Why did the R’s lose house seats this year?

[GAinDC]

You’d think they’d have made gains considering how poorly the Dems did in the Senate and presidential races? They certainly had some pickup opportunities. Why did they come up short?

[Tokugawa Sexgod Ieyasu]

The Democrats didn't really do that badly in the Senate. Or, rather, they did do badly, but mostly because they did badly at the top of the ticket; all the defeated Democratic incumbents (even Casey, who shat the bed in his own right) outperformed Harris, and of course there were four Trump/Senate Democrat states and no Harris/Senate Republican ones. (This is the main empirical argument for Harris having run a bad campaign; the main empirical argument for her having run a good campaign in a bad environment (which is my own position) is the fact that every swing state trended left.) The Senate is a matter of terrible geography at this point, and that's not a problem at the House level. So a bad Democratic year presidentially that was a wash downballot manifested more positively for Democrats in the the chamber where they don't have a geography/malapportionment problem.

[Smash255]

Democrats popularity in 2024 probably wasn't that much different than it was in 2022 so the overall House results were pretty similar.  In some cases they were probably in worse shape in 2022 than 2024.  NY for example, despite how much worse Democrats did than usual, and how strong Trump did, was still better for them than 2022 was, hence wining back seats they lost in 2022

[Compuzled_One]

You’d think they’d have made gains considering how poorly the Dems did in the Senate and presidential races? They certainly had some pickup opportunities. Why did they come up short?

I mean, the party that gained seats in the midterms has usually lost them in the election following that-this has been the case since 2010-2012. They also went down in the margin by 0.02%, so that didn't help (and I wouldn't discount the possibility the shave reveals they got less votes than the 2022 one had all seats been contested). Also, while they did skew the map with NC, Dems also got a better map in NY, and the SC made states add majority-minority seats.

[ottermax]

A very odd combination of factors: gerrymandering, non-white voter turnout, and geographic differentials in swing.

1. Gerrymandering
Most states with Republican governments have gerrymandered to maximize how many seats they can get almost to their detriment when they have a favorable environment. The two biggest states for them - Texas and Florida - have zero swing seats so they cannot take advantage of national swings when they occur. On the flip side the biggest Democratic states - California and New York - have a less partisan map which means you are starting with more swingy districts in the first place. The bulk of the changed districts in both 2022 and 2024 were in CA and NY for this reason primarily. Other states where we see less gerrymandered swing districts include WA, AZ, CO, MI, and PA. That being said the NC redraw really blunted Democratic gains in 2024.

2. Non-White Turnout
Many of the seats Democrats gained in CA and to a lesser extent in NY (Long Island and Lower Hudson Valley) depend heavily on Latinos and other non-White groups to vote. Midterm years have historically terrible turnout for these groups, and we saw Republicans hold onto or gains these seats in 2022. Even with the swings towards Republicans by Latinos and Asians the sheer number of increased votes when starting from a 60/40 split baseline, and combined with the fact that many of these non-White voters dropped off after the presidential ticket actually helped Democrats in lower ballot races. This also saved seats in NM, NV, TX, and NJ.

3. Geographic Swing Differentials
Trump made his biggest gains in urban and Latino or Asian areas. These areas started from a baseline of 70/30 support for Democrats so even with a pretty dramatic shift towards something like 57/43 this is still a clear majority for Democrats. Even at the state legislature level Republicans failed to gain many seats in Latino areas. Harris basically held steady in specifically swingy areas - suburban, more White areas - and this helped keep those vulnerable districts from voting Republican.

One danger zone for Democrats is expectations in 2026. All of these reasons that helped them in 2024 will actually make them vulnerable to some seat losses in 2026 just by virtue of being a midterm, and will limit potential gains. Some seats like Gray's in CA-13 are potentially DOA for 2026 if turnout drops, and others like MGP in WA-3 are at major risk due to their more Republican nature. And finally even if Democrats regain ground in states like Florida or Texas or even a state like NC that is more promising - the gerrymandering limits potential seats for them to gain unless there is a landslide scenario. In particular the number of Republican seats that were won by less than 10% is remarkably low for 2024 meaning that there will be very limited options for targets for Democrats. However all you need is to win 218 so it might not matter if Democrats win 220 vs 240 seats in the end.

[Vosem]

The simplest answer is that the generic ballot swung very slightly leftward from 2022 (from R+2.7 to R+2.6). Republicans actually "won" the redistricting cycle (on notional 2022 results, Democrats gained a single seat in AL/LA/NY, but Republicans gained 4 in NC), but that's not a significant enough gain to matter much. The seat which decided House control actually swung slightly right (IA-3, Nunn v. Axne, was R+0.7 in 2022*; PA-7, Mackenzie v. Wild, was R+1.0 in 2024).

The result was just basically the same. It's honestly even hard to say that Republicans campaigned much worse (in the decisive seat they won the House by more, after all); the perturbations were absolutely tiny. If anything, I think you could say Republicans overperformed 2022 by a little because they won most of the very closest races; among seats won by less than 0.7%**, in 2022 Republicans beat Democrats 5-1, but in 2024 Democrats beat Republicans 4-1. This kind of thing is dumb luck more than anything else.

