The AHCA vote likely increased Democrats’ chances of winning the U.S. House in 2018

Alaska and Hawaii became the 49th and 50th states, respectively, in 1959. As a result, 1962 was the first election year to reflect the current U.S. House of Representatives (House) configuration of 435 seats allocated across 50 states[1]. These were also the first House elections since John F. Kennedy won the presidency two years earlier. The rule of thumb is that the party of the president loses House seats in the first midterm (even-numbered years between presidential election years) elections following the first time that president is elected.

In 1962, the Democrats lost five seats, the Republicans won two seats, there was one new Independent, and two seats were eliminated (the previous two Houses had 437 seats in anticipation of Alaska and Hawaii seats and the post-1960 U.S. Census redistricting). To simplify matters, let’s call this a Democratic loss of two House seats.


On May 3, 2017, the House voted 217-213 to pass the American Health Care Act (AHCA). Of the 431 current House members, all 193 Democrats voted against the bill, as did 20 of the 238 Republicans, with one Republican (Dan Hewhouse of Washington’s 4th congressional district [WA-4]) not voting.

The AHCA was the third major attempt to overhaul the U.S. health care system in the last 25 years. On October 27, 1993, President Bill Clinton introduced his universal health care bill (designed in part by a task force led by First Lady Hillary Clinton) to Congress. After months of wrangling, it never came to a vote in either the House or the U.S. Senate (Senate), and the effort to pass the bill ended in late August 1994.

On September 9, 2009, President Barack Obama delivered a nationally-televised address to a joint session of Congress calling for Congress to pass health care reform, with the goal of universal coverage. Initial legislation passed the House on November 7, 2009 and the Senate on December 4, 2009. The Senate bill, the Patient Protection and Affordable Care Act (ACA), was amended by the House to allow the Senate to pass it on a simple majority vote under a process called budget reconciliation[2]. Ultimately the amended ACA passed both houses of Congress and was signed into law by President Obama on March 23, 2010.

Both of these months-long, highly-controversial efforts were undertaken by a newly-elected president in the two years prior to his first midterm elections.

Hold that thought.


As noted earlier, it has been common for a president’s party to lose House seats in that president’s first midterm elections. There have been eight such elections since 1960 (I exclude 1966, since Lyndon Johnson won RE-election to the presidency in 1964, and 1974, since Gerald Ford had not been elected to the presidency):

Figure 1: Net House Seat Losses for a President in First-Term Midterm Elections, 1962-2010

Net House Seat Loss by President’s Party in First Midterm, 1962-2010

Starting in 1962, the president’s party has lost an average of 21.5 House seats (median loss=13.5 seats) in that president’s first midterm elections (Figure 1)[3]. However, the rate of first-midterm House seat loss has accelerated over time. A simple weighting scheme[4] increases the average to 27.4 House seats lost. The simple average over the last five qualifying midterm elections is 28.6 House seats lost (median=26), and over the last three is 36.3 House seats lost (median=54). Finally, a simple linear regression model “predicts” that the average House seat loss for Republicans in 2018 would be 45.9!

Clearly, history is not always predictive. The president’s party lost an average of 13.8 House seats in the four qualifying midterm elections from 1962-1982, yet President George H.W. Bush’s Republicans only lost 8 House seats in 1990, while President Bush was still receiving plaudits for the first Gulf War and the end of the Cold War. Similarly, the 1962-1994 average was 19.5 House seats lost, yet President George W. Bush’s Republicans actually GAINED 8 House seats in 2002, in the first midterm elections following the 9-11 terrorist attacks.

Still, here is where the examples of 1994 and 2010 are instructive. As noted earlier, these were the first midterm elections following a newly-elected president, with majorities in both the House and Senate, attempting to overhaul the nation’s health care system; both overhaul attempts were highly controversial and extended over much of the president’s first two years in office.

Democrats lost an average of 58.5 House seats over those two elections, handing control of the House to Republicans on both occasions (in 1994, for the first time in 40 years!)[5].

Could the same thing happen to Republicans in 2018, following the AHCA vote in the House, and what appears likely to be as long a process as the previous two attempts to change the U.S. health care system? Analysts I respect are already speculating about the possibility, but just bear with me while I offer my own analysis.


