Aggregate Two-Party House of Representatives Vote Was 51.36% Republican, Yielding 50.57% of House Seats

The 51.36% fraction is of the voters who voted either Democrat or Republican.

This suggests that, whatever one thinks of gerrymandering and of geographic representation (as opposed to proportional representation), they didn't seem to have a particularly distortive effect on this year's race. (For the aggregate House vote totals, I'm looking at the Cook Political Report totals.)

As it happens, I don't much like gerrymandering, though I'm not sure what the optimal solution to it would be. (I'm not sure what I think on balance of geographic vs. proportional representation, if we were redesigning our political system from scratch.) I also appreciate that it's possible that, under some fair systems of district drawing, the 51.36% Republican vote would cash out into a minority in the House, while under other fair systems it would cash out into an even bigger majority. But I think these numbers should put into proportion arguments that Republican control of the House is the fault of "unfair maps."

Thanks to Richard Winger (Ballot Access News) for pointing this out. (His aggregate numbers are slightly different from the Cook Political Report's, but only slightly.)

Comments

[Martinned2]

This suggests that, whatever one thinks of gerrymandering and of geographic representation (as opposed to proportional representation), they didn't seem to have a particularly distortive effect on this year's race.

That assumes that, without gerrymandering etc., turnout and voting would have been the same. And I see no basis for that assumption, even leaving to one side the possibility of PR (which would create a massive push away from a two-party system).

[Brett Bellmore]

Which is precisely why there's about zero chance of the two parties permitting PR to be implemented; It's currently illegal for federal elections.

I agree that turnout and voting would likely be somewhat different if there were no gerrymandering, but it's hard to say how different. It probably locally discourages voting by the political minority, but since both parties do it, it does likely cancel out.

[FivebySixThree]

the courts would not permit it either as that voting system would possibly reduce representation among certain ethnic groups.

[JasonT20]

Which is precisely why there's about zero chance of the two parties permitting PR to be implemented...

This makes it sound like it is the party elite alone that would stand in its way. Which is also an assumption with no basis in fact. The truth is probably that partisan voters like gerrymandering when it is their party benefiting from it. They may even voice great displeasure with it, in theory, when asked a question about it from pollsters. But when the rubber hits the road and their party's control of their purple state's legislature is at stake, they oppose any changes.

[Don Nico]

In any case it is far more representative than the last UK election.

[Brett Bellmore]

According to studies I've seen comparing actual electoral outcomes to simulated gerrymandering free outcomes based on individual precinct numbers, nation-wide gerrymandering nets out at practically no partisan impact. But that's because most gerrymandering is for incumbent protection, not party advantage, and both parties engage in gerrymandering, so it tends to cancel out nationally, even if it's significant in certain states.

Of course, that's using a politically neutral definition of 'gerrymandering', which a lot of left-wing gerrymandering activists dislike.

[JasonT20]

But that's because most gerrymandering is for incumbent protection, not party advantage, and both parties engage in gerrymandering, so it tends to cancel out nationally, even if it's significant in certain states.

Gerrymandering (however defined) is no more acceptable if it tends to "cancel out nationally" than if it didn't. It shouldn't exist at all, and the only motivation to accept it is if one's preferred party is benefiting from it, even if only at the state level.

[Brett Bellmore]

This was a comment on why the number of seats so closely reflects the vote totals, mot an expression of approval.

[JasonT20]

Okay, so you'll be as eager to curb gerrymandering as anyone else, despite it seeming to cancel out nationally? Or did you have some other point to make with the claims about this that I didn't figure out?

[Brett Bellmore]

I said it cancels out nationally, but still has in-state effects, and that most of it is incumbent protection. The latter factors are reason enough to oppose it, even if the national effect is minimal.

[Rossami]

The problem with "it shouldn't exist at all" is that it assumes that any other state of being is possible. I don't like the idea of gerrymandering but the more I've thought about it over the years, the more I convince myself that the alternatives are, if not worse, certainly no better.

[MatthewSlyfield]

Another impact people don't consider to "it shouldn't exist at all" is that without it, there would be fewer if not no majority black districts in most states.

