Who are the most vulnerable House members heading into the 2020 elections?

Texas Rep. Will Hurd, the most vulnerable House Republican heading into 2020

Democratic House candidates performed surprisingly well in 2018, picking up a net of 40 seats. What’s not so surprising is that the number of flips hit double digits, or even that the Democrats were able to flip enough seats to take back the House majority; those things often happen in the first midterm after control of the White House changes hands. That’s because the party out of power (and its supporters) views the new arrangement as an existential threat and votes accordingly, while the party in power usually devolves into a mix of infighting and complacency.

What was somewhat surprising was that the number of flips went all the way up to 40. When toting up all the votes received by Democratic candidates nationwide versus those won by Republicans, Democrats came out on top by a rather wide 53.4 to 44.8 margin, a difference of 8.6 points. That gap, however, would ordinarily have translated into a gain of fewer than 40 seats.

For instance, political science professor Alan Abramowitz’s historic model for converting the national House vote (as predicted by what are known as generic congressional ballot polls) into House pickups predicted only a 32-seat increase, and some pundits thought even that sort of projection was too optimistic, since historic models don’t take into account the sophisticated level of Republican gerrymandering in many states. So if Abramowitz’s model is right, either a lot of Democratic candidates in close races last year got lucky breaks at the end, or Democratic strategists did an unusually good job of deploying their resources in a useful way around the margins.

Having control of the House with a sizable cushion, of course, is a great thing for Democrats, but there’s also the inevitable downside: The party picked up a lot of seats in swingy or outright Republican-leaning districts, so Democrats are probably going to be playing a lot of defense in 2020. It’s worth noting, though, that when a party has made major House gains in a midterm, it tends not to give a lot of them back in the following election. For example, after the wave elections of 1982, 1994, and 2010, the House elections of 1984, 1996, and 2012 were pretty close to washes.

One thing that 1984, 1996, and 2012 all have in common, however, is that the party that got pounded in the previous midterm bounced back rapidly, with its president winning re-election. A different template—that’s much more promising for today’s Democrats—would be the 1980 election, where the incumbent president lost by a wide margin.

In so doing, Jimmy Carter also dragged his party down in the House, as the Republicans gained 34 House seats that year in addition to winning the White House, on top of their gains from 1978. It’s worth noting, though, that GOP gains in the House in 1978 were unusually lackluster, given that Carter was already pretty unpopular at that point; they only gained 15 seats that year and didn’t come anywhere near a majority, so some of their 1980 pickups may have just been deferred from ‘78.

Given that Donald Trump’s re-election prospects look similarly dicey, further growth in the size of the Democrats’ House majority should be on the table.

So, next question: Which Democrats are most likely to find themselves in difficult fights in 2020? And who are the Republicans who, if the tide keeps pushing inland, would be next to fall? Luckily, we at Daily Kos Elections have developed a metric for assessing just that: the House Vulnerability Index, which combines data from previous elections to assess which seats are in the greatest danger in the next election.

The House Vulnerability Index doesn’t predict how many seats will be lost in the next election, and it doesn’t predict whether or not a particular seat will be lost. What it does, though, is predict the order in which seats will fall. We’ve deployed the HVI a number of times before (in 2016, 2014, and 2010), and it’s always worked well at describing which seats are the most vulnerable for each party.

For example, in 2018, of the 10 Republican-held seats that the HVI deemed most at-risk, Democrats picked up all of them. Of the top 25 most vulnerable GOP seats, the Democrats picked up 22. The only ones they missed out on were Texas’s 23rd, Pennsylvania’s 1st, and Nebraska’s 2nd district, each of which the Republicans won by only a couple points.

In fact, of the 42 seats that the Democrats picked up (keep in mind they lost two, for a net of 40), only three of those were outside of the top 67 most vulnerable seats. Those three came in Oklahoma’s 5th (which was a massive surprise), Utah’s 4th (which featured a very strong Democrat and a deeply flawed Republican incumbent), and Texas’s 32nd District (which exposes a design flaw in the model that we’ll get to later). On the Democratic side of the ledger, the two seats the party lost, Minnesota’s 1st and 8th Districts, were literally the two most vulnerable Democratic-held districts in the nation, according to the HVI.

