Re: Cry, cry, redistrict, cry

1

I am reliably informed that politics is not, in fact, beanbag.


Posted by: ogged | Link to this comment | 02- 6-18 9:06 AM
horizontal rule
2

If life gives you bean bags and some plywood, play cornhole.


Posted by: Moby Hick | Link to this comment | 02- 6-18 9:14 AM
horizontal rule
3

Just checked PA constitution. It looks like removal is parallel with the federal process, requiring 2/3 of the state senate, and GOP currently has... 68%. Fuck.


Posted by: Minivet | Link to this comment | 02- 6-18 9:52 AM
horizontal rule
4

I wonder how they got 2/3rds of the state senate in a state that elects a Democratic governor?


Posted by: Moby Hick | Link to this comment | 02- 6-18 9:54 AM
horizontal rule
5

So what is the level of federal illegality past which it would in fact be better for blue states to secede (assuming arguendo they could)? Literally rounding up and arresting Dem legislators? What if it's subtler like in Hungary?


Posted by: Minivet | Link to this comment | 02- 6-18 9:54 AM
horizontal rule
6

Fuck that. I want to win, not retreat.


Posted by: Moby Hick | Link to this comment | 02- 6-18 10:02 AM
horizontal rule
7

It's not what I want or would fight for any way ATM, but I want to have a plan for when to pivot.


Posted by: Minivet | Link to this comment | 02- 6-18 10:27 AM
horizontal rule
8

Gerrymandering is unfixable in the absence of good faith. Independent redistricting committees work, but they have to be allowed to exist.

Check out the 538 special on gerrymandering. It's chock full of good data, but does nothing to change my opinion above.


Posted by: F | Link to this comment | 02- 6-18 10:27 AM
horizontal rule
9

I want to burn the boats so that the Californians don't try to get out of conquering the metaphorical Mexico.


Posted by: Moby Hick | Link to this comment | 02- 6-18 10:29 AM
horizontal rule
10

Every balance has its check.


Posted by: fake accent | Link to this comment | 02- 6-18 10:33 AM
horizontal rule
11

How much could Democrats gain if they went full gerrymander?


Posted by: Eggplant | Link to this comment | 02- 6-18 10:35 AM
horizontal rule
12

Gerrymandering is unfixable in the absence of good faith. Independent redistricting committees work, but they have to be allowed to exist.

Well sure. Master's tools won't dismantle his house and all that.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 10:35 AM
horizontal rule
13

My thinking is still pretty disorganized on this weekend, but one interesting detail:
Poli sci people are divided on whether or not the 1982 VRA updates about creating minority-majority districts when possible benefited Democrats or Republicans, but one presenter made a pretty convincing statistical case that the 1990-census redistricting that created min-maj districts cost Ds about 10% of their elections. The idea is that the surrounding districts all took a hit when you created the min-maj district - they went from electing a white, moderately liberal guy to a white, moderately conservative guy.

Absolutely min-maj districts should be intentionally created, but a lot of redistricting goals are in tension with each other.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 10:41 AM
horizontal rule
14

And when it comes right down to it, single member districts of arbitrary size (i.e. US pop/435) and constantly changing shape make little sense as a form or representation. Define a polity and figure out how many representatives they should get (i.e. multi-member districts).


Posted by: F | Link to this comment | 02- 6-18 10:43 AM
horizontal rule
15

If you wouldn't be happy to make your congressional district into a county or city boundary, it makes no sense.


Posted by: F | Link to this comment | 02- 6-18 10:45 AM
horizontal rule
16

Absolutely min-maj districts should be intentionally created

I'm pulling out vague memories of an election law class I took in law school many many many years ago, but this actually isn't obvious. The VRA mandate, and what makes sense, is avoiding the dilution of minority voting power (don't hold me to the wording). Min-maj districts are a technique for that, but they're not necessarily the best one, and may in fact be counterproductive -- there's no independent value in having some districts with whites in the minority, unless that's effective in preserving minority voting power.