[JasonT20]

A couple of others have made arguments about "majority minority" districts, as if that somehow relates to partisan gerrymandering. I would say that even if Black voters in 1965, when the VRA was passed, were evenly split in their support for the Democratic Party and the Republican Party, it would still have been necessary to have majority Black districts in order for them to truly be represented in legislatures. That is because they were deliberately spread out so that all districts would have large white majorities. That way, candidates of all parties would win elections only by appealing only to white voters. Any candidate that made a point of listening to and advocating for the interests of Black voters was guaranteed to lose.

Racial gerrymandering won't be unnecessary until the interests of voters has no relation to their race. Put simply, the law can be fully color-blind when we have a color-blind society.

[Brett Bellmore]

"I would say that even if Black voters in 1965, when the VRA was passed, were evenly split in their support for the Democratic Party and the Republican Party, it would still have been necessary to have majority Black districts in order for them to truly be represented in legislatures."

That makes no sense at all; If, indeed, blacks were evenly split between parties, that would be evidence that they had no distinctive interest in need of representation. Really, this is just racial essentialism, a racist notion that people, (Or maybe just some people?) can only really be properly represented by members of their own race. Except whites, of course: You'd never follow the reasoning in THAT direction....

[Savagely Average]

I’m no expert but that’s consistent with my understanding as well. The substantive objection to gerrymandering is that it undermines accountability, not that it systemically favors one side or another.

Personally I’m skeptical of efforts to curb gerrymandering. It’s likely always going to happen, whether districts are drawn geographically or through other means. Similar to campaign finance reform, efforts to eliminate gerrymandering will probably just create distortions without compensating benefits.

[JasonT20]

Some states have made efforts to curb gerrymandering that might be doing something about it. Usually its bipartisan commissions drawing the maps or something like that. I haven't lived in a state that does this, so I haven't looked into the details. My state, Florida passed a state constitutional amendment years ago that was supposed to make it against the state constitution to engage in partisan gerrymandering. But given the 25 year lock on the governor's mansion that the Florida GOP has had, the state supreme court is a rubber stamp for whatever Ron DeSantis wants. (He appointed the majority of its Justices. I think Rick Scott appointed the others.)

[Lee Moore]

As you have spotted, any such "bipartisan" system gets captured. Wherever you lock up the honey, the bear will find it.

[JasonT20]

As you have spotted, any such "bipartisan" system gets captured.

I don't see where I spotted this. Florida doesn't have a bipartisan process to draw maps at all. It is still completely controlled by the legislature, with the governor able to veto the maps.

[Rossami]

Those "bipartisan commissions" just move the problem instead of fixing it. Human nature will not allow them to remain "bipartisan" for more than one (and arguably not even for one) election cycle.

[tkamenick]

"Bipartisan" commissions also make it that much easier to suppress any chance of a third party arising.

[JasonT20]

We have a two party system because of the combination of FPTP and single-member districts, not because of how maps are drawn. There is no incentive for a voter to go with a minor party if they want to be sure that their vote won't "spoil" the election and result in their least preferred candidate winning. Minor parties are always going to be the realm of the contrarians and protest votes as long as we structure elections this way.

Proportional representation isn't a cure-all for our politics by any stretch, but it is the only way to get more than two competitive parties.

[Brett Bellmore]

Sure, FPTP and single member districts makes life hard for 3rd parties, but the Whigs are still history. If a major party screws up badly enough under FPTP, they can still end up replaced.

Which is why, when 3rd parties started making a resurgence back in the 70's and 80's, the incumbent parties reacted by all sorts of institutional changes, such as campaign finance reform, or the Bipartisan Debate Commission, to make sure life went from hard to impossible.

That's a source of a lot of the dysfunction in our politics these days: The major parties know that it is now impossible for them to screw up so badly they get supplanted, and they act like it.

[Lee Moore]

I have mentioned before my patent scheme for an algorithmic drawing of districts that pays no attention whatesoever to party advantage. (Which is not to say that it would ever be adopted.)

Pure version :

1. do the census to determine how mant Reps each state gets
2. draw the districts to minimize the aggregate of district boundaries in the state - so long as each district has the same number of inhabitants (or registered voters, whatev.)

