Sabatos Crystal Ball

2020 House: Assessing the Open Seats So Far

Openings won’t match the volume of 2018, but Democrats may once again benefit more than Republicans

Kyle Kondik, Managing Editor, Sabato's Crystal Ball November 21st, 2019


Dear Readers: The Crystal Ball will be away next week. Our next issue will be released on Thursday, Dec. 5. We wish all of you a Happy Thanksgiving.

The Editors


— Of 28 open House seats, Republicans are defending 20 while Democrats are defending only eight.

— Of eight the Crystal Ball rates as competitive, Republicans are defending all but one.

— Open seats, along with pending redistricting in North Carolina, give Democrats a small buffer as they defend their majority.

The open seats so far

The 2018 House election cycle was defined by an unusually large number of open seats. The 2020 cycle almost certainly won’t feature as many, but there are still more than two dozen so far. And just like in 2018, the open seats that cover competitive turf provide more opportunities for Democrats than Republicans.

The start of the holiday season is a good time to take stock of these open seats, because the number of retirements often increases as members consider their own futures with their families and as filing deadlines loom. Impeachment adds an additional wrinkle to the retirement calculus: Depending on the timing, it could cut into holiday family time, and — whatever one thinks of it — we doubt the process is making the House a more pleasant place to work.

Whoever decides to not seek reelection to the House will add to the retirements we’ve already seen this year. So far, 28 House seats are going to be open in next year’s elections, meaning that there will not be an incumbent on either the primary or general election ballot. Additionally, there are four vacancies in the House right now. We’re not counting these as true open seats, because presumably new incumbents in these seats will be seeking full terms in their own right after winning forthcoming special elections.

The postwar average, according to data compiled by Vital Statistics on Congress, is for 397 House members to seek reelection each cycle. That means that, on average, there are 38 open seats per cycle. There are 28 so far, but there’s still plenty of time for this cycle’s tally to match or perhaps even exceed the postwar average.

The open seats have a decidedly Republican lean: Of the 28 open seats, more than two-thirds (20) are currently held by Republicans, while just eight are held by Democrats.

Most of these open seats will be easy holds for the incumbent party: 20 of the 28 are rated as Safe Republican or Safe Democratic.

However, eight of the open seats are competitive to at least some degree in our ratings, and all but one of those are currently held by Republicans. This gives Democrats more of an opportunity to net seats from the list of open seats, at least so far.

While the incumbency advantage historically observed for House members may be eroding, open seats still tend to be harder for the incumbent’s party to defend than if the incumbent him or herself was running. In 2018, 12 of the Democrats’ 43 pickups came in seats where the incumbent didn’t run again. So while the bulk of the Democratic pickups came by beating incumbents — they beat 30 last year, and also flipped an additional seat (SC-1) where the GOP incumbent, Mark Sanford, ran for reelection but lost his primary — Democrats won 12 of 39 GOP-held open seats. In other words, Democrats converted about 30% of their possible open-seat opportunities. Meanwhile, Democrats won about 15% of the GOP-held seats where incumbents ran again. So the Democrats’ batting average was significantly higher among the open seats than among the ones defended by Republican incumbents.

Meanwhile, Republicans did not defeat a single Democratic incumbent in 2018 — but they did flip three Democratic open seats.

This is a long way of saying that open seats are still often more vulnerable for the incumbent party than ones held by incumbents.

Let’s take a look at these districts so far by splitting them into three groups. The first group, shown in Table 1, includes the eight open seats that we rate as at least somewhat competitive. The second features the 20 open seats where, realistically, the only competition will be in the primary. The third group features the four vacant seats, where there will be special elections later this cycle.

Table 1: Open House seats rated as competitive

Sources: U.S. House of Representatives Press Gallery; Daily Kos Elections; Cook Political Report.