Posted by: LizardBreath | Link to this comment | 02- 6-18 10:45 AM
horizontal rule
17

Right - presumably the goal is proportional representation. To both 15/16. When I said that we must create min-maj districts, I was thinking hyper-concretely about the limited methods on the table in 2018. And it's still debatable.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 10:47 AM
horizontal rule
18

Actually here's what I meant: the argument about bleaching surrounding districts is not compelling enough (to me) to negate the existence of the minority-majority district.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 10:50 AM
horizontal rule
19

Can someone please explain, as if to a hamster, what a min-maj district is?


Posted by: Mossy Character | Link to this comment | 02- 6-18 10:50 AM
horizontal rule
20

A district with boundaries set such that minorities are a majority of the voters.


Posted by: LizardBreath | Link to this comment | 02- 6-18 10:52 AM
horizontal rule
21

A district which has been drawn so that a majority of the voters are black or Hispanic or whatever minority is otherwise disenfranchised in an at-large election.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 10:52 AM
horizontal rule
22

Shut up.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 10:53 AM
horizontal rule
23

Like for instance, if in some sentences in this conversation "minority" will refer to a minority of voters, and in others to POC, could the latter case be referred to simply as POC?


Posted by: Mossy Character | Link to this comment | 02- 6-18 10:53 AM
horizontal rule
24

Because 23 would make this hamster's brain hurt less.


Posted by: Mossy Character | Link to this comment | 02- 6-18 10:54 AM
horizontal rule
25

The idea is that if you've, say, got a city that's 80% white and 20% black, and 5 city council members, and voters have racially distinct preferences, you can split up the city two ways: One gives you 5 80-20 districts, and black voters don't elect anyone. The other gives you 4 all white districts and 1 all black district, and black voters have elected representation in proportion to their voting numbers.

That's how it's supposed to work, but in practice it can be counterproductive. If you pack black voters into a few districts with overwhelming black majorities, that can also dilute their voting power.


Posted by: LizardBreath | Link to this comment | 02- 6-18 10:55 AM
horizontal rule
26

23: Minority-majority is a Voting Rights Act term of art, so annoying to avoid. Now that you know what it means, it should all make sense.


Posted by: LizardBreath | Link to this comment | 02- 6-18 10:56 AM
horizontal rule
27

Guys, if you don't even mention how many voters you can fit into your cheek pouches, nobody is going to believe you're really talking to a hamster.


Posted by: Swope FM | Link to this comment | 02- 6-18 10:59 AM
horizontal rule
28

Just because lawyers are insane I'm not going to copy them.


Posted by: Mossy Character | Link to this comment | 02- 6-18 11:00 AM
horizontal rule
29

Anyway, race surely isn't the only thing gerrymanderers are looking at. The constitution lets the states draw borders however the hell they want, right?


Posted by: Mossy Character | Link to this comment | 02- 6-18 11:03 AM
horizontal rule
30

If states could really draw their borders however they want, I'm pretty sure that little pokey bit of West Virginia would be part of Ohio or Pennsylvania.


Posted by: Moby Hick | Link to this comment | 02- 6-18 11:07 AM
horizontal rule
31

29: well, the big defense by Republicans in North Carolina, Texas, Florida, etc, is that these are partisan gerrymanders, not racial gerrymanders, and therefore not a violation of the VRA. Of course, from the POV of a functional democracy, it's a staggering defense. But here we are.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 11:10 AM
horizontal rule
32

So the VRA is the only legal limit on gerrymandering? Meaning you need another amendment to deal with it. Meaning you're not going to deal with it.


Posted by: Mossy Character | Link to this comment | 02- 6-18 11:16 AM
horizontal rule
33

Something, something, Baker v. Carr, something, something, one person, one vote.


Posted by: Moby Hick | Link to this comment | 02- 6-18 11:19 AM
horizontal rule
34

32: It's the only statutory limit on gerrymandering. All decent people are waiting with bated breath for the pending decision from the SC in the Wisconsin gerrymandering case, where Justice Kennedy may decide for the first time that an unreasonable level of partisan gerrymandering is unconstitutional.


Posted by: LizardBreath | Link to this comment | 02- 6-18 11:21 AM
horizontal rule
35

32: Well, it's tied up in courts right now. They'll rule against the gerrymander and the states won't do anything anyway.

