It’s hard to remember that there’s a presidential election happening in 2016, what with the huge amounts of attention generated by the epic fight for control of the U.S. House (especially with the way the Daily Kos recommended list is constantly dominated with pitched battles between supporters of tried-and-true establishmentarian Brad Schneider and progressive upstart Nancy Rotering in the Democratic primary in Illinois’s 10th district). But bear with me: Let’s have one more diary looking at the Democrats’ chances in the lower chamber, and then at some point in the future we can explore the obscure mysteries of the presidential race.
The Democrats currently hold 188 seats in the House; they would need to flip 30 seats in order to obtain a bare majority. That’s about as many seats as they gained in the 2006 wave, so it’s not impossible, given enough momentum. There are two impediments that are greater now than there were in 2006, though: One is a more severely gerrymandered map, following state legislative losses in the 2010 election. A number of previously competitive seats got remade in a way that they’re now slightly out of reach (for instance, from 50-50 to, say, 45-55).
And the other problem is a related one: The decline in ticket-splitting. In the ‘70s and ‘80s, and to a lesser extent in the ‘90s and ‘00s, that 45-55 district could and often would elect a Democrat anyway … usually a moderate, though, which is why even when the Democrats did have a House majority, it was often a pretty uncooperative one. Today, the realignment is almost entirely complete, with House votes very likely to match presidential votes. And when the majority of congressional districts are one that the GOP presidential candidate wins (thanks to gerrymandering, but also the natural tendency of Democrats to distribute themselves inefficiently throughout the country, clustering in cities), that makes it an uphill climb even under the best of circumstances.
The good news, though, is that 2016 is a presidential year (or so I hear) … and the pattern in the last decade has been that the Democratic base is much likelier to show up in full force in those years. General election polls have also usually shown the Democratic candidates with a decent edge in November, which means potential coattails. With that in mind, combined with the fact that there were a few GOPers who were unexpectedly washed in to light-blue districts in 2014 who don’t seem to have much staying power, most analysts are expecting Democratic gains in the House in 2016.
How many gains, though … are we talking 10, or are we talking 30?
Most likely we’re talking more like 10, but there’s no way to predict coattails, and if there were, the general election polling we’re seeing right now is subject to a million unpredictable variables that’ll pop up in the next year. What we can predict with some confidence, though, is what order the seats are likely to fall in—in other words, who falls over, even with just one of those little sneaker waves that get your ankles wet, versus those on higher ground who’d need a huge tsunami in order to lose. And that’s where our House Vulnerability Index comes in.
If you’ve been a regular reader of Daily Kos Elections (and Swing State Project), you may have seen the HVI before; we ran it in 2010 and then 2014 , and both times it worked pretty well. For instance, in 2014, of the 13 Democratic seats that the model deemed most vulnerable, 10 of those seats became actual losses in November (accounting for the majority of all Dem losses). The top 13 were open UT-04, open NC-07, AZ-02, FL-18, AZ-01, open NY-21, TX-23, open ME-02, CA-07, GA-12, NH-01, NY-01, and WV-03. (Of those, we only won FL-18, AZ-01, and CA-07.)
Six other losses were higher up the totem pole: FL-26 at #21, open IA-01 at #22, IL-12 at #23, NV-04 at #31, NY-24 at #32, and IL-10 all the way up at #38. (If you’re remembering "Hey, wait a minute, the Dems only lost 13 seats in 2014,” that's because they also gained three elsewhere.) In most of those latter cases, too, there’s an explanation that the model couldn’t really account for: in FL-26, Joe Garcia’s 2012 margin was inflated by running against David Rivera and his myriad legal problems; in NV-04 and NY-24, the candidates were hampered by abysmal Dem turnout in the gubernatorial races (in Nevada because the Dems only put up a token candidate, in New York because, well, Andrew Cuomo); and in IL-10 because the challenger was the GOP ex-Rep. that Brad Schneider narrowly beat in 2012, in a socially liberal but affluent district that's one of the few places where ticket-splitting is still de rigeur.