*This is still actually true if you use notional results based on 2024 maps; in that case the decisive seat for 2022 is AZ-1, Schweikert v. Hodge, and still R+0.9, so less than the 2024 decisive seat.
**This is a bit of an arbitrary cutoff to prove my point.

[Spectator]

The simplest answer is that the generic ballot swung very slightly leftward from 2022 (from R+2.7 to R+2.6). Republicans actually "won" the redistricting cycle (on notional 2022 results, Democrats gained a single seat in AL/LA/NY, but Republicans gained 4 in NC), but that's not a significant enough gain to matter much. The seat which decided House control actually swung slightly right (IA-3, Nunn v. Axne, was R+0.7 in 2022*; PA-7, Mackenzie v. Wild, was R+1.0 in 2024).

The result was just basically the same. It's honestly even hard to say that Republicans campaigned much worse (in the decisive seat they won the House by more, after all); the perturbations were absolutely tiny. If anything, I think you could say Republicans overperformed 2022 by a little because they won most of the very closest races; among seats won by less than 0.7%**, in 2022 Republicans beat Democrats 5-1, but in 2024 Democrats beat Republicans 4-1. This kind of thing is dumb luck more than anything else.

*This is still actually true if you use notional results based on 2024 maps; in that case the decisive seat for 2022 is AZ-1, Schweikert v. Hodge, and still R+0.9, so less than the 2024 decisive seat.
**This is a bit of an arbitrary cutoff to prove my point.

Rs gained 3 in NC since Davis held on. Further, I don’t know in what world 2022 was a Republican overperformance when they were expected to win those close seats you’re talking about by much more. Engel and Hodge were essentially left for dead by the DCCC in 2022, but Lake and Masters nearly torpedoed Ciscomani and Schweikert

[Vosem]

The simplest answer is that the generic ballot swung very slightly leftward from 2022 (from R+2.7 to R+2.6). Republicans actually "won" the redistricting cycle (on notional 2022 results, Democrats gained a single seat in AL/LA/NY, but Republicans gained 4 in NC), but that's not a significant enough gain to matter much. The seat which decided House control actually swung slightly right (IA-3, Nunn v. Axne, was R+0.7 in 2022*; PA-7, Mackenzie v. Wild, was R+1.0 in 2024).

The result was just basically the same. It's honestly even hard to say that Republicans campaigned much worse (in the decisive seat they won the House by more, after all); the perturbations were absolutely tiny. If anything, I think you could say Republicans overperformed 2022 by a little because they won most of the very closest races; among seats won by less than 0.7%**, in 2022 Republicans beat Democrats 5-1, but in 2024 Democrats beat Republicans 4-1. This kind of thing is dumb luck more than anything else.

*This is still actually true if you use notional results based on 2024 maps; in that case the decisive seat for 2022 is AZ-1, Schweikert v. Hodge, and still R+0.9, so less than the 2024 decisive seat.
**This is a bit of an arbitrary cutoff to prove my point.

Rs gained 3 in NC since Davis held on. Further, I don’t know in what world 2022 was a Republican overperformance when they were expected to win those close seats you’re talking about by much more. Engel and Hodge were essentially left for dead by the DCCC in 2022, but Lake and Masters nearly torpedoed Ciscomani and Schweikert

I believe, under the 2024 maps, Davis would've lost in 2022 (even though it was a Clinton/Biden seat), so him winning reelection in 2024 was a notional Democratic gain. (Similarly, Suozzi winning reelection was a notional Democratic gain, but John Mannion picking up the Syracuse seat, along with Democrats winning the two new Southern black seats, along with Republicans winning the three new North Carolina vote sinks, were all notional holds; if those seats had existed in 2022, they would've voted the same way they did in 2024).

2024 was an overperformance not relative to expectations; just relative to running that national environment a bunch of times. Republicans winning the closest seats in 2022 5-1 is luck, not skill; it suggests that if you run that national environment a bunch of times most of the time their House performance would've been very slightly worse. Obviously, relative to expectations, they did quite poorly. (Similarly, Democrats winning the closest seats in 2024 4-1 is luck, not skill.)

[Spectator]

The simplest answer is that the generic ballot swung very slightly leftward from 2022 (from R+2.7 to R+2.6). Republicans actually "won" the redistricting cycle (on notional 2022 results, Democrats gained a single seat in AL/LA/NY, but Republicans gained 4 in NC), but that's not a significant enough gain to matter much. The seat which decided House control actually swung slightly right (IA-3, Nunn v. Axne, was R+0.7 in 2022*; PA-7, Mackenzie v. Wild, was R+1.0 in 2024).

The result was just basically the same. It's honestly even hard to say that Republicans campaigned much worse (in the decisive seat they won the House by more, after all); the perturbations were absolutely tiny. If anything, I think you could say Republicans overperformed 2022 by a little because they won most of the very closest races; among seats won by less than 0.7%**, in 2022 Republicans beat Democrats 5-1, but in 2024 Democrats beat Republicans 4-1. This kind of thing is dumb luck more than anything else.

