It is impossible to draw a proportional map under first-past-the-post if you want it to be proportional when one party wins a very large majority of the vote.
If one party is winning more than two-thirds of the vote, then the minority party needs to have very high concentrations of votes to win any seats at all. Consider it nationally - a D+33 national result would make every single district blue - TX-13 is the reddest district in the nation and it's "only" R+33. An R+33 national result would result in 17 districts staying Democratic (in order of PVI: FL-24, GA-05, MA-07, NY-09, CA-34, CA-44, NJ-10, NY-08, CA-12, CA-37, NY-05, IL-07, NY-07, CA-13, PA-03, NY-13, NY-15). I'm using Cook PVI.
Note that only three of those 17 Democratic districts are in states where the boundaries are drawn by a GOP legislature (FL-24, GA-05 and PA-03) - they're not really the result of "packing", but of the intense concentration of Democratic votes in the most urban areas, and the most African-American areas - unless I've missed one, these are all majority-minority districts and all but NJ-10 (Newark) are in the inner core of cities big enough to have multiple congressional districts.
A proportional set of boundaries would have 145 districts at least this Republican and 145 at least this Democratic. That's utterly impossible.
The most partisan district in the nation is NY-15 - D+44. If the Republicans won that seat purely on the national vote, that would be an R+44 election; that's a 72-28 national generic ballot. Under a proportional system, there should be 121 districts at least that partisan, and the most partisan district should be something like D+99.8 (and equally one at R+99.8)
Within the constraints of single-member districts of equal size that cannot cross state boundaries, you cannot get proportionality outside of a relatively narrow band around 50-50. Fortunately, US election results are almost always in a relatively narrow band around 50-50.
Require proportional results (to within a margin of error of something like 5%, ie it can be off by no more than one seat in any state that isn't California or Texas) for any statewide result where no party exceeds 60% of the vote statewide, and accept that the system will inflate statewide results above that level in terms of representation.
Massachussetts was 75-21 in 2020 to the Democrats, of course they won all nine seats.
I just want to throw this out there for informational purposes:
The supreme court's 2019 ruling on gerrymandering is very misunderstood. The common summary of it was that the supreme court said that political gerrymandering was ok. What they actually said is that it's bad but they didn't want to make a standard to judge what is and isn't gerrymandering. However, they were very explicit that if someone came up with a standard, they would enforce it.
In particular, they were very clear that congress could come up with this standard because the constitution reads: "The Times, Places and Manner of holding Elections for Senators and Representatives, shall be prescribed in each State by the Legislature thereof; **but Congress may at any time make or alter such Regulations,** except as to the Place of chusing Senators." The supreme court decided that drawing districts fell under "manner", so congress could set a standard that states had to follow.
Now, one dirty little secret is that many (most?) states already have requirements on districts, but they're routinely ignored. For example, the Pennsylvania constitution states "Unless **absolutely necessary** no county, city, incorporated town, borough, township or ward shall be divided in forming either a [state] senatorial or [state] representative district." (Article II section 16.) If you look at, say, PA state senate districts, they divide a lot more county lines than necessary. However, almost no one bothers to challenge maps under existing standards! It's insane. People want the supreme court to intervene with a new standard when not bothering to use the tools that already exist. Other seriously gerrymandered states (NC, WI, etc.) have similar laws, but again, no one bothers to use them. WTF?
Any reason why expanding the House doesn't get more play? There's no obvious reason we are stuck with 435 representatives the Reapportionment Act of 1929 stuck us with. It wouldn't be perfect, but it would reduce the advantage rural states have, and it would allow major cities to have more representation proportionally.
I think the gerrymandering stuff would be tricky constitutionally for the feds to enforce, but House size is totally in their remit. I know in order to achieve constituents per representative at 1929 levels we'd need a couple thousand members, but, I don't know, it seems like an interesting idea that isn't talked about much.
"Proportional maps would lead to more rather than fewer Black Democrats winning elections simply because more Democrats would win elections, and a healthy share of them would be white."