Here’s how the model works. It combines two factors: The average lean of the district in the previous two presidential elections, and the incumbent’s margin of victory in the previous House election. This way, it downplays members who are in fairly safe districts but who had a close call last time (possibly because they were running for the first time, and usually an election is closer when it’s an open seat or you defeat an incumbent), and also members who are in swing districts but who have gotten entrenched and usually win by decent margins. Instead, it pinpoints members who are both in swingy districts and had a rough previous election: They’re the ones who, history indicates, are the ones likeliest to lose the next time. We always switch to “0” for margin score in open seat races, since historically that’s when districts are most vulnerable to switching, since you can’t rely on an entrenched incumbent’s goodwill.
So what does the model tell us about 2016? We’ll start with the vulnerable Republicans, since they’re the ones playing most of the defense next year.
DISTRICT |
REP. |
Dist. lean |
‘12 margin |
TOTAL hvi |
FL-10 |
Webster |
2 |
1 |
3 |
VA-04 |
Forbes |
1 |
2 |
3 |
FL-13 |
OPEN |
7 |
0 |
7 |
IA-01 |
Blum |
4 |
6 |
10 |
IL-10 |
Dold |
3 |
7 |
10 |
NY-19 |
OPEN |
12 |
0 |
12 |
NV-04 |
Hardy |
6 |
8 |
14 |
FL-26 |
Curbelo |
11 |
3 |
14 |
NV-03 |
OPEN |
18 |
0 |
18 |
ME-02 |
Poliquin |
8 |
12 |
20 |
PA-08 |
OPEN |
22.5 |
0 |
22.5 |
CO-06 |
Coffman |
13 |
15 |
28 |
NH-01 |
Guinta |
19 |
10 |
29 |
MN-02 |
OPEN |
33 |
0 |
33 |
IL-12 |
Bost |
14 |
20 |
34 |
IA-03 |
Young |
17 |
19 |
36 |
NJ-03 |
MacArthur |
20 |
17 |
37 |
CA-21 |
Valadao |
9 |
32 |
41 |
NY-01 |
Zeldin |
31 |
14 |
45 |
CA-10 |
Denham |
24 |
25 |
49 |
TX-23 |
Hurd |
45 |
5 |
45 |
AZ-02 |
McSally |
47 |
4 |
51 |
FL-06 |
OPEN |
51 |
0 |
51 |
IL-13 |
Davis |
16 |
36 |
52 |
PA-16 |
OPEN |
52 |
0 |
52 |
NY-24 |
Katko |
5 |
48 |
53 |
MI-06 |
Upton |
25 |
31 |
56 |
PA-06 |
Costello |
29 |
27 |
56 |
MI-01 |
OPEN |
59.5 |
0 |
59.5 |
MI-08 |
Bishop |
35 |
26 |
61 |
MI-07 |
Walberg |
39 |
24 |
63 |
NY-21 |
Stefanik |
15 |
59 |
74 |
NJ-05 |
Garrett |
53 |
23 |
76 |
VA-10 |
Comstock |
44 |
34 |
78 |
MI-11 |
Trott |
48 |
30 |
78 |
MI-10 |
OPEN |
78 |
0 |
78 |
WI-07 |
Duffy |
28 |
51 |
79 |
NY-11 |
Donovan |
38 |
42 |
80 |
NJ-02 |
LoBiondo |
10 |
77 |
87 |
WI-06 |
Grothman |
61 |
33 |
94 |
MI-04 |
Moolenaar |
59.5 |
38 |
97.5 |
MI-03 |
Amash |
57 |
46 |
103 |
WA-03 |
Herrera Beutler |
36 |
69 |
105 |
IN-09 |
OPEN |
105 |
0 |
105 |
MN-03 |
Paulsen |
30 |
78 |
108 |
CA-49 |
Issa |
54.5 |
55 |
109.5 |
AR-02 |
Hill |
98 |
13 |
111 |
MT-AL |
Zinke |
83 |
29 |
112 |
NY-23 |
Reed |
40 |
72 |
112 |
You might be surprised to see Dan Webster and Randy Forbes at the top of the list. They aren’t freshmen, or guys who had hard-fought elections the last few times, or moderates with a long track record of sticking it out in swing districts. Possibly you haven’t even heard of them before—they're just your basic Southern social conservative backbenchers. The reason they’re suddenly on top is because what gerrymandering giveth, gerrymandering also taketh away—at least when the courts get involved.
