Who are the most at-risk House members in November? Our House Vulnerability Index will tell you

Texas Rep. Mayra Flores is the most endangered Republican incumbent in 2022, according to the Daily Kos Elections House Vulnerability Index.

One of the many projects at Daily Kos Elections that we undertake every two years is calculating our home-grown House Vulnerability Index, which we’re pleased to launch for the 2022 election cycle. This deceptively simple tool tells us a great deal about which House members of both parties are going to face the most difficult re-election fights in their next elections, using data from previous elections.

To do so, it relies on just two data points: the margin of victory in each incumbent House member’s own race in the last election, and the average presidential margin of victory in that same congressional district over the last two presidential elections. This accounts for two separate issues. One, with the recent historical decline in ticket-splitting by voters, the presidential lean of a House district is very determinative of that district’s vote for Congress. There are, in fact, almost no representatives left who represent a district won by more than 10 points by the presidential candidate from the opposite party, and few who hold districts won by the other party’s presidential nominee by any margin at all.

The other consideration is that, despite the dwindling of ticket-splitting, candidates still matter. Candidate quality can’t help an unusually talented and hard-working Democrat beat a Republican in a district that’s 60-40 for the GOP at the presidential level. But it does matter around the margins, in that fairly narrow slice of swing districts where all the action is. That’s why we also factor in an incumbent’s previous election, since it can help isolate who are the chronic underperformers or overperformers—factors that can, at least to an extent, mitigate a district’s lean.

The House Vulnerability Index (HVI) isn’t a crystal ball to the extent that it tells us who specifically will win or lose in the next election. Rather, it tells you the likely order in which one party’s House seats are apt to fall. For instance, if you expect that Democrats are on track to lose 15 House seats, the HVI provides a rank ordering of which 15 are the likeliest seats for them to lose.

This cycle, however, presents a rare hurdle: redistricting. Because borders were redrawn for the nation’s House maps, many incumbents are running in districts that have been significantly reshaped from those they used to occupy.

That creates a number of problems. One, because of reapportionment, a handful of districts have simply disappeared (goodbye, New York’s 27th and Pennsylvania’s 18th!), while new districts have popped up to replace them in other states. We don’t include these seven brand-new districts in our analysis at all, on the theory that neither party is “defending” them, so they can’t be considered “vulnerable.”

(Two, however, are especially competitive—Oregon’s 6th and Colorado’s 8th—while another pair—Montana’s 1st and Florida’s 15th—bear watching, if not this year, then in the future. North Carolina’s 14th and Texas’ 35th, meanwhile, should be easy Democratic wins, while Republicans will have no trouble picking up Texas’ 38th.)

Another issue is that the presidential lean of many districts has changed considerably. In some cases, representatives who were used to coasting now find themselves in competitive races, or in a few cases, as outright underdogs in districts that now go the “wrong” way. In a lot of other cases, a number of House members who had rough 2020 races had their districts shored up thanks to gerrymandering by friendly lawmakers and face much easier races than last time.

To account for this, we’ve made a small mathematical adjustment to incumbents’ 2020 margins, to reflect how much their districts changed. We do this by comparing the average of the 2016 and 2020 presidential elections in each district under the old lines with the average of how those two elections would have gone under the new lines.

Sometimes that shift is very small. Maine’s 2nd is one of the toughest districts held by a Democrat, Rep. Jared Golden. But it also barely changed during redistricting, since lawmakers only made minimal changes in order to balance out the population between the state's two districts. As a consequence, Golden’s district went from 52.3-44.8 for Donald Trump in 2020 using the old lines to 51.6-45.5 for Trump under the current lines. Similarly, the 2016 results also moved a hair, from 51.4-41.1 for Trump to 50.6-41.6 for Trump.

On average, then, the district saw an improvement (from the Democratic perspective) of 1.3 points. In 2020, Golden won re-election by 6.1%, so we bump that figure up to 7.4% to reflect the fact that the 2nd District is now a bit bluer.

But you can see a much bigger effect in, say, Texas’ 24th District in Dallas’ northern suburbs, an area that’s moved decidedly to the left in the last half-decade. Republican freshman Beth Van Duyne won her first race by only 1.3 points in 2020, against the backdrop of a 51.9-46.5 win for Joe Biden in her district on the very same day—a big jump from Trump’s 50.7-44.5 victory just four years earlier.

However, Republicans in the legislature performed major surgery on the 24th to include more rural areas, transforming it back to one that Trump would have won 55.4-43.0 in 2020—and Clinton would have lost by a punishing 59.3-35.4. Needless to say, Van Duyne is looking much safer this year. In one of the nation’s most extreme adjustments, we’ve adjusted her 1.3-point win in 2020 all the way to a 19.1-point win for our purposes, given the 17.7-point improvement in her favor in the circumstances in her district.