In the 2016 midterm elections, Republicans won 241 House seats and Democrats won 194 House seats, meaning that Democrats would have to gain 24 House seats to win a majority of House seats in 2018. I have already written about the challenge to such a takeover presented by the large majority of very safe seats held by both political parties. In that earlier post, I found only 48 seats (28 won by Republicans[6]) decided by less than 10 percentage points and/or the winning candidate won less than 55% of the total vote (“close”); there were only 23 districts won by a Republican House candidate AND by Democratic presidential nominee Hillary Clinton. In total, there are 41 House districts in either/both of these two “vulnerable” categories.

The easiest thing for Democrats would be to win 24 of these 41 House seats in 2018 (while losing none of their seats). And, in an intriguing sign of things to come, 15-term Republican Congresswoman Ileana Ros-Lehtinen (FL-27) announced that she would not seek reelection in 2018. In 2016, Ms. Clinton won her House district, and she won her House seat by “only” 9.8 percentage points. Democrats are now favored to win this seat in 2018.

There are other warning signs for Republicans as well. After Republican Mike Pompeo won his House seat (KS-4) in 2016 by 31.1 percentage points, Republican Ron Estes won a special election for that seat on April 11, 2017 by just 6.8 percentage points, a pro-Democratic shift of 24.3 percentage points. Republican Tom Price won his House seat (GA-6) in 2016 by 23.4 percentage points, but Democratic candidate Jon Ossoff is essentially tied with Republican Karen Handel in advance of the June 20 runoff in the special election to fill that seat, a shift of 23.4 percentage points. A total of 93 Republicans won their House seats in 2016 by less than 23 percentage points; if Democrats won only one-third of these Republican seats in 2018, that would still be seven more than they need to recapture the House.


It is remarkable political irony that the ACA is now more popular than ever. On April 1, 2010, just after the ACA was signed into law, 42.7% supported it and 50.0% opposed it. Seven years later, those numbers had reversed: 48.4% support and 41.9% opposition, a 13.8 percentage point marginal shift.

At the same time, the AHCA has been less popular than the ACA ever was. The public disapproves of Congress’ performance (62.9% disapprove, 17.2% approve) and of the way President Donald Trump has handled health care (51.3% disapprove, 37.7% approve).

Still, 217 of 238 House Republicans in a historically unpopular body, acting at the behest of an unpopular president, voted in favor of an unpopular bill whose purpose was to “repeal and replace” an increasingly-popular law.

This is exactly what the creation of very safe seats does to democracy—it creates a perverse incentive to be more afraid of primary challenges from party extremes than of the broader pool of general election voters.

Table 1: Distribution of House Votes for the AHCA on May 4, 2017 By Vulnerability of House Member

  Close, Clinton Won Close, Trump Won Not Close, Clinton Won Not Close, Trump Won TOTAL
No 5 1 4 10 20[7]
Yes 5 17 9 186 217
TOTAL 10 18 13 196 237
% No 50.0% 5.6% 30.8% 5.1% 8.4%
Ratio 9.80 1.09 6.03

Overall, according to Table 1, only 8.4% of Republican House members (excluding Newhouse’s non-vote) voted no on the AHCA, as did an even smaller 5.1% of the 196 Republicans from the safest districts (2016 House vote margin>10 percentage points and/or the winning percentage>55%[8]) AND Trump won the district). Surprisingly, only 1 Republican House member (5.6%) from a close district that Trump won (Brian Fitzpatrick [PA-8]) voted no on the AHCA, while the other 17 voted yes, implying they are more scared of Trump voters in those districts than of all other voters.

On the other hand, four Republican House members (30.8%) from districts that were not close, but which Clinton won—John Katko (NY-24), Ryan A. Costello (PA-6), Patrick Meehan (PA-7), Dave Reichert (WA-8)—voted no, while nine voted yes. And, of the 10 Republican House members whose 2016 election was close AND whose district was won by Clinton, half voted no (Mike Coffman [CO-6], Ros-Lehtinen, Leonard Lance [NJ-7], Will Hurd [TX-23], Barbara Comstock [VA-10]) and half voted yes.