Webster may be familiar to fans (or detractors) of Alan Grayson, though. Webster’s the guy who beat Grayson in the old swingy FL-08 in Orlando in the 2010 wave; population growth in central Florida led to the creation of a new district in 2012, so Webster got a much safer version of his district, now called FL-10, while an entirely new Hispanic-majority district next door was created (FL-09), to which Grayson returned. However, that 2012 Florida map was a pretty half-assed attempt by the state’s Republican legislature at complying with the state’s Fair Districts initiative. Challenges to the map worked their way through the Florida courts, and recently the new map was made official. It blew up the atrocious old FL-05, which packed hundreds of thousands of African Americans in the northern half of the state into one worm-like district that stretched from Jacksonville to Orlando. The Orlando portion of the 5th got added to the 10th, leaving Webster in possession of what’s now a dark-blue district.
That doesn’t mean we’ve seen the last of Webster, though: He's considering moving into next-door FL-11, which is still a dark-red district (at 61 percent Obama in 2012 under the new lines), and now open thanks to the retirement of Rich Nugent (it’s so red that it’s only 63rd on the vulnerability list, despite being an open seat). If he wins the GOP primary there, he can keep on doing what he does. However, the point is an African American Democrat is almost certain to take over the 10th, regardless of whether Webster bails, or stays and fights (it may, in fact, be Corrine Brown, who’s planning to vacate the 5th and move to the 10th).
A similar story is playing out in Virginia, where the courts have ruled that the mostly African American VA-03, which stretches from Richmond to Norfolk, must also be unpacked. That moves many African American votes into the 4th, where Randy Forbes has been comfortably parked for the last decade and a half. (While the Virginia map isn’t truly official, both proposed versions of the new map released by the special master are similar, and they both dramatically remake VA-04.) Again, the new district is blue enough that Forbes will almost certainly be replaced by a Democrat regardless of whether he stays or goes.
And similarly, Forbes does have a bailout option: He’s been cited as a possible primary challenger to Dave Brat in next-door VA-07. Brat, of course, is the tea partier who shocked Eric Cantor in the 2014 GOP primary. The Republican establishment has been casting about for an establishment challenger to Brat, and Forbes would certainly fit the bill, though Henrico Co. Sheriff Mike Wade may already have that position staked out.
Slotted in at third place is open FL-13, another district where the court-ordered redistricting has a big effect. This was already a swingy district, but the court’s move of the African American parts of St. Petersburg into the 13th from FL-14 makes this more of a light-blue district. At 55 percent Obama in 2012 (up from 50 percent under the old lines), that was enough to prompt Republican incumbent David Jolly to seek redder pastures, though conveniently for him, he’s angling for a promotion—the open Senate seat vacated by Marco Rubio. With Charlie Crist (who, if nothing else, is well-known) running as the Democrat in FL-13, this is another likely pickup.
So, we have to travel down to fourth place to find the first two really conventional races. Rod Blum is the most vulnerable incumbent (or at least the most vulnerable one who’s likely to stick around in his district), in eastern Iowa’s IA-01. His victory was one of the more unexpected ones in 2014, facilitated by the open seat vacated by Bruce Braley, as well as a lackluster campaign from would-be Dem replacement Pat Murphy. (Apparently, there can only be one Patrick Murphy in the House at a time.) Blum also has been acting all term like he’s one-and-done, anyway. Rather than trying to moderate his stances in order to hold on to a light-blue district, he voted against John Boehner for speaker (which means the NRCC is doing nothing to help him in 2016), and he’s the only swing-district dweller who’s in the House Freedom Caucus (the rest of whom are in blood-red districts, mostly in the south).
The other person tied for fourth place is Bob Dold, a moderate GOPer who’s a much more skillful (or less down-with-the-ship) politician than Blum. He held IL-10 in Chicago’s northern suburbs in 2010 when it was open, narrowly lost it to Democrat Brad Schneider in 2012, and narrowly won it back from Schneider in 2014. They’ll have their third face-off in 2016, assuming Schneider makes it through the Dem primary.