I'll repeat my oft-stated point: being popular and convincing Americans to vote for you is more important than fiddling with gerrymandering rules.
Between 2002 and 2020, the Democrats won the national 2-party vote for House seats five times. They won a majority of House seats four times (the exception being the 2012 vote that everyone, including Matt, focuses on, when they got a bare majority with 50.6% of the 2-party vote but the Republicans got a majority of the seats). In the other five elections, they got a minority of the 2-party vote and a minority of the seats.
In 2018 and 2020, the Democrats won 54.4% and 51.6% of the popular vote, and took 54.1 and 51.2% of the House seats.
I know people like Wasserman say that the Republicans can retake the House just due to redistricting, but I disagree. Have a booming economy and defeat COVID, don't let the activists define the party, and maybe have party leaders pick some especially juicy Sister Souljah moments, and get really good candidates who are well-fitted to their districts. Win a majority of the votes!
If anything, getting some of those MA Dems to go out and very publicly fall on their sword -- "This bill will almost certainly mean that some of us don't get re-elected because it will mean MA elects some Republicans, proportional to how many MA Republicans actually vote, but we're voting for it anyways because the US House _should_ have some MA Republicans, balanced out by having the fair number of NC Democrats" -- would probably be good PR for getting the bill passed. We should be promising those MA Dems whatever it takes to get them to play ball.
Any federal law to limit gerrymandering runs into a buzzsaw of challenges on states rights grounds that would come to a gerrymandering friendly Supreme Court.
The fundamental asymmetric problem with fighting gerrymandering is that liberals want districts to be *fair* while conservatives want districts to give them a permanent majority like they have in Michigan and Wisconsin. Not that liberals don't gerrymander when given the opportunity. Maryland has some of the worst shaped districts in the country.
Arguing about proportion seems unwinnable. Why not strive for cohesion and community and let partisanship fall where it may? Excessively gerrymandered districts lump people from legitimately different communities instead of allowing them to share representation. To avoid districts like the MD-3, one could start with a map of all school boundaries in a state (since they always exist are fairly granular and schools are public spaces used for polling). A computer would not have much trouble drawing a map with two rules. 1) Make the population of n districts as even as possible, and 2) minimize the edge to volume ratio. One would get compact, cohesive and even districts. Let the politics and demographics fall where they may.
Is anyone out there working on the actual text of a "really good anti gerrymandering bill"? If so, who? If not, why doesn't somebody in these comments write it so we can e-mail it to Schumer and Manchin? I'm like 50% serious here. It doesn't seem like it would be all that difficult to write, nor would it need to be all that long.
As a resident of a state (Wisconsin) in which voting has very little to do with political outcomes below the statewide level, I'm all in favor of anti-gerrymandering legislation. But I think you're a little too optimistic about its likely effect. Given the geographic concentration of Democratic voters in Wisconsin, you'd need some very, very strange districts to produce maps that make the "median district’s partisan composition...as close as possible to the mean partisanship of the state," as you advocate in your footnote. Proportional representation is the only way I see to address the issue, and there's no way I see to get from here to there. BTW, Nirvana's Krist Novoselic (I am not making this up) does good work on proportional representation.
I think proportional representation with multi member districts is really the best solution here, but obviously not one at all under consideration in the current discourse. What do you think about this idea and what can be done to inject it into the conversation as a real solution?
I think this is really important and often overlooked:
"Last but by no means least, this whole line of thought rests on a fairly arbitrary effort to draw a line between “campaign-related disbursements” and other ways money can influence politics."
People talk about campaign donations quite a bit but so much of our media (news and entertainment) contains political messaging, often overtly, and is created by wealthy people and companies. Cop shows influencing views on policing is a version of money in politics as is making a "Trump like figure" the villain of your TV show sometime often the last 4 years.
If people only care about the idea that politicians are being "bought" by donations that is one thing but if they care about rich people and companies having a wildly outsized influence and platform to disseminate political messaging that is going to require a lot more thought.