Some states have additional anti-gerrymandering legislation - nothing is preventing the rest of the states from doing so, besides the fact that the elected officials don't want to and have drawn their own districts so nothing will ever change.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 11:22 AM
horizontal rule
36

I think Utah - or maybe Arizona? - has something on their books about drawing districts to make them competitive if possible. Which is also a good goal, but again is in tension with other good goals.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 11:23 AM
horizontal rule
37

25: a couple of points: I don't think there is a single other democratic country in the world that draws its electoral boundaries like this;

and is it really true to say that POC voters in the first case don't elect anyone?
It's true that there is no elected representative with a majority-POC district but that isn't the same thing. It also seems to be based on an assumption that no white voters will ever vote for a POC candidate.
But say you have a district with 40% white Whigs, 40% white Federalists, and 20% black voters. We will assume racial preference in voting; as in, the white Whigs ideally want a white Whig congressman, and the same with the Federalists, but if they can't get one they will prefer someone of their own party to someone from the other party. The black voters most of all want a black congressman.
The best strategy is to run a black guy, because he'll win 60-40. Most of his votes will have come from white voters, but is it really true to say in that case the black voters didn't elect him? He wouldn't have won if not for the black vote.
And this isn't just a weird hypothetical - it's a commonplace of reporting on US elections that, in a lot of cases, Democrats wouldn't win without the black vote.

The current system of majority-minority districts doesn't seem to be working very well to elect POC representatives judging by a comparison of the House to the actual US population.


Posted by: ajay | Link to this comment | 02- 6-18 11:35 AM
horizontal rule
38

Most of the current math work is being done where you take a random sampling of districtings and use features of those to analyze whether a particular Republican gerrymander could have happened by chance.

One unexpected result: most maps do not produce proportional representation on average. It's totally dependent on how the minority population is clustered, but even in the best possible circumstance, the average map is an under-representation. Ie, the example that I was walked through, in a fairly "good" distribution of 40% minority, the average maps tend to produce 20-25% minority representation.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 11:36 AM
horizontal rule
39

and is it really true to say that POC voters in the first case don't elect anyone?
It's true that there is no elected representative with a majority-POC district but that isn't the same thing. It also seems to be based on an assumption that no white voters will ever vote for a POC candidate.

Well, you need data to demonstrate the degree of RPV - racially polarized data. Some places are extremely racially polarized, others less so.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 11:38 AM
horizontal rule
40

Most countries with single-member constituencies (UK) have had them neutrally drawn for ages without much controversy, no? It's perfectly technically feasible, though when you have multiple priorities like compactness, PoC representation, competitive seats, matching existing communities, etc., you're bound to lose out on some of them.

(Interesting twist when I looked into the details in India: "The government had suspended delimitation in 1976 until after the 2001 census so that states' family planning programs would not affect their political representation in the Lok Sabha. This had led to wide discrepancies in the size of constituencies, with the largest having over three million electors, and the smallest less than 50,000.")


Posted by: Minivet | Link to this comment | 02- 6-18 11:39 AM
horizontal rule
41

And this isn't just a weird hypothetical - it's a commonplace of reporting on US elections that, in a lot of cases, Democrats wouldn't win without the black vote.

Another big takeaway I had is that absolutely nothing should be generalized beyond the parameters of the situation. Math people want to make all these broad theorems, and what's useful is to collect local data about how officials are actually elected, and use it to craft locally fair districts.

For example, voter turnout varies wildly. If you're assuming you need 60% black voters to create an African American opportunity district, you're unintentionally packing a lot of black people into the district and wasting their votes, because in a lot of places, 45% will create an opportunity district due to good turnout.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 11:41 AM
horizontal rule
42

It's totally dependent on how the minority population is clustered

Which is itself non-random, given the urban/rural divide. Population not being uniformly distributed is why single-member districts would tend to have unrepresentative (read reactionary) outcomes even if you eliminated gerrymandering entirely.


Posted by: Minivet | Link to this comment | 02- 6-18 11:41 AM
horizontal rule
43

37

Yakima, WA had a similar problem recently where all councilmembers were elected city-wide, which meant that the substantial Hispanic minority was shut out of city government. Link.


Posted by: F | Link to this comment | 02- 6-18 11:43 AM
horizontal rule
44

I'm not clear what you mean in the last sentence of 42.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 11:43 AM
horizontal rule
45

They'll rule against the gerrymander and the states won't do anything anyway.