Here’s how the math works for Blum, for example: The 1st had an average 15.9 percent Democratic presidential margin in 2008 and 2012 (56.2 percent for Obama in 2012, 58.5 percent for Obama in 2008), making this the fourth-bluest Republican-held district, after the aforementioned VA-04, FL-10, and IL-10. And he had the third-narrowest margin of victory in 2014, after Martha McSally in AZ-02 (who won by 0.1 percent) and Will Hurd in TX-23 … though that gets bumped down to sixth, because we’ve added a fudge factor to VA-04, FL-10, and FL-26, where, after adjustments, the incumbents won by negative margins!
That fudge factor means that, for candidates in Florida and Virginia districts that got fixed by the courts, we’ve subtracted the change in presidential margins between the old and new districts, from their 2014 House margins. That’s an imprecise way of reflecting that a lot of the people who voted both for Mitt Romney and a Republican House member are no longer in those districts. It’s possible that they’re skilled enough campaigners that they could generate some ticket-splitting and over-perform their district leans … but we’re not going to assume that.
So, if you’re wondering what the underlying two factors look like when you break them apart, here they are. We’ll start with the district lean, first, showing who’s in the most mismatched districts:
District |
Rep. |
‘12 RESULT |
‘08 RESULT |
AVG. MARGin |
VA-04 |
Forbes |
62.1/38.0 |
61.2/38.1 |
-23.6 |
FL-10 |
Webster |
60.7/38.4 |
60.9/38.5 |
-22.4 |
IL-10 |
Dold |
57.5/41/1 |
63.0/35.9 |
-21.8 |
IA-01 |
Blum |
56.2/42.5 |
58.5/40.4 |
-15.9 |
NY-24 |
Katko |
57.0/41.1 |
56.0/42.0 |
-15.0 |
NV-04 |
Hardy |
54.4/43.7 |
56.4/41.4 |
-12.9 |
FL-13 |
OPEN |
54.6/43.9 |
55.5/43.0 |
-11.6 |
ME-02 |
Poliquin |
52.9/44.4 |
54.6/43.4 |
-9.9 |
CA-21 |
Valadao |
54.6/43.5 |
52.0/46.0 |
-8.6 |
NJ-02 |
LoBiondo |
53.5/45.4 |
53.2/45.5 |
-7.9 |
FL-26 |
Curbelo |
55.4/43.9 |
51.5/47.9 |
-7.6 |
NY-19 |
OPEN |
52.1/45.9 |
53.0/45.0 |
-7.1 |
CO-06 |
Coffman |
51.6/46.5 |
53.6/44.9 |
-6.9 |
IL-12 |
Bost |
49.7/48.2 |
54.7/43.6 |
-6.3 |
NY-21 |
Stefanik |
52.2/46.1 |
52.0/47.0 |
-5.6 |
And now the Republicans who had the closest calls in 2014 (note the asterisks; these, as mentioned before, are where we adjusted the margin to reflect the new district):
DIStrict |
Rep. |
‘14 result |
margin |
FL-10 |
Webster |
68.5/31.5 |
-5.5 * |
VA-04 |
Forbes |
60.2/37.5 |
-2.3 * |
FL-26 |
Curbelo |
51.5/48.5 |
-1.5 * |
AZ-02 |
McSally |
50.04/49.96 |
0.1 |
TX-23 |
Hurd |
49.8/47.7 |
2.1 |
IA-01 |
Blum |
51.1/48.8 |
2.3 |
IL-10 |
Dold |
51.3/48.7 |
2.6 |
NV-04 |
Hardy |
48.5/45.8 |
2.8 |
WV-02 |
Mooney |
47.1/43.9 |
3.2 |
NH-01 |
Guinta |
51.8/48.2 |
3.6 |
UT-04 |
Love |
50.9/45.8 |
5.1 |
ME-02 |
Poliquin |
47.1/41.8 |
5.2 |
AR-02 |
Hill |
51.9/43.6 |
8.3 |
NY-01 |
Zeldin |
54.4/45.6 |
8.9 |
CO-06 |
Coffman |
51.9/43.0 |
8.9 |
Finally, let’s switch over to looking at the most vulnerable Democrats. We’ll use a shorter list here, because there simply aren’t as many plausibly vulnerable Dems. That’s not so much a matter of how awesome they all are, as: a) it’s a presidential year, so turnout should be better and the wind should be at their backs more, and b) most of the precariously placed Dems in 2014 got cleaned out, so there’s not much left to be vulnerable. The last holdout Blue Dogs in rural, heavily evangelical, southern districts (like John Barrow in GA-12 and Nick Rahall in WV-03) finally saw their luck run out in 2014.