You leave out the voter registration and ballot access provisions of the bill. Since state level Republicans are in a mad rush to restrict access to both, the matter deserves some consideration. One of the problems with Democrats elected from "safe" seats is they tend to view their office as a lifetime entitlement. This has a negative effect on the incentives to work for their constituents. Whatever the content of the policies advocated by Justice Democrats, shaking up that entitlement attitude is a positive development.
It seems to me that people tend to write about HR1 and the Democratic majority as though they are (still) a heard of cats devoid of political skills. It's as if Trump never happened and McConnell is somehow still secretly the majority leader.
I've read numbers takes that start "HR1 was a messaging bill, but now it's serious". Really? Why isn't it still a messaging bill? And plenty of "Joe Manchin is standing in the way of <DEM_PRIORITY>" takes. Really? Maybe he is just playing the foil?
The Democrats advance HR1 and then: a) Republicans filibuster it, giving Democrats the public support needed to reform the filibuster, with Manchin's "reluctant approval". b) Republicans, fearing the Democrats are just crazy enough to blow up the filibuster---oh, Manchin, will he/won't he?---offer amendments and try to negotiate some kind of compromise. c) Democrats are crazy enough to blow up the filibuster and pass HR1, triggering a fundraising bonanza (for both sides) and also some real reforms. d) What Matt said; play chicken with HR1 until Manchin "is convinced" and then start all over with a gerrymandering bill.
a-d are all winners for Democrats; what is the e) Democrats lose and DeSantis nominates Trump to the Supreme Court in 2025 scenario?
Democrats are all ready testing the waters by daring Republicans to filibuster a bill meant to curb Asian hate-crimes.
Also, does anyone remember back when bills were debated, amended and improved before passing?
Reducing the political power of super-rich people and moneyed corporate interests, and increasing the power of people with a few hundred dollars of disposable income to donate to their preferred political candidates, is good, actually.
It is impossible to draw a proportional map under first-past-the-post if you want it to be proportional when one party wins a very large majority of the vote.
If one party is winning more than two-thirds of the vote, then the minority party needs to have very high concentrations of votes to win any seats at all. Consider it nationally - a D+33 national result would make every single district blue - TX-13 is the reddest district in the nation and it's "only" R+33. An R+33 national result would result in 17 districts staying Democratic (in order of PVI: FL-24, GA-05, MA-07, NY-09, CA-34, CA-44, NJ-10, NY-08, CA-12, CA-37, NY-05, IL-07, NY-07, CA-13, PA-03, NY-13, NY-15). I'm using Cook PVI.
Note that only three of those 17 Democratic districts are in states where the boundaries are drawn by a GOP legislature (FL-24, GA-05 and PA-03) - they're not really the result of "packing", but of the intense concentration of Democratic votes in the most urban areas, and the most African-American areas - unless I've missed one, these are all majority-minority districts and all but NJ-10 (Newark) are in the inner core of cities big enough to have multiple congressional districts.
A proportional set of boundaries would have 145 districts at least this Republican and 145 at least this Democratic. That's utterly impossible.
The most partisan district in the nation is NY-15 - D+44. If the Republicans won that seat purely on the national vote, that would be an R+44 election; that's a 72-28 national generic ballot. Under a proportional system, there should be 121 districts at least that partisan, and the most partisan district should be something like D+99.8 (and equally one at R+99.8)
Within the constraints of single-member districts of equal size that cannot cross state boundaries, you cannot get proportionality outside of a relatively narrow band around 50-50. Fortunately, US election results are almost always in a relatively narrow band around 50-50.
Require proportional results (to within a margin of error of something like 5%, ie it can be off by no more than one seat in any state that isn't California or Texas) for any statewide result where no party exceeds 60% of the vote statewide, and accept that the system will inflate statewide results above that level in terms of representation.
Massachussetts was 75-21 in 2020 to the Democrats, of course they won all nine seats.