These adjustments can also result in a perverse outcome: Some incumbents will appear to have negative margins of victory in 2020. Democratic Rep. Al Lawson is one such congressman. Thanks to Florida Gov. Ron DeSantis’ hyper-aggressive gerrymander, Lawson wound up in a district an astonishing 37.4 points redder than his old one, based on presidential margins. That shrunk his 30.3-point margin in 2020 all the way to -7.3—and no, that’s not a typo.

One other adjustment that we make every time we run the House Vulnerability Index is that when a seat becomes open (whether due to retirement, vacancy, or an incumbent’s loss in a primary), the number in the incumbent strength column reverts to nil. That reflects the fact that House seats are ordinarily at their most vulnerable when there’s no incumbent seeking re-election. Open swing seats are usually the first seats to flip in a wave election because whatever goodwill or name recognition the previous office-holder had accrued no longer applies—voters face a blank slate.

Because the aim of HVI is to produce a rank ordering, we also rank each of its two components, then add them together, with lower numbers indicating greater vulnerability. Georgia’s open 6th, for instance, was the target of an extreme GOP gerrymander that transformed the seat into one Trump would have carried 56.7-41.8 in 2020 and 61.0-34.4 in 2016. As such, Trump’s average 20.7-point win makes this the reddest seat held by a Democrat, earning a “President Rank” of “1.”

Similarly, Al Lawson’s -7.3 House margin gives him a “House Rank” of “1.” That’s not actually the lowest possible, though, under our system. As we noted above, open seats have no rank at all, so they get a House Rank of “0.” (The Presidential Rank, however, can never go below 1 because, of course, every presidential election is conducted in every district.) Combining President Rank and House Rank gives us an “HVI Score,” which we again translate into a ranking as “HVI Rank.”

So let’s take a look at who’s most vulnerable this year, according to the House Vulnerability Index. The full set of data is available as a Google doc, with separate tabs for Republicans and Democrats. There you can also see all of our underlying calculations, but here are the topline numbers for the 30 most at-risk seats held by Democrats. (Note that HVI Ranks can result in ties.)

HVI Rank	CD	Incumbent	President Rank	House Rank	HVI Score
1	GA-06	OPEN	1	0	1
2	TN-05	OPEN	2	0	2
3	FL-02	Al Lawson	4	1	5
4	FL-04	OPEN	6	0	6
5	AZ-02	Tom O'Halleran	5	2	7
6	FL-13	VACANT	8	0	8
7	FL-07	OPEN	9	0	9
8	AK-AL	Mary Peltola	3	7	10
9	WI-03	OPEN	11	0	11
10	MI-10	OPEN	12	0	12
11	AZ-06	OPEN	13	0	13
12	IA-03	Cindy Axne	16	5	21
12	NY-19	OPEN	21	0	21
14	NJ-07	Tom Malinowski	19	3	22
14	PA-07	Susan Wild	18	4	22
16	OH-09	Marcy Kaptur	14	9	23
16	VA-02	Elaine Luria	15	8	23
18	OH-13	OPEN	24	0	24
18	PA-08	Matt Cartwright	10	14	24
20	ME-02	Jared Golden	7	19	26
20	PA-17	OPEN	26	0	26
22	KS-03	Sharice Davids	20	10	30
22	TX-15	OPEN	30	0	30
24	MN-02	Angie Craig	25	6	31
25	MI-07	Elissa Slotkin	17	15	32
26	OR-05	OPEN	33	0	33
27	NY-03	OPEN	35	0	35
28	NH-01	Chris Pappas	23	13	36
29	IL-17	OPEN	37	0	37
30	WA-08	Kim Schrier	29	11	40

As you can see, the open seats really pile up at the top of the list. And this year, between the usual retirements and the added effect of redistricting, Democrats have a lot of them. In fact, given their narrow five-seat majority, the Democrats could lose control of the House simply via open seats.

The first few slots, in fact, are races that no one is really even viewing as competitive. They’re all just about certain to be uneventful Republican pickups, all through the stroke of the gerrymandering pen. Georgia’s 6th in Atlanta’s suburbs, Tennessee’s 5th in Nashville, and northern Florida’s 5th all got transformed into dark-red districts. While Georgia’s Rep. Lucy McBath will be returning to Congress—she defeated fellow Rep. Carolyn Bourdeaux in a member-on-member primary in the adjacent, formerly swingy but now dark-blue 7th District; Tennessee Rep. Jim Cooper decided to retire.