The seat that stands out here is TX-23, which runs from the outskirts of El Paso to the San Antonio area and has been competitive throughout the decade. Rep. Will Hurd (R) is retiring after three terms; this is just one of three districts held by Republicans that Hillary Clinton carried in the 2016 election, and it’s also the only seat in the country where we make the incumbent party an underdog. There are many candidates on both sides, but veteran Gina Ortiz Jones (D), who came within half a point of beating Hurd last year, is running again and is sitting on a $1.4 million warchest. Meanwhile, the leading GOP candidate appears to be veteran Tony Gonzales (R), who Hurd recently endorsed. He only has a little over $100,000 in the bank. Money isn’t everything, but Ortiz Jones’ big financial edge, residual name ID from her near-miss last time, and the likelihood (although not certainty) that the Democratic presidential nominee will once again carry the district combine to make the Democrats the favorite in this district.

The one Democratic-held open seat, IA-2, gives the GOP a Toss-up target in a Trump-won district. Democrats have rallied around Rita Hart, a former state senator who was the party’s lieutenant gubernatorial nominee last year, while Republicans have state Sen. Mariannette Miller-Meeks (R) and former Rep. Bobby Schilling (R), who represented IL-17 from 2011-2013. Miller-Meeks unsuccessfully challenged retiring Rep. Dave Loebsack (D, IA-2) in 2008, 2010, and 2014, coming within striking distance of him in those latter two elections, while Schilling’s old Illinois district is separated from IA-2 by the Mississippi River.

Otherwise, these open seats largely will test the GOP’s ability to defend suburban turf in places like Long Island, Dallas-Ft. Worth, Houston, and Atlanta.

All told, the Republicans would be very fortunate to avoid a net loss among the seats listed here. But there’s always the possibility that future retirements could endanger Democratic seats and change the overall calculus. For instance, Rep. Collin Peterson (D, MN-7) holds a seat Trump won by 31 points and often flirts with retirement. If he retired, his district would move from a Toss-up to an easy Republican pickup.

Table 2: Open seats rated as Safe

Notes: * N/A means that the district did not feature two-party competition in 2018; Cook faced a fellow Republican in the general election, while Kennedy and Lowey did not face Republican opponents.

Sources: U.S. House of Representatives Press Gallery; Daily Kos Elections; Cook Political Report.

Table 2 shows the open seats rated as safe. This group includes a mix of long-serving members on both sides — although there are clearly more Republicans heading for the exits than Democrats — as well as some more recently-elected members who are seeking greener pastures: Reps. Bradley Byrne (R, AL-1), Joe Kennedy (D, MA-4), Roger Marshall (R, KS-1), and Ben Ray Luján (D, NM-3) are seeking Senate seats, and Rep. Paul Cook (R, CA-8) is running for a county supervisor position, which may actually be a greener pasture depending on one’s perspective (there have been a handful of other sitting U.S. House members from California who have attempted to make such a switch in recent years). Rep. Tulsi Gabbard (D, HI-2) is pursuing, rather quixotically, the Democratic presidential nomination.

Some of these seats could hypothetically be competitive in really bad political environments for the incumbent party. For instance, retiring Rep. Peter Visclosky’s (D, IN-1) post-industrial northwest Indiana district fell from voting for Barack Obama by 24 points in 2012 to Hillary Clinton by 13 in 2016, although it should remain Democratic in 2020. If Indiana Republicans wanted to aggressively gerrymander the state after 2020 to add to their 7-2 statewide U.S. House delegation edge, though, this district might be an attractive target to break up.

Note that there are several open Republican seats in Texas, listed both in Tables 1 and 2. For newly-elected members from these seats, redistricting could scramble things next cycle as the Republicans who probably will retain line-drawing power redesign the current districts and add a few new ones to account both for the state’s exploding population growth and for the growing Democratic strength in the state.

Table 3: Current House vacancies

Sources: U.S. House of Representatives Press Gallery; Daily Kos Elections; Cook Political Report; Greg Giroux.

Finally, Table 3 shows the vacant seats that will hold special elections later this cycle. Of these forthcoming special elections, only CA-25, from which Katie Hill (D) recently resigned, seems truly competitive. Gov. Gavin Newsom (D-CA) just set the all-party primary for March 3, concurrent with the state’s presidential primary. A big, Democratic-leaning first-round turnout could allow state Assemblywoman Christy Smith (D) to win the vote majority required to win the special election outright, but she may not have the Democratic field entirely to herself: Cenk Uygur, a progressive commentator, also is apparently running, but he is taking flak for past controversial comments. Former Rep. Steve Knight (R) is running to reclaim the seat he lost to Hill by nearly nine points last year, among other Republicans.