How the index works is extremely simple. It relies on only two factors: the average lean of the district in the previous two presidential elections, and the incumbent’s margin of victory in the last House election. Each factor is rank-ordered, meaning that the bluest Republican district according to presidential lean gets a rank of 1, the next-bluest a rank of 2, and so on. (For Democratic seats, the reddest is 1, and so forth.)

Likewise, the race with the closest margin of victory gets a rank of 1, the second-closest a rank of 2, etc. For the 2018 edition of the HVI, those margins came from the 2016 elections, except in cases where more recent special-election results were available, such as in Georgia’s 6th.

This, incidentally, is why the model can understate the vulnerability of a district like Texas’ 32nd. In 2016, Republican Rep. Pete Sessions faced no Democratic opponent, allowing him to rack up what would prove to be an inflated 52-point margin over a hapless Libertarian foe. Oftentimes, when an incumbent fails to draw a major-party opponent, it’s a sign of strength (or just overwhelming district lean), but Sessions—who lost to Democrat Colin Allred—proved to be the exception.

There’s also one final, important tweak: For open seats, we consider the previous margin to be zero and accordingly give them all a rank of “0.” Historically, open seats are when a party is at its most vulnerable, because the incumbent’s name recognition and goodwill no longer apply.

The goal with the index is to surface the incumbents who are in swingy districts or districts that are trending away from them (as seen with the presidential numbers), and who’ve had their own difficult elections recently. That way, it downplays incumbents who are in fairly safe districts but who won by a diminished margin last time, as is often the case with freshmen who narrowly win open-seat races but might be more protected by their districts’ natural lean in the future.

So let’s take a look at who’s actually the most vulnerable, starting with the 25 most vulnerable Democrats for 2020. (You can also click through to our full spreadsheet, where you can see every member as well as see our full calculations and underlying data.)

HVI RANK	DISTRICT	Incumbent	DIST. LEAN	‘18 MARGIN	TOTAL HVI
1	UT-04	McAdams	1	1	2
2	GA-06	McBath	5	3	8
3	OK-05	Horn	3	6.5	9.5
4	SC-01	Cunningham	4	6.5	10.5
5	NM-02	Torres Small	7	10.5	17.5
6	NY-22	Brindisi	9	9	18
7	VA-07	Spanberger	8	10.5	18.5
8	MN-07	Peterson	2	17	19
9	ME-02	Golden	24	4	28
10	MI-08	Slotkin	13.5	15	28.5
10	TX-07	Fletcher	6	22.5	28.5
12	NJ-03	Kim	25	5	30
13	IL-14	Underwood	10	20.5	30.5
14	VA-02	Luria	19	12.5	31.5
15	CA-45	Porter	18	16	34
16	TX-32	Allred	11	28	39
17	NJ-07	Malinowski	21	20.5	41.5
18	CA-48	Rouda	12	30	42
19	MI-11	Stevens	13.5	29	42.5
20	IA-03	Axne	31	12.5	43.5
21	NY-11	Rose	20	27	47
22	MN-02	Craig	27	25	52
22	NY-19	Delgado	28	24	52
24	CA-39	Cisneros	39	14	53
25	KS-03	Davids	15	40	55

Perhaps not surprisingly, only one of the 25 most vulnerable Democrats is not a freshman. If you’re a regular Congress watcher, you also won’t be surprised who it is: Collin Peterson, who’s held Minnesota’s rural 7th District since time immemorial even as his seat has grown redder at the presidential level, delivering him ever-decreasing margins in his last few elections.

Peterson (who’s 74, and one of the few remaining core Blue Dogs) seems like a likely retirement possibility—if not in 2020, then in 2022, when Minnesota is on track to lose a seat in reapportionment following the 2020 census. If he calls it quits in 2020, the 7th (currently the second most Republican seat held by a Democrat, using presidential numbers) would move into a tie for the most vulnerable Democratic seat. (If he bails in 2022, it’ll be less of an issue, because there won’t be a 7th district in Minnesota, at least not in its current configuration.)

The single most vulnerable Democrat is a freshman, and it’s Ben McAdams, the former Salt Lake County mayor who turfed out Republican incumbent Mia Love by a margin of just 0.2 percent in Utah’s 4th district. Not only was that the narrowest 2018 win by a Democrat, but Utah’s 4th is also now the reddest Democratic-held district, according to our method of averaging out the margins in the previous two presidential levels.

