strongly about it. But the potential for the battle to harden partisan attitudes may have down-ballot effects for some members of Congress, as noted above.

But don’t be surprised if, for all the noise, the impeachment inquiry — and even a successful House vote for impeachment and subsequent Senate trial — does not lead to sharp changes in public attitudes. It is true from limited polling that, at the very least, Democrats are coalescing around impeachment after the revelation of the now-famous telephone readout between Trump and the Ukrainian president. What seems to be happening is that Democrats are taking their cues from party leadership, which has resisted calling for impeachment until now, and increasing their own support for impeachment as a result. There has been some movement in favor of impeachment among independents and Republicans, although one would have to cherry-pick data to argue that overall support for impeaching and removing the president is significantly more than mixed.

Meanwhile, the president’s approval rating — as it seems to do — has remained largely fixed where it’s been, in the low-to-mid 40s, with disapproval over 50%. Could the Ukraine bombshell and subsequent discoveries from the impeachment process cause it to dip over time? Sure.

But after years of observing the president’s durability in polls, thanks in large part to strong GOP support, it’s safer to expect continuity as opposed to change in the president’s standing.