*This is still actually true if you use notional results based on 2024 maps; in that case the decisive seat for 2022 is AZ-1, Schweikert v. Hodge, and still R+0.9, so less than the 2024 decisive seat.
**This is a bit of an arbitrary cutoff to prove my point.

Rs gained 3 in NC since Davis held on. Further, I don’t know in what world 2022 was a Republican overperformance when they were expected to win those close seats you’re talking about by much more. Engel and Hodge were essentially left for dead by the DCCC in 2022, but Lake and Masters nearly torpedoed Ciscomani and Schweikert

I believe, under the 2024 maps, Davis would've lost in 2022 (even though it was a Clinton/Biden seat), so him winning reelection in 2024 was a notional Democratic gain. (Similarly, Suozzi winning reelection was a notional Democratic gain, but John Mannion picking up the Syracuse seat, along with Democrats winning the two new Southern black seats, along with Republicans winning the three new North Carolina vote sinks, were all notional holds; if those seats had existed in 2022, they would've voted the same way they did in 2024).

2024 was an overperformance not relative to expectations; just relative to running that national environment a bunch of times. Republicans winning the closest seats in 2022 5-1 is luck, not skill; it suggests that if you run that national environment a bunch of times most of the time their House performance would've been very slightly worse. Obviously, relative to expectations, they did quite poorly. (Similarly, Democrats winning the closest seats in 2024 4-1 is luck, not skill.)

You were not talking about “notional gains” in your original post, you were talking about real gains from redistricting. It is okay to admit you’re sometimes wrong. If we’re gonna pretend the original post was about notional gains, then sure. Davis probably would not have won under the new lines in 2022.

[Vosem]

The simplest answer is that the generic ballot swung very slightly leftward from 2022 (from R+2.7 to R+2.6). Republicans actually "won" the redistricting cycle (on notional 2022 results, Democrats gained a single seat in AL/LA/NY, but Republicans gained 4 in NC), but that's not a significant enough gain to matter much. The seat which decided House control actually swung slightly right (IA-3, Nunn v. Axne, was R+0.7 in 2022*; PA-7, Mackenzie v. Wild, was R+1.0 in 2024).

The result was just basically the same. It's honestly even hard to say that Republicans campaigned much worse (in the decisive seat they won the House by more, after all); the perturbations were absolutely tiny. If anything, I think you could say Republicans overperformed 2022 by a little because they won most of the very closest races; among seats won by less than 0.7%**, in 2022 Republicans beat Democrats 5-1, but in 2024 Democrats beat Republicans 4-1. This kind of thing is dumb luck more than anything else.

*This is still actually true if you use notional results based on 2024 maps; in that case the decisive seat for 2022 is AZ-1, Schweikert v. Hodge, and still R+0.9, so less than the 2024 decisive seat.
**This is a bit of an arbitrary cutoff to prove my point.

Rs gained 3 in NC since Davis held on. Further, I don’t know in what world 2022 was a Republican overperformance when they were expected to win those close seats you’re talking about by much more. Engel and Hodge were essentially left for dead by the DCCC in 2022, but Lake and Masters nearly torpedoed Ciscomani and Schweikert

I believe, under the 2024 maps, Davis would've lost in 2022 (even though it was a Clinton/Biden seat), so him winning reelection in 2024 was a notional Democratic gain. (Similarly, Suozzi winning reelection was a notional Democratic gain, but John Mannion picking up the Syracuse seat, along with Democrats winning the two new Southern black seats, along with Republicans winning the three new North Carolina vote sinks, were all notional holds; if those seats had existed in 2022, they would've voted the same way they did in 2024).

2024 was an overperformance not relative to expectations; just relative to running that national environment a bunch of times. Republicans winning the closest seats in 2022 5-1 is luck, not skill; it suggests that if you run that national environment a bunch of times most of the time their House performance would've been very slightly worse. Obviously, relative to expectations, they did quite poorly. (Similarly, Democrats winning the closest seats in 2024 4-1 is luck, not skill.)

You were not talking about “notional gains” in your original post, you were talking about real gains from redistricting. It is okay to admit you’re sometimes wrong. If we’re gonna pretend the original post was about notional gains, then sure. Davis probably would not have won under the new lines in 2022.

All gains from redistricting are notional at first! Notionally, Republicans gained a net of a single seat in redistricting between the 2022 and 2024 cycles. This includes NC-1, the Don Davis seat. (Actually, they lost a single seat: the NY-3 special election). Thus, NC-1 was a notional Democratic gain in 2024, since it is being compared to notional 2022 results; so is NY-3, even though both were actually Democratic holds where the Democratic victories were achieved by incumbents.

(In 2022, throwing out Mary Peltola (D-AK) -- special election winners who flip seats are always notional gainers -- there were five notional gains, where incumbents who would've lost their races in 2020 had 2022 maps been used, but who still won reelection: Susan Wild (D-PA), Mike Garcia (R-CA), Michelle Steel (R-CA), David Valadao (R-CA), and Mariannette Miller-Meeks (R-IA). None of these were actual gains, but all of these were notional gains).