Practical version :

1. do the census etc
2. permit, say, any member of the state legislature to propose district boundaries etc, and pick the one with the smallest aggregate boundaries etc

The practical advantage of the practical version is that you don't get endless court cases with new proposals for even shorter boundaries. The choice is limited to a hundred or two max. In practice probably no more than a dozen. You just get court cases on whether the chosen plan does indeed have shorter boundaries than any other competitor, and does have equal population in each district.

Obviously this doesn't pay any lip service to local communities, or racial groupings, or tradition, or county lines etc. Which is the point. It pays attention only to the measuring tape (or modern equivalent.)

And since in most cases many of the bits of lines will be the same as in the other possible plans it isn't even necessary to measure them.

In x + y + z + 37 < x + y + z +41, it doesn't matter what x, y and z are.

The main practical difficulty would be in finding some lawyers and judges who are comfortable with arithmetic.

For ties, we use a coin toss. Though as it's State Reps, single combat to the death would be more fun.

[BadLib]

This is along the lines with what I've proposed as well.

However, since the determination of "optimal" is clearly defined and the "score" can be easily determined by a program given the maps and detailed census data (the same information available to those proposing the maps), any resident of the state should be able to propose one map and if theirs is "the most optimal", it's the map.

If there is a tie, random selection would resolve the outcome. Although to reduce the effect of 1000 colluding residents submitting identical (or effectively identical) maps to increase the odds of that map being selected in a random drawing, perhaps some algorithm could be developed to determine the "distance" between two maps and group substantially similar maps together in the first round and each group (including groups of one) get an equal chance.

Proposed maps could be made public in advance and if someone comes up with a "better" map than all existing proposals, the contest opens again with that (and subsequently selected "equal") map being the starting point and all existing proposals being tossed. No one who submitted a map in an earlier round could submit another one in later rounds.

[Brett Bellmore]

What I've proposed is that a computer be used to generate thousands of maps starting from random seeds, that all comply with the requirements of compactness and equal population, and respecting natural boundaries, while NOT being based on voting patterns. Doing so is a well established process at this point, the software can be open source.

Then in a process like voir dire, each of n ballot qualified parties gets to eliminate 1/(n+1) of the maps, on any basis they want.

Finally, the map that actually gets used is selected from the remaining maps by bingo cage in public.

This guarantees that it is impossible to actually gerrymander, each party gets to eliminate the maps most damaging to its interests anyway, while the resulting churn in district borders every 10 years will give challengers a chance.

[tkamenick]

I agree that solutions like this are the most ideal. However, they'd probably have net negative effects on racial minority representation compared to the status quo so they'd never fly.

Of course, the people who scream the loudest about how bad partisan gerrymandering is have zero problem engaging in racial gerrymandering to benefit favored groups.

[Brett Bellmore]

Well, yeah: If you're racially gerrymandering to artificially inflate minority representation in Congress, (Which we are doing under the Voting Rights Act.) then ending gerrymandering would reduce said representation.

But you can't do anything about gerrymandering while mandating it, that's a circle the Court couldn't square, so they just threw up their hands. The racial gerrymanders have to go away with the political ones, or its not happening at all.

[MatthewSlyfield]

And there in lies the problem with efforts to outlaw Gerrymandering. Current case law on the Voting Rights Act and minority representation effectively mandates at least some Gerrymandering.

[JasonT20]

It was a long time ago, but I do remember reading about someone that took an algorithmic approach like this to drawing districts and compared the likely results of that with actual results. I don't remember how it turned out, but it would be interesting to see.

If there are going to be single member districts of any shape, there is value in having the districts keep some contiguous communities together, rather than drawing boundaries right through them. It wouldn't make sense to split a small town in half between two different legislative districts, for instance. This is why the wording of districting rules like those in the Florida constitution talk about "respecting existing communities with shared interests" or something like that.

Rather than going with really small units, maybe chunk populations in reasonable ways centered around neighborhoods or other characteristics. But then the chunks would need to be pretty small, still, in order for the algorithm to create districts of close to the same size.

[Brett Bellmore]

"It was a long time ago, but I do remember reading about someone that took an algorithmic approach like this to drawing districts and compared the likely results of that with actual result"

Chen and Cotrell. https://electionlawblog.org/?p=91074

[Dr. Ed 2]

What's not being mentioned here are the "Majority Minority" setaside House districts which are mandated by the US DoJ's interpretation of the Voting Rights Act.