This would be a bad outcome from a legitimacy of government perspective.


Posted by: | Link to this comment | 02- 6-18 11:43 AM
horizontal rule
46

37: Things are different now because they have gotten so polarized, but at the time the VRA was done, it certainly would have been the case in much of America that majorities of white people in both both parties would not have voted for a black person even if that meant switching parties. That's why they all switched parties.


Posted by: Moby Hick | Link to this comment | 02- 6-18 11:43 AM
horizontal rule
47

All of 538s attempts to create "fair" districts favored Republicans.


Posted by: F | Link to this comment | 02- 6-18 11:46 AM
horizontal rule
48

37: It all depends. Exactly the situation I'm describing: a vast majority of the white voters are racists and are voting for white supremacist candidates, black voters are voting for candidates opposed to white supremacy, who may or may not be black themselves, was a non-controversial description of big parts of the South at the time of passage of the VRA. And if that's what's going on, maj-min districts help, along the lines of my 25.

Now, things are subtler, but that sort of thing is still happening. Whether maj-min districts are still a good idea is arguable, but it comes down to the specific dynamic in the particular area.


Posted by: LizardBreath | Link to this comment | 02- 6-18 11:47 AM
horizontal rule
49

Or what heebie said in 41.


Posted by: LizardBreath | Link to this comment | 02- 6-18 11:48 AM
horizontal rule
50

47: Which is not the same as saying the current gerrymandering is anything close to "fair".


Posted by: Moby Hick | Link to this comment | 02- 6-18 11:48 AM
horizontal rule
51

From the same link, maximizing the number of min-maj districts did not significantly change the partisan advantage compared to other criteria.


Posted by: F | Link to this comment | 02- 6-18 11:49 AM
horizontal rule
52

50
Totally in agreement.


Posted by: F | Link to this comment | 02- 6-18 11:50 AM
horizontal rule
53

The current maps generate ~15 extra safe R seats.


Posted by: F | Link to this comment | 02- 6-18 11:51 AM
horizontal rule
54

44: If the Democratic areas are 80% Democratic while the Republican areas are 60% Republican, any geographic map you could draw will disadvantage Democrats.


Posted by: Minivet | Link to this comment | 02- 6-18 11:52 AM
horizontal rule
55

30

Panhandles are the worst from a rational boundary perspective. Florida would be a different state politically without the Redneck Riviera.


Posted by: F | Link to this comment | 02- 6-18 11:54 AM
horizontal rule
56

Two districts were recently thrown out in Texas, one of which is my district. The Republican legislature claimed they'd swapped one Hispanic opportunity district for another, so the net was fair. In reality, they'd taken a very highly RPV region and lumped them in with a mostly white district, 23, in order to protect one guy, Will Hurd. In exchange, they'd gerrymandered Lloyd Doggett's district, 35, and said it was now a Hispanic opportunity district, although it has a much lower degree of RPV. The courts basically said that the hispanic population of Nueces county, at the bottom of Hurd's district, had been disenfranchised and that different communities are not interchangeable.

Does everybody see how much I learned? Is everyone clear on this?


Posted by: heebie-geebie | Link to this comment | 02- 6-18 11:55 AM
horizontal rule
57

54: It'd depend on the specifics, but I bet you could gerrymander any map there is to get to a level of partisan advantage that corresponds to overall partisan identification (that is, something 'fair'). It'd just involve being willing to tolerate non-compact districts.


Posted by: LizardBreath | Link to this comment | 02- 6-18 11:55 AM
horizontal rule
58

54: I think you're referring to the idea that Democrats self-pack in cities, and this hurts them? This is actually not true (that it hurts them).


Posted by: heebie-geebie | Link to this comment | 02- 6-18 11:56 AM
horizontal rule
59

56: Scribbling quickly. "Texas has at least two different non-exchangeable types of Hispanic people."


Posted by: Moby Hick | Link to this comment | 02- 6-18 11:57 AM
horizontal rule
60

Yep. Check out the long skinny districts.


Posted by: F | Link to this comment | 02- 6-18 11:58 AM
horizontal rule
61

That is fun.


Posted by: Moby Hick | Link to this comment | 02- 6-18 12:07 PM
horizontal rule
62

58: Source? The math is persuasive, but I can imagine it might not match the reality. I remember thinking this was also the explanation of the Electoral College consistently swinging Republican with random state boundaries (and more than just the extra 2 votes per state could account for).