District |
rep. |
dist. lean |
‘12 margin |
total hvi |
FL-02 |
Graham |
1 |
1 |
2 |
AZ-01 |
OPEN |
4 |
0 |
4 |
FL-18 |
OPEN |
5 |
0 |
5 |
CA-07 |
Bera |
9 |
3 |
12 |
NE-02 |
Ashford |
3 |
10 |
13 |
FL-09 |
OPEN |
17 |
0 |
17 |
MN-08 |
Nolan |
13 |
4 |
17 |
NY-18 |
Maloney |
10 |
7 |
17 |
CA-24 |
OPEN |
23 |
0 |
23 |
CA-52 |
Peters |
15 |
9 |
24 |
MN-07 |
Peterson |
2 |
23 |
25 |
MN-01 |
Walz |
6 |
22 |
28 |
CA-36 |
Ruiz |
7 |
21 |
28 |
CA-26 |
Brownley |
22 |
8 |
30 |
MD-06 |
Delaney |
26 |
5 |
31 |
TX-15 |
OPEN |
32 |
0 |
32 |
CA-03 |
Garamendi |
19 |
15 |
34 |
CT-05 |
Esty |
16 |
19 |
35 |
NY-03 |
Israel |
11 |
25 |
36 |
NY-04 |
Rice |
20 |
16 |
36 |
AZ-09 |
Sinema |
8 |
33
|
41 |
IA-02 |
Loebsack |
28 |
14 |
42 |
NH-02 |
Kuster |
18 |
27 |
45 |
WA-01 |
DelBene |
21 |
28 |
49 |
CA-31 |
Aguilar |
39 |
11 |
50 |
Freshman Blue Dog Gwen Graham, in FL-02, would be in first place even if nothing had happened to her district: She only narrowly defeated incumbent Rep. Steve Southerland by less than one percent, in 2014 (one of only two incumbent GOPers who lost that year), and it was already a district that Romney won in 2012. However, the same court-ordered re-map that’s gifting us FL-10 and FL-13 had to stick those displaced Republicans somewhere, and many of them ended up in Graham’s remodeled district, moving her from very-endangered-first-place to hopeless-case-first-place.
Remember how we talked about FL-05 losing the African-American portion of Orlando? Well, FL-05 now instead encompasses the African-American populations in Jacksonville and Tallahassee. Taking Tallahassee out of FL-02 takes the already light-red district and moves the 2nd, which now just takes in the rural parts of the Panhandle, into dark-red territory. Graham hasn’t announced her plans for 2014 yet, but they probably don’t involve running for re-election in the 2nd. Possibly, she could run in the new 5th, and try to win the primary by hoping the black vote gets split between multiple candidates from Jacksonville and Tallahassee. Or, she might just quit and get a jump on full-time fundraising for the 2018 Florida gubernatorial race, for which she’s already been tipped as the Dems’ best option.
After Graham come two open seats. Two talented Democrats who proved surprisingly resilient at keeping their light-red seats in 2014 — Ann Kirkpatrick in AZ-01 in rural northern Arizona and Patrick Murphy in FL-18 in Palm Beach County — are seeking promotions to the Senate. Murphy has a good shot at the open seat abandoned by Marco Rubio, while Kirkpatrick’s engaged in a roll of the dice: She wouldn’t have good odds in November against John McCain, but is gambling on him losing his primary and facing a combustible tea partier in the general, instead. Unfortunately, that leaves both their seats open, which will be tough to hold.
So, that takes us up to Ami Bera and Brad Ashford, who are the most vulnerable incumbent Dems. Bera only narrowly won re-election in CA-07, a swingy district in Sacramento’s suburbs, thanks in part to a face-off with ex-Rep. Doug Ose. Ashford won a little more convincingly, beating the hapless Rep. Lee Terry (the other incumbent GOPer to lose in 2014), but his Omaha-based district of NE-02 is also a little more Republican-leaning.