This is how you have people like Ayanna Pressley in the House -- she's in Massachusetts' set-aside Congressional District. I believe that Occasional Cortex also is in NY, and there are others who are only in Congress because they are in incredibly low-turnout districts.

The late Abby Thernstrom wrote about this in _Voting Rights and Wrongs_ -- a book I highly recommend.

[JasonT20]

What's not being mentioned here are the "Majority Minority" setaside House districts which are mandated by the US DoJ's interpretation of the Voting Rights Act.

It really would be awesome if legislative districts could be drawn without regards to race, ethnicity, or any other individual characteristic. Too bad the U.S. had a history of reducing the voting power of non-Whites to the point where making districts like that was the only way for them to get representation in their government.

[Dr. Ed 2]

Ever hear of a man named Barack Obama?

He lost a race for a set-aside seat and then won a regular one.

That aside, the way to end racial discrimination is to stop discriminating on the basis of race.

[DixieTune]

Minorities would still get representation in government, and have the right to vote for whomever they want. It just wouldn't be guaranteed to be someone who looks like them.

[JasonT20]

You're missing the point of allowing minority groups to have a majority within a district. It isn't about being able to elect someone that "looks like them." It is about electing someone that has to represent their interests in order to win and be reelected. Black people making up 20% of every district in a state might as well be without a vote at all in the Jim Crow South, where no candidate of any party could get to a majority if they represented their Black constituents and pushed for their equal rights. White voters that didn't want Blacks to have equal rights made up a majority of people that voted in every district in those states at that time. That was actually a tactic segregationists used to dilute the influence of Black voters, even if the efforts of voter suppression failed. Spread out Black voters into majority-white districts, and they could safely be ignored.

In the ideal world, voters of different races, ethnic groups, or religious traditions would not be pitted against each other, so that no voters would need their group to be a majority in a district in order for that group to have real representation from the winner. We still don't live in that ideal world.

[Brett Bellmore]

And you're missing that it's racist to think that only blacks, (Or white Democrats...) can represent blacks.

[JasonT20]

I never said that, though. Black people, as a voting bloc, are perfectly capable of choosing a White, Hispanic, Asian, Native American, mixed-race person, or anything else to represent them if they want. A majority-minority district doesn't stop anyone not Black from running, any more than a majority-white district prevents a Black person from running.

The problem is whether the voters themselves will accept someone that doesn't "look like them." There weren't many Whites in the Jim Crow South that would, and few enough White people were trusted by Blacks to represent them that it turned out that they preferred Black candidates. For all of the progress we've made since then, there are still a significant number of voters of all races that will only feel that someone that shares their life experience will represent them. And that "sharing their life experience" means having lived in society with the same skin color. I'd say that this is an unfortunate truth, and that we should all be ignoring each other's skin color in everything. But it is a reality that someone that grows up in America White experiences life in this country differently than someone that grows up Black, or Hispanic, Asian, or something else. And that this is true even if other circumstances, like income, wealth, or education are the same.

I don't see any of these observations as being "racist". You seem to be pulling from the same deck as many others I've seen over the years as they argue about this. The premise is that we 'fixed' racism at some point in the past after the Civil Rights Act and Voting Rights Act, therefore it is "racist" not only to treat people of different races differently, even with the goal of trying to right those historic wrongs, but to believe that the historically oppressed groups are not, in the present, truly equal in all of their opportunities to the groups that used to be the oppressors.

I am not sure that this is what you think, but I do want to be clear that I reject this premise. Society doesn't go from systematic discrimination to truly equal opportunity only by outlawing discrimination. The Civil Rights movement accomplished changing the law to make discrimination illegal, but building up the will to enforce the law everywhere it was being violated took decades. And there were gaps in the anti-discrimination laws, plus the fact that there isn't always objective evidence of discrimination when it happens. And then there are law enforcement policies that were race-neutral on their face that impacted poor, minority-heavy populations much more than they did White communities (War on Drugs). All of that and more is why we still don't have the color-blind society with equal opportunity that would be ideal.

[Katall]

creative ways of referring to elected officials' names is a sure way of convincing people that you are not a buffoon. Well done.