Posted by: Minivet | Link to this comment | 02- 6-18 12:12 PM
horizontal rule
63

57: Yes, but it's still difficult and less guaranteed, especially with all the other priorities as we note.

An interesting facet of the current debate is that it seems to be becoming more of a bedrock value, in the left of center, at least, that if a legislature does not on average reflect voter proportionalities there's something illegitimate about it. (Nobody I'm aware of on the left is saying, look, voters choose individuals to represent their communities, if that happens to be a different mix of ideologies or parties that's par for the course.) The next step following this principle is, why not just switch to PR or MMP instead of juryrigging the district system to match that outcome?


Posted by: Minivet | Link to this comment | 02- 6-18 12:17 PM
horizontal rule
64

63 gets it right. No matter how much Politics Wolf tweets about it, I will never understand what a "fair map" is. Maximizing competitiveness mandates incoherent and weirdly shaped districts.

And is maximizing competitiveness good anyway? If all 18 districts in a state are split 50-50, then what - every year either Democrats win all 18 or Republicans win all 18 depending on the President's approval ratings? Nobody wants that.

So we don't want to maximize competitiveness at all, maybe we want to reflect the state's political demographics, by jerry-rigging it into 10 safe Democratic districts and 8 safe Republican districts. And then there's no reason to vote except in the primary. Nobody wants that.

And how do we explain that it's bad to gerrymander in any way, except that it's mandatory and also good to gerrymander on the basis of race?

What we want turns out to be proportional representation.


Posted by: Cryptic ned | Link to this comment | 02- 6-18 12:26 PM
horizontal rule
65

62: M/oo/n D/uchin is convinced, but I haven't read a paper on it. And when they actually take samplings of randomly drawn districts, it strongly seems to bear out the idea that clustered minorities are more easily represented than diffused minorities. And in fact, Massachusetts is 30% Republican, but they are so evenly distributed that you actually cannot draw a district that they'd be a majority of, even if you suspend compactness, etc.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 12:27 PM
horizontal rule
66

And is maximizing competitiveness good anyway?

It's not the end-all be-all, but the argument goes that when you get runaway skewed noncompetitive districts, the only election becomes the primary, and you get wildly extreme Tea Party members in power.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 12:30 PM
horizontal rule
67

What we want turns out to be proportional representation.

But yes, I agree with where you're going with this.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 12:31 PM
horizontal rule
68

I just like to comment before I get to the end of comments that I'm responding to, because it's very, very important that I feel very important with my newfound knowledge.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 12:32 PM
horizontal rule
69

The consensus seems to be that you want every district to be fine-tuned to be about 60-40, so if the majority party nominates an unpopular maniac they might possibly lose. And you want every state to be fine-tuned so these 60-40 districts are jerry-rigged to produce proportional representation. Except that most states are too small to do that so either you have to have 1 district containing the major city where the Democrats live and then 2 or 3 or 4 incredibly Republican districts, or you have to give up and have no Democrats. (Wait... why does a district have to be limited to the boundaries of 1 state?) Also the majority-minority districts have totally different rules and are required to be de facto guaranteed Democratic, so you have to fit that in.

With precise knowledge of where all the voters are nowadays, and computers drawing the maps, and the US having exactly two parties so there is no chance of unpredictable interactions between the multiple parties like how the SNP collapsing suddenly led to 18 more Tory seats in Scotland, but the UKIP collapsing did NOT lead to more Tory seats... the whole thing is impossible.


Posted by: Cryptic ned | Link to this comment | 02- 6-18 12:41 PM
horizontal rule
70

Very impressive, heebie! Lots of knowing stuff!


Posted by: peep | Link to this comment | 02- 6-18 12:42 PM
horizontal rule
71

63 is sort of true, but this is a result of extreme partisan polarization which (people now forget) is really fairly new in the US. We've only had truly ideological parties, especially at the State/district representative level, since maybe 1994. That means that the concerns about fairness in representation by party (and the costs of extreme partisan gerrymandering AND the incentive to do it) are fairly different now than they were in the recent past.