Florida’s 5th, however, is a particularly unusual case, and one we’ve had to double count in our spreadsheet for it to make sense. The old, dark-blue 5th used to stretch from Tallahassee to Jacksonville to create a predominantly Black seat, leaving two dark-red districts on either side: the 2nd in the rural Panhandle and the 4th in the white parts of Jacksonville. Under the new map, Republicans stripped the 5th for parts, deliberately undermining the cause of Black representation to create three somewhat less red but still likely Republican districts.

The old 4th is now predominantly the new 5th, where GOP Rep. John Rutherford is running. The old 5th is now the Jacksonville-centered 4th, a 53-46 Trump seat, but Democratic Rep. Al Lawson chose not to run there but instead to face off against Republican Rep. Neal Dunn in the 2nd, which would have gone 55-44 Trump. That made sense for Lawson, whose Tallahassee base is in the 2nd, but given the district lean, he faces an exceptionally tough climb against Dunn. His decision also leaves the 4th open for a likely pickup by Republican state Sen. Aaron Bean.

That makes Rep. Tom O’Halleran the next most-vulnerable incumbent Democrat. He, too, is at least somewhat of a victim of redistricting, though he would have had a competitive race this year even under the old lines; he used to represent the 1st District in rural northern Arizona, which went 50 Biden-48 Trump.

He still represents the bulk of the same territory, now numbered the 2nd District, but it shifted to 45 Biden-53 Trump. It’s worth noting that this wasn’t due to Republican legislative chicanery: Arizona uses an independent redistricting commission, and to an extent, O’Halleran ended up drawing the shortest straw. And neither side thinks the ultimate outcome is a foregone conclusion: The NRCC spent $1.7 million to help Republican Eli Crane last month, and the DCCC just went up with its first ad to defend O’Halleran.

Florida’s 7th and 13th also appear to be likely losses for Democrats, again thanks to DeSantis’ map. After that, the next most vulnerable incumbent is Rep. Mary Peltola, who picked up Alaska’s at-large seat in a hugely surprising special election this summer. This is a case where the House Vulnerability Index may, however, somewhat overstate her vulnerability. Yes, it’s one of the reddest districts held by a Democrat, and, yes, she won her race narrowly, but the unusual circumstances of her election may well repeat. The election will still be held using the same ranked-choice format against the same two Republican opponents, who are continuing to go hammer-and-tongs at one another rather than seek the support of each other's voters.

A little further down the list is another case where an incumbent might not be quite as vulnerable as she looks, and that’s in Ohio’s 9th district, a Toledo-area district that got quite a bit redder thanks to yet another GOP gerrymander Here, longtime Democratic stalwart Marcy Kaptur faces Republican J.R. Majewski in what had initially looked like a very difficult race. However, the HVI can’t take into account the recent implosion of Majewski’s campaign amidst a grievous stolen valor scandal and the NRCC’s apparent triage of this race. It’s still a difficult district at 51-48 Trump, so even taking this implosion into account, Kaptur isn’t truly out of the woods yet; it’s just more that she’s probably at less risk relative to, say, Elaine Luria or Angela Craig, who have more “normal” opponents.

By contrast, one race a bit lower on the list where the HVI might understate the Democratic vulnerability is in Texas’ 15th, an open seat where incumbent Rep. Vicente Gonzalez is instead running next door in the bluer 34th district. (If you scroll down, you’ll notice Gonzalez is listed at number 96 in his new district.) This is one of those Rio Grande Valley districts that swung hard in the Republican direction between 2016 and 2020, so averaging those out makes the district look bluer than it would if you just took 2020 by itself.

Clinton won the district under the new lines 55-42, but then Donald Trump won the exact same turf 51-48 four years later. If you assume that’s the new norm for this district and not a one-time aberration, this would look more like the 10th-most vulnerable district rather than the 22nd. Along with Florida’s 2nd, this is the one other district we’re double-counting—more on that below.

Of course, sometimes a race that’s nowhere near the top of the HVI list flips because of race-specific circumstances, and one contest that doesn’t make the top 30 but is still worth mentioning is the battle in Rhode Island’s open 2nd. Although this is a pretty bluish district—Biden won here 56-42 in 2020, which is what keeps this at number 39 on the list despite being an open seat—polling suggests the GOP could well flip it. That’s largely a candidate quality issue: The Republican nominee here is former Cranston Mayor Allan Fung, who’s widely known after two turns as the GOP’s nominee for governor in 2014 and 2018, both races where he overperformed the state’s blue lean.

Now let’s take a look at the most vulnerable Republican seats, with the top 30 arrayed below.