Digging into the numbers, this district went for Donald Trump by only 7 points in 2016 (thanks largely to Mormon hostility to Trump’s amoral ways), but it went for Mitt Romney by 37 points in 2012 (a margin that was inflated due to Romney’s favorite-son status), for an average of 22 points. It’s also worth noting that Minnesota’s 7th is the reddest Democratic-held district if you look only at the most recent election: Romney won the 7th by 10 points, but Trump carried it by 31.

Our method tries to strike a balance between districts that are very Trumpy, like Minnesota’s 7th, versus those that are very Romney-y, like Utah’s 4th, which is not only a well-educated suburban seat but one that’s heavily Mormon. That’s also why Jared Golden, the freshman Democrat in Maine’s 2nd District, may look safer than you’d initially expect: His seat went for Trump by 10 but for Obama by 9, so it averages out into what looks like a rather generic, evenly divided district, even though it’s actually prone to wide swings.

Following McAdams are the three freshmen who won probably the most surprising victories: Lucy McBath in Georgia’s 6th, Kendra Horn in Oklahoma’s previously mentioned 5th, and Joe Cunningham in South Carolina’s 1st. Incidentally, McBath, Horn, and Cunningham are three of the five winners in races that Daily Kos Elections had rated “Lean Republican or “Likely Republican” last year but where the Democrat nevertheless won. The others were Max Rose in New York’s 11th, which clocks in at 21st on the HVI, and T.J. Cox in California’s 21st—more on him in a moment.

By contrast, McAdams’ win wasn’t as surprising, even though we knew it’d be close; there was ample warning from district-level polls that Love was in some serious trouble, and we considered that race a “Tossup.” They, plus Xochitl Torres Small in New Mexico’s 2nd and Anthony Brindisi in New York’s 22nd, will probably be the Democrats’ most difficult holds in the 2020 cycle.

One quick note about someone who isn’t very vulnerable: that would be Cox, who also won one of 2018’s most surprising victories, narrowly defeating Republican Rep. David Valadao. Despite the Democratic lean in California’s 21st at the presidential level, few saw this win coming, between district-level polls, outside spending, and Democrats’ previous poor track record in this heavily Latino district in midterm years. But a Latino turnout surge last year helped get Cox over the top, and now, despite having won only the second narrowest victory of 2018, he’s pretty well insulated, thanks to the district’s status as only the 69th reddest Democratic-held district. As a result, Cox is just the 34th most vulnerable overall.

Now let’s turn to the 25 most vulnerable remaining Republicans:

HVI RANK	DISTRICT	Incumbent	DIST. LEAN	‘18 MARGIN	TOTAL HVI
1	TX-23	Hurd	3	2.5	5.5
2	PA-01	Fitzpatrick	2	8	10
3	IL-13	Davis	5	5.5	10.54
4	NE-02	Bacon	9	7	16
5	MN-01	Hagedorn	17	4	21
6	NY-24	Katko	1	25.5	26.5
7	PA-10	Perry	20	9.5	29.5
8	NY-01	Zeldin	13	17	30
9	MI-06	Upton	10	22.5	32.5
9	WA-03	Herrera Beutler	7	25.5	32.5
11	OH-01	Chabot	14	21	35
12	NY-02	King	4	32.5	36.5
13	NC-09	VACANT	37	0	37
14	MN-08	Stauber	11	28.5	39.5
15	IL-12	Bost	16	32.5	48.5
16	OH-12	Balderson	32	18	50
17	FL-15	Spano	21	30.5	51.5
18	NC-13	Budd	23	30.5	53.5
19	GA-07	OPEN	53	0	53
20	FL-18	Mast	15	43	58
21	VA-05	Riggleman	25	35	60
22	PA-16	Kelly	43	19.5	62.5
23	MO-02	Wagner	47	16	63
23	NC-02	Holding	34.5	28.5	63
25	NY-23	Reed	22	42	64

Not surprisingly, Will Hurd, in Texas’ Latino-majority 23rd District, is the most vulnerable Republican in 2020. Hurd won by only four-tenths of a point against Gina Ortiz Jones in 2018, and he sits in one of just three remaining districts that Hillary Clinton won in 2016 that are still GOP-held. After squeaking into office against a Democratic incumbent in 2014, Hurd somehow managed to survive perfect storms in both 2016 and 2018; we’ll have to see if his luck keeps holding in 2020.