Overall, 2020 is not going to see the number of open seats we saw in 2018. Last cycle, only 375 of the 435 House seats featured an incumbent running for reelection, meaning that there were 60 open seats. That was the second-highest number of open seats in the post-World War II era, eclipsed only by 1992. That was a national redistricting year, and redistricting often leads to more open seats than normal years.

Last cycle, only Pennsylvania drew new lines, as the state’s Democratic-controlled Supreme Court threw out a Republican gerrymander and replaced it with a more balanced but arguably Democratic-leaning map. Democrats converted a 13-5 GOP delegation into one that was split 9-9, with three of those four Democratic flips coming in open seats and the fourth coming when Rep. Conor Lamb (D) beat Rep. Keith Rothfus (R) in a redrawn Pittsburgh-area seat (Lamb had won a special election under the old district lines before winning the incumbent vs. incumbent bout).

A similar situation is playing out in North Carolina, where a state court just threw out a Republican gerrymander. The state currently features a 10-3 Republican-dominated U.S. House delegation. North Carolina’s GOP-controlled state legislature passed a replacement map that likely would result in an 8-5 GOP delegation, netting Democrats two seats. It’s unclear whether this map will ultimately be adopted: It may be that the Democrats can get an even better map. Rep. George Holding (R, NC-2), whose district likely would become unwinnable for him under a new map, is already telegraphing his retirement. Rep. Mark Walker (R, NC-6) might also find himself in a Democratic district, although he may challenge a sitting GOP colleague elsewhere. In any event, the North Carolina remap may very well contribute to the open seat total, too. We will issue new ratings in North Carolina once the new map is finalized.

One final note: In 2018, Americans elected a 235-200 Democratic House majority. So that means Republicans need to win 18 more seats than they won in 2018 to win a House majority in 2020.

However, since the 2018 election, a few developments have effectively risen the bar for needed GOP gains. As noted above, Democrats seem likely to net at least a couple of seats because of the North Carolina remap. Also, Republicans are at least a small underdog in one of their current open seats — TX-23 — and Rep. Justin Amash (I, MI-3) left the party. So instead of Republicans having to win at least 18 currently-Democratic seats to win the House, the real number may be more like 21, and that doesn’t even include whatever effort they will need to exert in Amash’s seat.

Republicans do have a path to win the House, which largely goes through flipping a significant number of the 31 Trump-won districts held by Democrats. Such a result cannot be ruled out in an age where ticket-splitting is on the decline.

But so far this cycle, Democrats stand to benefit more from retirements than Republicans. Also, significant one-off events, like Amash’s defection and the North Carolina redistricting, are making life harder for the Republicans.

That’s why the Democrats continue to be favorites to hold the House of Representatives majority.

The Governors: Party Control Now Near Parity

GOP holds slim edge, but more Americans live under Democratic governors; crossover governors obscure stronger partisan alignment below the top ballot line

Kyle Kondik, Managing Editor, Sabato's Crystal Ball November 21st, 2019



— Following the 2019 elections, Republicans retain a narrow 26-24 edge in governorships.

— But that’s a big shift from mid-2017, when Democrats held just 15.

— Gubernatorial races are likelier to defy federal partisanship than House and Senate races.

The gubernatorial scorecard

The Louisiana gubernatorial runoff closed the book on 2019’s statewide races. Democrats held Louisiana and narrowly won Kentucky, while Republicans held Mississippi.

Map 1 shows the current party control of the governorships. They are split roughly evenly, with Republicans holding 26 and Democrats holding 24. A majority of Americans, a little less than 55%, will live under Democratic governors once Gov.-elect Andy Beshear (D-KY) takes office next month.

Map 1: Party control of governorships following 2019 elections

Since mid-2017, when Gov. Jim Justice’s (R-WV) party switch left Democrats with just 15 governorships, the party has gained nine total, while in that timeframe the only GOP gain came in Alaska, where a Republican took over from an independent last year. This trend is in keeping with what happens when a party controls the White House: They often lose ground in down-ballot offices.