Thus, compared to “safe” Republican House members, Republican House members from the least safe (close, Clinton) districts were 9.8 times more likely to vote no on the AHCA, while Republican House members from the not-close districts Clinton won were 6.0 times more likely to vote no.

Simply put, Republican House members from districts won by Hillary Clinton in 2016 (regardless of how “close” their own 2016 election was) were 7.7 times more likely to vote no on the AHCA than Republican House members from districts won by Donald Trump.

But that still leaves 31 Republican House members from districts which were close in 2016 and/or Clinton won in 2016 who voted YES on the AHCA:

Close AND Clinton won (n=5)

Jeff Denham              CA-10

Steve Knight             CA-25

Darrell Issa                CA-49

Carlos Curbelo         FL-26

Kevin Yoder               KS-3

Close and Trump won (n=17)

Martha Robey           AL-2

Don Young                AK-At large

Scott Tipton               CO-3

Brian Mast                 FL-18

Mike Bost                  IL-12

Trey Hollingsworth IN-9

Rod Blum                  IA-1

David Young             IA-3

Bruce Poliquin         ME-2

Jack Bergman            MI-1

Don Bacon                 NE-2

John J. Faso                NY-19

Claudia Tenney        NY-22

Tom Reed                  NY-23

Lloyd K. Smucker    PA-16

Mia Love                   UT-4

Not Close and Clinton won (n=9)

Martha E. McSalley              AZ-2

David Valadao                     CA-21

Ed Royce                               CA-39

Mimi Walters                        CA-45

Dana Rohrabacher               CA-48

Peter Roskam                       IL-6

Erik Paulsen                         MN-3

Tom Culberson                    TX-7

Pete Sessions                        TX-32

Put it this way: if Ossoff wins the runoff on June 20 and a Democrat wins Ros-Lehtinen’s seat, the Democrats would need to win just 22 remaining seats to recapture the House (assuming they lost none of their own 20 close seats). There are 31 seats where voters can truly make an impact by continuing to hold their Republican House members accountable for their AHCA “yes” vote (in clear opposition to the overall political “lean” of their constituents).

And that does not count 9 other Republican House members who voted no on the ACHA but are still potentially vulnerable.

I am not predicting that the Democrats will recapture the House in 2018. I am only saying that a VERY strong opportunity to do so, based on accelerating historical trends, the closeness of 41 (of 241) seats, and the demonstrated impact of tinkering with national health care, is there, IF the Democrats can successfully capitalize on all of these trends.

Until next time…

[1] There had been a total of 435 House seats since 1912, after Oklahoma (1907), Arizona and New Mexico (both 1912) became the 46th through 48th states.

[2] This became necessary after the Democrats lost their filibuster-proof 60-seat Senate majority following the death of Massachusetts Senator Edward Kennedy on August 25, 2009 and the eventual victory of Republican Scott Brown (January 19, 2010) to replace him.

[3] Using the 1966 and 1974 data—10 first midterm election for any president—increases the average net House seat loss to 26.8.

[4] Where the 1962 value is weighted “1,” the 1970 value is weighted “2,” and so forth up to the 2010 value weighted “8.” Adding 1966 and 1974 increases the weighted average to 34.5 net House seats lost.

[5] Let me be very clear. I am NOT saying that the pursuit of national healthcare reform legislation was the sole reason for the Democrat’s very high House seat loss in 1994 and 2010. I would argue that, in 1994, the House Post Office scandal and the omnibus budget bill of 1993 (with its large tax increases) also played a huge role, as did backlash to the first black president and cap-and-trade legislation (which passed the House but died in the Senate) in 2010.

[6] This is actually a correction to that post, as I had neglected to count one close Republican House seat.

[7] Overall, 13 no votes were cast by Republicans from states Clinton won in 2016 (CO, NJ, NY, PA, VA, WA), six were cast by Republicans from states Clinton lost by less than 10 percentage points (AZ, FL, NC, OH, TX) and the other was cast by Thomas Massie of the 4th CD of Kentucky, a state Clinton lost by 29.8 percentage points. Go figure.

[8] This represents 81.3% of all Republican House seats won in 2016.

7 thoughts on “The AHCA vote likely increased Democrats’ chances of winning the U.S. House in 2018

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s