[JasonT20]

This suggests that, whatever one thinks of gerrymandering and of geographic representation (as opposed to proportional representation), they didn't seem to have a particularly distortive effect on this year's race. (For the aggregate House vote totals, I'm looking at the Cook Political Report totals.)

Gerrymandering is done by individual states. It is entirely likely that the gerrymandering in states with Democratic Party majority legislatures is nearly cancelling out the gerrymandering in states with Republican Party majority legislatures. That would still be a large problem because the whole concept of gerrymandering takes some of the power to choose leaders away from the voters. In addition to that, it isn't just about how the number of seats compares to the total votes, it is about how competitive the races are. The aggregate number of votes for House candidates for each party can't that much when 37 of the 435 House races only had a candidate for one of the two major parties on the ballot. Add in token opposition in a most other races, and what do you get? (69 competitive races according to the Cook Political Report ahead of the election - https://www.cookpolitical.com/ratings/house-race-ratings?mod=article_inline)

[JasonT20]

Looks like 17 seats flipped out of 435. 418 districts returned a member of the same party.

[The Margrave of Azilia]

So now that the data is in - did third parties steal the election from the Democrats, and was this third-party activity all a plot by deep-pocketed Republicans?

[NvEric]

Geographically - from a chosen center of a district expand out as equally as possible, population dependent. Think of bubbles in a head of beer, ale, or stout.

Prudently, the numbers in the House are grossly small and have not increased in 100 years, save Alaska and Hawaii, the true deficiency in our system. Regional entities can address this problem by having Regional Congresses. Or, by reducing the federal government itself.

Governance is too important to be left to government.

[davidk]

Ironically, of course, it's gerrymandering that allows a first-past-the-post electoral system to be proportional. Suppose X% of people are Republican supporters. Imagine two ways of distributing votes to district-- in one, every district perfectly reflects the country as a whole, so each district has X% Republicans. In the other, districts are all gerrymandered so that X% of districts are 100% Republican and 100-X% of districts are 100% Democratic. Under FPTP, the perfectly-ungerrymandered scenario generates a congress that's either 100% Republican or 100% Democratic, depending on whether X is >= 50%. Meanwhile the gerrymandered scenario always ends with exactly X% Republicans in the legislature.

I hate FPTP and I hate gerrymandering-- I'm a big PR supporter-- but as a PR supporter, if we have FPTP... gerrymandering really isn't so bad, at least as long as it's done by both parties.

[JasonT20]

I hate FPTP and I hate gerrymandering-- I'm a big PR supporter-- but as a PR supporter, if we have FPTP... gerrymandering really isn't so bad, at least as long as it's done by both parties.

There could be worse results than gerrymandering, sure. But that doesn't mean that the best we can do with FPTP is for partisan legislatures to draw maps that favor the party in power, and then hope that it evens out nationally for the U.S. House. That still leaves a lot of state legislative majorities able to reinforce their own control over their state legislatures. What's bad for democracy on one scale is just bad for democracy, period.

The worst thing to do is to become complacent at let it be seen as normal to engage in the kind of maneuvering that gerrymandering exemplifies. What's 'normal' can become 'acceptable' or even 'preferred' if those that like it can keep getting their way long enough.

[Brett Bellmore]

Gerrymandering only has that ironic effect so long as both parties are roughly at equal strength; Once one of the parties dominates a state during redistricting, that state effectively gets locked in for them against changes in voter opinion.

We saw that nation-wide in the US for a long while; Democrats had dominated for a while after the 1929 crash, and gerrymandered like mad. The result is that the dominated Congress completely for better than half a century, until the 1994 election.

That's where the downside of gerrymandering for more seats kicked in: You get more seats by concentrating your opponents' votes, and spreading yours around into as many districts that you just barely win as possible. Then if a wave election comes along for the opposition, your wide, shallow base is flooded over, and they win massively.

That's why most gerrymandering is actually for incumbent protection, NOT to increase a party's seats: The parties know that latter application of gerrymandering is very risky if you carry it too far.

[DixieTune]

That's what happened to the Virginia House of Delegates in 2017.

[Dan Schiavetta]

There are many districts which are so heavily Republican that Democrats know it’s a waste of time to even put up a candidate. More so than the other way around. This changes the analysis and the conclusion.

