What the Republicans are doing right now on gerrymandering is despicable - but there should be real soul searching on the Dem side for why Democrats were so unfocused and epistemically unsound when we had a chance to fix this four years ago.

@asthanaprav Both blue and red states gerrymander but on net the house currently has a 1.5% pro-Republican bias and it's about to get substantially worse due to Trump's historically quite unusual multi-state push.

@HarmstonVe56329

@brianbeutler @mattyglesias Regardless of how you feel about Manchin/Sinema, during the whole HR 1 process we repeatedly lied to the public and lied to our donors and shouted down experts who tried to tell the public the truth. I think that’s bad for its own sake.

@CathleenIsabel3 If you're genuinely interested in the extent to which Democratic and Republican states gerrymander and the net effect - I made an interactive web applet that measures how bad each state is on two measures of partisan bias and explaisn what they are!

@brianbeutler @mattyglesias Big dollar donors were being told that Manchin was open to things right up until maybe a month before the floor vote. Maybe they were just getting lied to - if so that's bad too! The whole thing was epistemically terrible and I don't see why you're changing the subject.

@jerrymayEwriter @tom_paine1737 The net partisan impact of not counting undocumented immigrants in the census is ~ zero and it honestly reflects poorly on the analytical capacity of the conservative movement that basically none of them seem to realize it

@steve__case @jerrymayEwriter @tom_paine1737 Republicans have made giant gains with low income non-white voters in the last eight years so it doesn’t have a partisan impact anymore or if anything the signs have flipped - both coalitions have not adopted yet - campaign finance is similar

@SpeakingBee @tom_paine1737 @asthanaprav I encourage you to look through this!

@SaveFarrisLSU @joeykatzen If you go through this state by state it is clear that Republicans are net beneficiaries

@HFlashmanVCKCB

@Dbarenholtz1

@VirginiaHo33518

@agentunknownuno

@BrendonPeak1

@DCCyclone @brianbeutler @mattyglesias There was a serious and gross insider/outsider vibe where people in the know knew it wouldn’t pass almost immediately while the donor class was cynically strung along for money for a year. I don’t know how anyone who was close to the process could not think mistakes were made.

@tbone9070

@weinerimer @asthanaprav It's really not for two reasons: 1) house vote share doesn't take into account unopposed districts or varying turnout in unopposed districts well 2) presidential vote is more correlated with one cycle out congressional vote than lagged congressional vote

@weinerimer @asthanaprav If you go look at the court documents of Republican map maker correspondance in places like TX or NC, they heavily referenced presidential vote to do their benchmarking - this has been standard practice for several decades. You have to assume they know what they're doing!

@weinerimer @asthanaprav If you're genuinely curious about this, the autoregressive structure is basically if you have a race with an incumbent, cong_t ~ pres_{t-2} + 0.7*(cong_{t-2}-pres_{t-2}) , while it's cong_t ~ pres_{t-2} for open races. IE, there's strong mean reversion to presidential.

@DCCyclone @brianbeutler @mattyglesias Relatedly though - my critique of policy demanders is honestly less about them demanding policy and more that they often choose their policy demands via a taylorist donation/grant maximizing hellscape machine that incentivizes them to lie to electeds/donors/press all the time

@DCCyclone @brianbeutler @mattyglesias It’s totally fashionable to hate on using polling to make decisions but for some reason making strategic decisions based on what raises you the most money or boosts your list building metrics for grantees gets no criticism even though it’s much more common

@Dbarenholtz1 What people in blue states will tell you is that if they don’t try to make their maps good then they’re unilaterally disarming. But a mid-decade redistricting is different. Republicans are just trying to max out the chance that they hold power with a minority of the vote.

@Alsadius @LiamTuc43138389 I just really think "the non-white people who disproportionately live in cities should always have their political power diluted in perpetuity because that is better than having a map making procedure that looks at election results or race" is such a crazy belief to say out loud

@Alsadius @LiamTuc43138389 I will say if you go throughout history the track record of anti-majoritarianism is really very bad.

@Alsadius @LiamTuc43138389 My preferred system is Ireland’s multi member district system with STV - super proportional, not party based in any way, fairly robust to gerrymandering. But if you’re going have single member districts you should draw them to be fair rather than choosing not to do that.

@BrendonPeak1 Which to be clear is exactly what the spreadsheet said! I did a lot of work to actually collect all this data!

@BrendonPeak1 You have to use presidential for these things - it's more correlated with one cycle out vote choice and otherwise you don't have any way to handle unopposed districts

@BrendonPeak1 What's common is instead for people to use congressional vote by state instead - but then you run into issues with unopposed seats. Everyone who does this for a living and actually has to predict elections as their job knows to look at presidential vote.