Still, party control of the governorships defies federal partisanship (as expressed by the 2016 presidential results) more so than does Congress. About a quarter of all the governorships — 12 of 50 — are held by a member of the party that differs from the party that won it for president.

There are eight Democratic governors of Donald Trump-won states, and four Republican governors of Hillary Clinton-won states. Four of the eight crossover Democrats govern states that are very Republican at the presidential level (Kansas, Kentucky, Louisiana, and Montana), while the other four govern presidential battlegrounds (Michigan, North Carolina, Pennsylvania, and Wisconsin).

One of the four Republicans in a Clinton state holds a swing state (New Hampshire), while the other three are in dark blue states (Massachusetts, Maryland, and Vermont).

Meanwhile, just 11% of senators (11 total, nine Trump-state Democrats and two Clinton-state Republicans) and 8% of House members (34 total, 31 Trump-district Democrats and three Clinton-district Republicans) represent crossover territory.

Party control of state legislatures, unlike the governorships, aligns almost perfectly along presidential lines. After Democrats captured both chambers of Clinton-state Virginia’s state legislature earlier this month, there are only two state legislative chambers out of 99 total — the Alaska state House and the Minnesota state Senate — that are not controlled by the same party that won them for president in 2016. There are asterisks with both: Alaska has a Republican majority, but it has an independent speaker — who is a former Democrat — thanks to a GOP split, and Minnesota’s Senate might have flipped to the Democrats if it had been contested in 2018 (Nebraska’s single legislative chamber is technically nonpartisan, but Republicans effectively hold the majority there).

This partisanship below the top of the ballot was reinforced in Kentucky and Louisiana. Despite losing the gubernatorial races, the GOP won every other statewide race in Kentucky (the state legislature was not on the ballot this year) and, in Louisiana, Republicans expanded their state legislative majorities, winning a supermajority in the Senate and narrowly missing out on one in the House. That failure to win a House supermajority could have an impact on redistricting next year, as Gov. John Bel Edwards (D-LA) could try to push for a second Democratic U.S. House district in the state. Republicans currently hold a 5-1 majority in the state’s U.S. House delegation, but the state could accommodate a second Democratic district. Republicans could’ve overridden an Edwards district map veto had they won a supermajority, although it’s hard to know how a redistricting fight will evolve following the 2020 census, and party-line votes on that issue cannot be assumed with certainty.

There are only a relative handful of gubernatorial races next year. The big prize is North Carolina, where Gov. Roy Cooper (D-NC) is a modest favorite to win a second term in the only large state that will feature competitive races for president, Senate, and governor next year. The GOP’s best target is the open seat in Montana, and that’s also the governorship likeliest to flip.

Map 2: 2020 Crystal Ball gubernatorial ratings

Shades of Alabama in Louisiana’s Gubernatorial Runoff

The similarities between the Deep South Democratic victories by John Bel Edwards and Doug Jones

J. Miles Coleman, Associate Editor, Sabato's Crystal Ball November 21st, 2019



— In his narrow reelection win on Saturday, Gov. John Bel Edwards (D-LA) replicated some of the voting patterns displayed in Sen. Doug Jones’ (D-AL) special election victory two years ago.

— Both Edwards and Jones won statewide races despite each carrying only a single congressional district apiece.

How John Bel Edwards won a second term

For political junkies, the main event of last weekend’s Bayou State festivities was the closely fought gubernatorial runoff, where Gov. John Bel Edwards (D) held on against wealthy businessman Eddie Rispone (R). The president invested substantial political capital in this race, notably holding three campaign rallies in the state since October. Coming on the heels of a rebuff in Kentucky’s gubernatorial contest earlier this month, where he rallied for the now-ousted Gov. Matt Bevin (R), the pair of races, taken together, show the limitations of the president’s endorsement. Still, while these gubernatorial contests are certainly setbacks for the president in the short-term, the Crystal Ball keeps both Kentucky and Louisiana safely in the red column for him in 2020.