[Brett Bellmore]

You literally have that completely backwards. Republicans mostly have districts where they rack up a modest majority of the vote, while Democrats have many districts that they totally dominate.

When you graph vote percentages by precinct, the Democrats' side of the graph has a long tail, and the Republicans' side a short tail. Precincts that are more than 65% Republican are quite rare, precincts that are more than 65% Democratic are quite common.

[Frank Drackman]

Like the Late John Lewis's Shithole GA-5, which went for Cums-a-lot 90/10, it's like D.C. without the Historical Sites (Sherman burned them) and more crime.

[John F. Carr]

In Massachusetts Democrats win big due to inefficient distribution of Republicans. If the state had two or three multi-member districts Republicans would get a couple House seats, which they deserve if you think parties deserve representation in proportion to their voters.

[larrydj9000]

I wonder how the vote totals are calculated in districts where there may be two competitors from the came party. For example, California mandates the top two vote getters in a jumbo primary face off in the general election. In a very blue district you may have two Dems competing, no GOP. Are both counted in the Dem column for the national totals, and zero for GOP? Or some other method? The concept of a national popular vote gets distorted once one adopts jumbo primaries and subsequent runoffs. Rank choice voting is another example.

[MonitorsMost]

I think that is how you can end up with differences between the calculations between different sources. You are correct that a national popular vote gets distorted based on these types of things, but in the aggregate they will still be approximately correct. The net distortion is not going to be significant if the underlying question is "did the popular vote largely map onto the number of seats won." The answer is yes.

[Peter Kreeft taught me LOGIC, BEWARE !!!!!!!]

If you have a Black, male, Catholic, homosexual, conservative and you make any decision based on any trait, you are voting for him and not in any way 'enabling' his vote. Why? Because what you think is his identification is only your view of him and not his view of himself -- which at any rate changes over time and over issue.

[MonitorsMost]

This has been the case for quite a while. The consistently larger discrepancies between popular vote and number of house seats petered out in the early 90s.
https://www.brookings.edu/wp-content/uploads/2017/01/vitalstats_ch2_tbl2.pdf
As others have noted, the the existence of gerrymandering could impact the number of votes cast in the first place. But it is also worth noting that alternatives to partisan map drawing such as independent commissions do not necessarily make the discrepancy between popular vote and number of seats better. California has consistently massive gaps between the popular vote and number of seats, despite having an independent redistricting commission. ProPublica even had an article about how the process was manipulated by partisans.

https://www.propublica.org/article/how-democrats-fooled-californias-redistricting-commission

If matching the number of seats to popular vote is your number one priority, the answer is moving to a parliamentary system. Such a system does have the downside of having less of constituency for representatives. But there are going to be trade-offs in all systems.

[Brett Bellmore]

Multi-member districts and proportional representation can match seats to votes while still having particular constituencies for members.

[edbeau99]

When you go all the way up to the national level, the effect of gerrymandering will be swamped by other factors, and since both parties use it to some extent, that will even out the effect.

But let's look at the excellent example of New York State. In 2020, under the map that had been gerrymandered by the Democrats, they won 19 seats to 8 Republicans. But in 2022, as a result of judicial intervention, a non-partisan map was drawn and the results produced 15 Democrats and 11 Republicans (one seat lost due to redistribution). Smarting from having their gerrymandering overturned, the Democrats changed the law to limit judicial challenges, put in another gerrymandered map, and produced 19 Democrats, 7 Republicans. So the gerrymandering gave them 4 (you could argue 5) unearned seats, significant even at the national level given how close 2024 was for the House.

[edbeau99]

In Canada, the concept of gerrymandering is almost entirely unknown (there was one case in one province back in the 1980's where a single seat had questionable boundaries - ever afterwards known as the "Affair of Gracie's Finger").

That is because an independent and non-partisan federal agency, Elections Canada, sets out the boundaries of each constituency. When adjustment is needed due to population changes, new boundaries go through a process of political and public review. In the event of something attracting widespread criticism, the decision can be referred to judicial arbitration, but that almost never happens. Similar processes work at the provincial level.

Another excellent reason why Canadians would never want to be tainted by corrupt American election practices.