The other two “Clinton-GOP” districts left are New York’s 24th in the Syracuse area and Pennsylvania’s 1st, in the Philadelphia suburbs. The former is the bluest district still held by a Republican: Clinton won the 24th by 4 points, while Obama carried it by a wide 16-point margin in 2012. However, this one is lower on the list, at sixth most vulnerable, because Republican Rep. John Katko won by a more substantial margin in 2018—partly because he’s popular locally and fairly moderate, partly because he faced a weaker opponent compared to Hurd’s challenger.

At the other end of the spectrum is Georgia’s 7th District, located mostly in the rapidly diversifying suburb of Gwinnett County. Republican Rep. Rob Woodall seemed to coast through his last campaign (alarming GOP election-watchers even before Election Day) and only barely won against Democrat Carolyn Bourdeaux by just a 50.07-49.93 margin after a recount—the closest House race in the nation.

Woodall apparently saw the handwriting on the wall and recently announced his retirement. But even factoring in what is now an open seat, the HVI still gives the GOP a decent shot at retaining the district, which clocks in at only the 53rd bluest district held by a Republican, especially if the party lands a less somnolent candidate. However, that might be deceptive: Like many well-educated Sun Belt suburbs, the 7th moved rapidly in the Democrats’ direction between 2012 and 2016, going from a 22-point Romney victory to only a 6-point Trump win, and in last year’s race for governor, Democrat Stacey Abrams actually carried it, 50-49. If that trajectory continues, the GOP will face serious trouble here.

Also of interest is New York’s 2nd District, and, just hanging off the edge of this table, Texas’s 24th District, which shows up as the 29th most vulnerable seat. These are the two most vulnerable districts, where the 2018 races were ones that we (and almost all other prognosticators) rated “Safe Republican.” These districts ended up with closer-than-expected races in 2018, though, and the district leans in each case give the Democrats some hope as well. New York’s 2nd has always been a swingy district, and Texas’ 24th is a suburban district that moved rapidly toward Clinton in 2016. On top of that, New York Rep. Peter King is rumored to be considering retirement in 2020, which would (given its tilt) move this district right to the top of the vulnerability index.

Those races contrast with two that we had at “Tossup” going into 2018, but where the district lean probably protects the incumbents going into 2020: Kentucky’s 6th District, which featured an unusually compelling Democratic challenger in Amy McGrath, and Kansas’s 2nd District, which was an open seat and also drew a strong Democratic candidate, Paul Davis. But these are pretty solidly red districts at the presidential level, so despite winning narrowly in 2018, the Republican incumbents in each case look safer for 2020, with Kentucky’s 6th and Kansas’ 2nd clocking in at 31st  and 34th most vulnerable, respectively.

Similarly, the lonely two Republican freshmen who picked up seats—Jim Hagedorn in Minnesota’s 1st and Pete Stauber in Minnesota’s 8th—appear to be less at risk than a number of veterans who are in swingier districts. Hagedorn and Stauber are only 5th  and 14th most vulnerable, despite their freshman status. It’s worth noting, though, that these districts are in transition, not historically Republican: Obama won them both in 2012, though Trump won both by double digits in 2016. The average margin could therefore prove misleading if these districts return to their previous Democratic form, though a Republican victory in a recent state Senate special election in territory overlapping the 8th suggests that we shouldn’t hold out much hope of seeing this.

Finally, of course, these numbers aren’t set in stone. There’s one important variable that really hasn’t come into play yet, and that’s retirements, which lead to open seats. Except for the previously mentioned 7th District in Georgia, and a few districts that aren’t likely to be competitive regardless of whether they’re open, like Utah’s 1st, we just don’t have many open seats yet.

But every time an incumbent does decide to hang up their spurs, or ride off into the sunset, or whatever other Western-themed metaphor you’re keen on, we’ll immediately update our spreadsheet so that we can see how the playing field is affected—so keep us bookmarked. And we’ll be checking in again on the House Vulnerability Index once again at this point next year, to see how things have changed once you factor in open seats.