In fact, these two really seem like the only Dems in trouble for now. Everyone lower on the list has drawn so-so opposition, will benefit more from presidential-level turnout, or both. Bera ran a fairly progressive campaign in 2012 when first elected, but labor has sought to make an example of him in particular for backing Trade Promotion Authority and that could hurt both his fundraising and ground game. Meanwhile, Ashford is a Blue Dog who’s an indifferent fundraiser and loath to engage in negative campaigning … but, on the other hand, his affable, low-key moderation may be a good fit for his district.
Again, let’s break down the various factors, starting with the Dems in the most Republican-friendly districts. Notice that, after the thinning of the herd in 2014, only five Dems are left in Romney districts:
district |
rep. |
‘12 result |
‘08 result |
avg. margin |
FL-02 |
Graham |
34.0/64.6 |
35.2/63.4 |
-29.4 |
MN-07 |
Peterson |
44.1/53.9 |
47.1/50.4 |
-6.6 |
NE-02 |
Ashford |
45.7/52.9 |
50.0/48.8 |
-3.0 |
AZ-01 |
Kirkpatrick |
47.9/50.4 |
47.8/51.0 |
-2.9 |
FL-18 |
Murphy |
47.7/51.4 |
51.2/47.9 |
-0.2 |
MN-01 |
Walz |
49.6/48.2 |
50.8/46.7 |
2.8 |
CA-36 |
Ruiz |
50.7/47.5 |
50.0/47.0 |
3.1 |
AZ-09 |
Sinema |
51.1/46.6 |
51.3/47.4 |
4.2 |
CA-07 |
Bera |
50.8/46.8 |
51.0/46.0 |
4.5 |
NY-18 |
Maloney |
51.4/47.1 |
52.0/47.0 |
4.7 |
NY-03 |
Israel |
50.8/48.2 |
54.0/46.0 |
5.3 |
OR-05 |
Schrader |
50.5/47.1 |
53.0/44.2 |
6.1 |
MN-08 |
Nolan |
51.7/46.2 |
53.1/44.5 |
7.1 |
OR-04 |
De Fazio |
51.7/45.0 |
54.2/42.7 |
9.1 |
CA-52 |
Peters |
52.1/45.7 |
55.0/43.0 |
9.2 |
And finally, let’s look at the closest 2014 margins for returning Democratic House members:
District |
rep. |
‘14 result |
margin |
FL-02 |
Graham |
50.4/49.6 |
-22.9 * |
NY-25 |
Slaughter |
50.2/49.8 |
0.5 |
CA-07 |
Bera |
50.4/49.6 |
0.8 |
MN-08 |
Nolan |
48.5/47.1 |
1.4 |
MD-06 |
Delaney |
49.7/48.3 |
1.5 |
CA-16 |
Costa |
50.7/49.3 |
1.5 |
NY-18 |
Maloney |
49.7/47.8 |
1.9 |
CA-26 |
Brownley |
51.3/48.7 |
2.7 |
CA-52 |
Peters |
51.6/48.4 |
3.2 |
NE-02 |
Ashford |
49.0/45.7 |
3.3 |
CA-31 |
Aguilar |
51.7/48.3 |
3.5 |
HI-01 |
Takai |
51.9/48.1 |
3.9 |
CA-09 |
McNerney |
52.4/47.6 |
4.7 |
IA-02 |
Loebsack |
52.5/47.4 |
5.1 |
CA-03 |
Garamendi |
52.7/47.3 |
5.4 |
Interestingly, many of the closest calls were in pretty reliably blue districts, and aren’t likely to present the same problems in 2016. Some you can blame on gubernatorial anti-coattails (like Louise Slaughter in NY-25 and John Delaney in MD-06); many of the California districts (like CA-16 and CA-09, especially) were heavily-Latino districts where turnout lagged. And HI-01 was an open seat where the Republican opponent was the only GOPer capable of making this dark-blue district competitive, ex-Rep. Charles Djou.
All of the underlying data is available in spreadsheet form, in case you're interested in seeing the guts of the model, or just looking to check out a non-competitive district that didn't make any of the tables above.