HVI Rank	CD	Incumbent	President Rank	House Rank	HVI Score
1	IL-13	OPEN	4	0	4
1	TX-34	Mayra Flores	1	3	4
3	CA-45	Michelle Steel	5	2	7
4	CA-27	Mike Garcia	3	5	8
5	NM-02	Yvette Herrell	8	1	9
5	NY-22	OPEN	9	0	9
7	CA-22	David Valadao	2	8	10
8	MI-03	OPEN	11	0	11
8	OH-01	Steve Chabot	7	4	11
10	NC-13	OPEN	14	0	14
11	NY-01	OPEN	20	0	20
12	AZ-01	David Schweikert	16	6	22
13	CA-03	OPEN	23	0	23
14	FL-27	María Elvira Salazar	6	18	24
14	FL-28	Carlos Giménez	10	14	24
16	IA-01	Mariannette Miller-Meeks	18	7	25
17	WA-03	OPEN	26	0	26
18	NE-02	Don Bacon	13	16	29
19	IA-02	Ashley Hinson	22	12	34
20	NY-02	Andrew Garbarino	21	15	36
21	OH-07	OPEN	37	0	37
22	CA-41	Ken Calvert	19	23	42
23	CA-40	Young Kim	15	29	44
23	NY-11	Nicole Malliotakis	31	13	44
23	WI-01	Bryan Steil	17	27	44
26	NC-11	OPEN	49	0	49
26	NJ-02	Jeff Van Drew	27	22	49
28	PA-01	Brian Fitzpatrick	12	38	50
28	PA-10	Scott Perry	29	21	50
30	MN-01	Brad Finstad	41	10	51
30	TX-23	Tony Gonzales	25	26	51
30	VA-05	Bob Good	34	17	51

There aren’t as likely to be as many Republican losses as Democratic losses this year, which is generally what just happens in a midterm when one party controls both the White House and Congress. While the Democrats’ current lead of about one point on the generic House ballot is a positive development, that’s not really enough to overcome the Republican lean of the national congressional map. That said, there almost certainly will be at least a handful of Democratic pickups.

Democrats’ best shot at a flip is in Illinois’ 13th District, which is in one of very few states where a Democratic-controlled legislature managed the redistricting process. Lawmakers turned the 13th from a previously swingy district into one that would have gone 54-43 for Biden in 2020; Republican Rep. Rodney Davis bailed out and ran in a primary in the nearby much-redder 15th district but lost to fellow GOP incumbent Mary Miller. That leaves the 13th the bluest GOP-held open seat by far.

Two other open seats aren’t quite so friendly but look like reasonable possibilities for pickups. The first of these is New York’s 22nd District, a Syracuse-based seat that used to be the 24th last decade, which got bumped up to 53 Biden-45 Trump in redistricting, prompting Republican Rep. John Katko to retire. The other is Michigan’s 3rd, previously a light-red seat in the Grand Rapids area, which got bumped up to 53-45 Biden in redistricting (now handled by an independent commission), and also saw Republican Rep. Peter Meijer knocked off in a primary by a much more extreme opponent, turning it into an open seat.

Then there’s Texas’ 34th, which we mentioned above and has been represented since June by Republican Mayra Flores, following her victory in a special election prompted by Democrat Filemon Vela’s resignation. While Republicans made Gonzalez’s 15th significantly redder in redistricting this year (prompting Gonzalez’s switch), the 34th correspondingly become much bluer, making it much less likely that Flores will repeat her special election feat.

The old version of the 34th (where the special election was held) went for Biden 52-48, but the new and current version would have given Biden a much wider 57-42 win, with a similar shift in the 2016 numbers. This has the effect, when applying the adjustment for changing district lines, of turning Flores’ 7.5-point special election victory into a 5.6-point loss, thanks to the 13.1-point swing in the district composition. Put another way, Flores wouldn’t have won that special election in the first place if it had been held under the current district lines, all other things being equal. That leaves her very much the most vulnerable Republican incumbent.

Finally, one other thing you might notice is the absence of the Republicans’ biggest and most egregious names from the list above. After all, you might be thinking, “Well, Marjorie Taylor Greene is out there in the news media making a giant spectacle and nuisance of herself every day, and her opponent, Marcus Flowers, is one of the top fundraisers among all Democratic challengers; that race has to be competitive, right?”

No … far, far from it. If you scroll all the way down the list, you’ll see Greene is in the 23rd slot … up from the very bottom, as the 185th-most vulnerable GOP member. That’s all thanks to the friendly confines of rural Georgia’s 14th District, which under the new lines and old, would have delivered massive wins for Trump (an average of about 40 points under the former and 50 under the latter). Greene herself also won her last race by 49 points. No amount of money, no scandal, and no “black swan event” would change the outcome of this race, which should underscore how bad of an investment putting any of your money here would be.