I just want to throw this out there for informational purposes:
The supreme court's 2019 ruling on gerrymandering is very misunderstood. The common summary of it was that the supreme court said that political gerrymandering was ok. What they actually said is that it's bad but they didn't want to make a standard to judge what is and isn't gerrymandering. However, they were very explicit that if someone came up with a standard, they would enforce it.
In particular, they were very clear that congress could come up with this standard because the constitution reads: "The Times, Places and Manner of holding Elections for Senators and Representatives, shall be prescribed in each State by the Legislature thereof; **but Congress may at any time make or alter such Regulations,** except as to the Place of chusing Senators." The supreme court decided that drawing districts fell under "manner", so congress could set a standard that states had to follow.
Now, one dirty little secret is that many (most?) states already have requirements on districts, but they're routinely ignored. For example, the Pennsylvania constitution states "Unless **absolutely necessary** no county, city, incorporated town, borough, township or ward shall be divided in forming either a [state] senatorial or [state] representative district." (Article II section 16.) If you look at, say, PA state senate districts, they divide a lot more county lines than necessary. However, almost no one bothers to challenge maps under existing standards! It's insane. People want the supreme court to intervene with a new standard when not bothering to use the tools that already exist. Other seriously gerrymandered states (NC, WI, etc.) have similar laws, but again, no one bothers to use them. WTF?
Any reason why expanding the House doesn't get more play? There's no obvious reason we are stuck with 435 representatives the Reapportionment Act of 1929 stuck us with. It wouldn't be perfect, but it would reduce the advantage rural states have, and it would allow major cities to have more representation proportionally.
I think the gerrymandering stuff would be tricky constitutionally for the feds to enforce, but House size is totally in their remit. I know in order to achieve constituents per representative at 1929 levels we'd need a couple thousand members, but, I don't know, it seems like an interesting idea that isn't talked about much.
"Proportional maps would lead to more rather than fewer Black Democrats winning elections simply because more Democrats would win elections, and a healthy share of them would be white."
Is this last word suppose to be black?
I'll repeat my oft-stated point: being popular and convincing Americans to vote for you is more important than fiddling with gerrymandering rules.
Between 2002 and 2020, the Democrats won the national 2-party vote for House seats five times. They won a majority of House seats four times (the exception being the 2012 vote that everyone, including Matt, focuses on, when they got a bare majority with 50.6% of the 2-party vote but the Republicans got a majority of the seats). In the other five elections, they got a minority of the 2-party vote and a minority of the seats.
In 2018 and 2020, the Democrats won 54.4% and 51.6% of the popular vote, and took 54.1 and 51.2% of the House seats.
I know people like Wasserman say that the Republicans can retake the House just due to redistricting, but I disagree. Have a booming economy and defeat COVID, don't let the activists define the party, and maybe have party leaders pick some especially juicy Sister Souljah moments, and get really good candidates who are well-fitted to their districts. Win a majority of the votes!
If anything, getting some of those MA Dems to go out and very publicly fall on their sword -- "This bill will almost certainly mean that some of us don't get re-elected because it will mean MA elects some Republicans, proportional to how many MA Republicans actually vote, but we're voting for it anyways because the US House _should_ have some MA Republicans, balanced out by having the fair number of NC Democrats" -- would probably be good PR for getting the bill passed. We should be promising those MA Dems whatever it takes to get them to play ball.
Any federal law to limit gerrymandering runs into a buzzsaw of challenges on states rights grounds that would come to a gerrymandering friendly Supreme Court.
The fundamental asymmetric problem with fighting gerrymandering is that liberals want districts to be *fair* while conservatives want districts to give them a permanent majority like they have in Michigan and Wisconsin. Not that liberals don't gerrymander when given the opportunity. Maryland has some of the worst shaped districts in the country.