The contours of Edwards’ triumph line up strikingly well with Sen. Doug Jones’ (D) upset win in the 2017 Alabama special election over former state Supreme Court Chief Justice Roy Moore (R), an extraordinarily weak candidate. To Edwards’ delight, his beloved Louisiana State University Tigers beat the University of Alabama Crimson Tide in Tuscaloosa, AL a couple of weeks ago in college football’s game of the year, but the Louisiana governor’s playbook looked a lot like the Alabama senator’s.

Both Jones and Edwards were able to energize the black electorate in their states to a level that is uncommonly intense in the post-Obama political landscape. Breaking down both their races by congressional district, a prominent feature stands out: they won statewide despite each carrying just a single congressional district apiece.

Map 1: Louisiana governor 2019 and Alabama Senate 2017 by congressional district

In Louisiana, Edwards carried the Second Congressional District, running from New Orleans to Baton Rouge, with an overwhelming 85%. Jones put up a similarly strong 80% in Alabama’s Seventh Congressional District, which includes parts of Birmingham and Montgomery. Both LA-2 and AL-7 are majority-minority and were the sole districts in their states to support Hillary Clinton in 2016.

The turnout data also shows a similarity between the races: in both states, the Sixth Congressional Districts cast the most votes (Table 1).

Table 1: Louisiana 2019 and Alabama 2017 vote totals by congressional district

Aside from their shared moniker, the districts are demographically comparable — LA-6 is centered on suburban Baton Rouge while much of AL-6 comes from suburban Birmingham. White, wealthy suburbanites are the demographic that Democrats have made the deepest inroads with in the Trump era, so it’s logical that these districts would see relatively high voter engagement. LA-6 was Edwards’ best Republican-held district: He came within three points of carrying it, and, while Jones took just 45% in AL-6, it was a major upgrade from the punishing 26% that Clinton received there.

Conversely, looking at each state’s Fourth Congressional Districts, voter turnout suggested that both Moore and Rispone inspired cool reactions in GOP base regions. AL-4, which covers a wide rural swath of the state’s north, supported Moore by a hefty 38%, but cast the fewest votes of any district. Facing a presidential electorate, it’s unlikely Jones will benefit from this turnout dynamic, which factors into the Crystal Ball’s Leans Republican rating for the Alabama Senate race coming up next year. Days prior to the Louisiana runoff, the president held a rally in the Shreveport area, which sits in LA-4. Not only did LA-4 cast the fewest votes in the state, but the precinct where Trump held his rally actually flipped to Edwards.

Given the nature of his opposition and the hardening of 2016’s electoral fault lines, Edwards was not in a position to replicate the 56% he received in 2015 against then-Sen. David Vitter (R), a weak GOP nominee. At the precinct level, the state nearly uniformly swung red in between the elections, but the movement lined up well with what observers of national politics have come to expect (Map 2).

Map 2: Change between Louisiana gubernatorial runoffs, 2015 to 2019

In the southwestern corner of the state, Cameron Parish moved hardest against Edwards; after taking 49% in 2015, he was down to 25% on Saturday. Such acute shifts to the right were common throughout Acadiana — a heavily blue-collar region where cultural issues and the national Democrats’ perceived hostility to the oil industry likely contributed to the shift.

Edwards’ gains from 2015 came primarily in the New Orleans metro; Jefferson Parish was especially salient. One of the most striking features of the governor’s coalition was east Jefferson Parish, which includes the large suburban community of Metairie. Among the first areas in the state to vote consistently GOP, it’s produced national figures such as Rep. Steve Scalise (R, LA-1) and Vitter, Edwards’ 2015 opponent. On Saturday, Edwards essentially tied Rispone in east Jefferson, losing it by just 73 votes.

It’s easy to gloss over Orleans Parish, as it’s so safely Democratic, but its vote also was informative. Outside of a few swampy precincts in the eastern part of the parish, Edwards carried every precinct there, gaining several from 2015. In New Orleans’ wealthy Uptown neighborhood, the precinct that houses Tulane and Loyola universities saw a 118% turnout increase from 2015’s totals on Election Day — Edwards’ percentage share consummately rose from 65% to 91%. Though ostensibly just one precinct, the collegiate voters in Uptown show that, in the pursuit of his second term, Edwards was able to mobilize his necessary constituencies.