Arguing about proportion seems unwinnable. Why not strive for cohesion and community and let partisanship fall where it may? Excessively gerrymandered districts lump people from legitimately different communities instead of allowing them to share representation. To avoid districts like the MD-3, one could start with a map of all school boundaries in a state (since they always exist are fairly granular and schools are public spaces used for polling). A computer would not have much trouble drawing a map with two rules. 1) Make the population of n districts as even as possible, and 2) minimize the edge to volume ratio. One would get compact, cohesive and even districts. Let the politics and demographics fall where they may.
Is anyone out there working on the actual text of a "really good anti gerrymandering bill"? If so, who? If not, why doesn't somebody in these comments write it so we can e-mail it to Schumer and Manchin? I'm like 50% serious here. It doesn't seem like it would be all that difficult to write, nor would it need to be all that long.
As a resident of a state (Wisconsin) in which voting has very little to do with political outcomes below the statewide level, I'm all in favor of anti-gerrymandering legislation. But I think you're a little too optimistic about its likely effect. Given the geographic concentration of Democratic voters in Wisconsin, you'd need some very, very strange districts to produce maps that make the "median district’s partisan composition...as close as possible to the mean partisanship of the state," as you advocate in your footnote. Proportional representation is the only way I see to address the issue, and there's no way I see to get from here to there. BTW, Nirvana's Krist Novoselic (I am not making this up) does good work on proportional representation.
I think proportional representation with multi member districts is really the best solution here, but obviously not one at all under consideration in the current discourse. What do you think about this idea and what can be done to inject it into the conversation as a real solution?
I think this is really important and often overlooked:
"Last but by no means least, this whole line of thought rests on a fairly arbitrary effort to draw a line between “campaign-related disbursements” and other ways money can influence politics."
People talk about campaign donations quite a bit but so much of our media (news and entertainment) contains political messaging, often overtly, and is created by wealthy people and companies. Cop shows influencing views on policing is a version of money in politics as is making a "Trump like figure" the villain of your TV show sometime often the last 4 years.
If people only care about the idea that politicians are being "bought" by donations that is one thing but if they care about rich people and companies having a wildly outsized influence and platform to disseminate political messaging that is going to require a lot more thought.
You leave out the voter registration and ballot access provisions of the bill. Since state level Republicans are in a mad rush to restrict access to both, the matter deserves some consideration. One of the problems with Democrats elected from "safe" seats is they tend to view their office as a lifetime entitlement. This has a negative effect on the incentives to work for their constituents. Whatever the content of the policies advocated by Justice Democrats, shaking up that entitlement attitude is a positive development.
It seems to me that people tend to write about HR1 and the Democratic majority as though they are (still) a heard of cats devoid of political skills. It's as if Trump never happened and McConnell is somehow still secretly the majority leader.
I've read numbers takes that start "HR1 was a messaging bill, but now it's serious". Really? Why isn't it still a messaging bill? And plenty of "Joe Manchin is standing in the way of <DEM_PRIORITY>" takes. Really? Maybe he is just playing the foil?
The Democrats advance HR1 and then: a) Republicans filibuster it, giving Democrats the public support needed to reform the filibuster, with Manchin's "reluctant approval". b) Republicans, fearing the Democrats are just crazy enough to blow up the filibuster---oh, Manchin, will he/won't he?---offer amendments and try to negotiate some kind of compromise. c) Democrats are crazy enough to blow up the filibuster and pass HR1, triggering a fundraising bonanza (for both sides) and also some real reforms. d) What Matt said; play chicken with HR1 until Manchin "is convinced" and then start all over with a gerrymandering bill.
a-d are all winners for Democrats; what is the e) Democrats lose and DeSantis nominates Trump to the Supreme Court in 2025 scenario?
Democrats are all ready testing the waters by daring Republicans to filibuster a bill meant to curb Asian hate-crimes.
Also, does anyone remember back when bills were debated, amended and improved before passing?
Reducing the political power of super-rich people and moneyed corporate interests, and increasing the power of people with a few hundred dollars of disposable income to donate to their preferred political candidates, is good, actually.
Do you have an example of a proportional Massachusetts map?
From what I understand, it's not really easy to design such a map in MA - the GOP voters are too interspersed throughout the state.