I have a few responses to other stuff but this is sort of what I do for part of my day job. Heebie and LB make very good points. The 538 maps help show the different trade-offs in redistricting which are of course real, but are somewhat misleading, in that at the national level there is an inherent natural pro-Republican tilt because of state boundaries (there are a lot of essentially non-competitive Republican states). Small states and the need to represent single-states in the House make the analysis harder than it would be otherwise.

More generally the worries about intellectual complexity of gerrymandering, like in 64, are sort of real but also pretty misleading about what the stakes are in most current debate/action around gerrymandering, and can lead to a needless "everything is hopeless, throw up your hands." The action in gerrymandering, at the state and federal level, is all in taking larger, mostly competitive states and districts and making them mostly non-competitive. And that's also the main source of outrage -- when politically-split states like PA, MI, WI, or NC are set up to be entirely dominated by Republicans, people (rightly) feel disenfranchised. Solving THAT problem doesn't require PR, isn't particularly hard to solve conceptually, and doesn't require any particularly remarkable tradeoff -- it just requires not drawing extreme partisan districts. People wouldn't be particularly mad about a small but persistent Republican advantage based on compact boundaries. But the current system (in many states) goes so far beyond that.

At the local level, you can have lots of interesting intra-party fights about boundaries and preferences. Those issues are pretty complex and interesting. But that's also different than the concern with extreme gerrymandering in larger states, which is where the legal and political action is on this issue these days.


Posted by: Robert Halford | Link to this comment | 02- 6-18 12:43 PM
horizontal rule
72

The conference had a really big focus on local politics, which was great. The Yakima situation mentioned in 37 above is really, really common, and a great place to take action.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 12:51 PM
horizontal rule
73

And in fact is the case in Heebieville: the city council is composed of 7 at-large seats, and there is exactly 1 Latino member.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 12:53 PM
horizontal rule
74

70: thank god. I thought I was just showing off into the abyss.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 12:54 PM
horizontal rule
75

Many of the VRA cases arise from situations like Yakima, because a very common move in the South (and elsewhere, but largely the South) after 1965 was to merge local governance into multi-member districtwide voting so that the white machine could run its candidate and the 30% of the population that was black would have no representative at all. That's a very different problem than the extreme partisan gerrymanders at the state or federal congressional level. And most actual VRA litigation is about local-level politics.


Posted by: Robert Halford | Link to this comment | 02- 6-18 12:56 PM
horizontal rule
76

the city council is composed of 7 at-large seats

This is totally weird. We have that in my city. It just went from 5 Republicans to 5 Democrats. I don't get why ANYONE would design a system like that. If there aren't any districts anyway and no ability to say "this councilman is my councilman from my ward; the other 6 are not", then city council is where proportional representation would really work.


Posted by: Cryptic ned | Link to this comment | 02- 6-18 12:56 PM
horizontal rule
77

I don't get why ANYONE would design a system like that.

Mostly, so that a local machine that can reliably get 50%+ of the vote citywide can run its slate and win every time.


Posted by: Robert Halford | Link to this comment | 02- 6-18 12:58 PM
horizontal rule
78

city council is where proportional representation would really work.

In municipal government, party lines tend to be less relevant (even setting aside traditional nonpartisan status of these races). I like district-based ranked-choice voting in this case, since the voting system allows for preference-sorting of individual candidates while the districts facilitate minority representation.


Posted by: Minivet | Link to this comment | 02- 6-18 1:05 PM
horizontal rule
79

Looks like the local Republican machine forgot to change the rules before the city started growing again with overflow from nearby larger cities. Oh well!


Posted by: Cryptic ned | Link to this comment | 02- 6-18 1:06 PM
horizontal rule
80

I think there used to be a lot more diversity in city structure - in the 1910s-20s Oakland seems to have elected people to specific portfolios, like Commissioner of Streets, Commissioner of Revenue and Finance, etc. - but researching this I got hung up at a 1924 city directory that devotes six pages to names and addresses of "Secret Societies".


Posted by: Minivet | Link to this comment | 02- 6-18 1:10 PM
horizontal rule
81

OT: Falcon Heavy is lifting off in about 25 minutes. http://www.spacex.com/webcast
First private-sector spaceflight beyond Earth orbit and the first to carry a $200k sports car as payload.


Posted by: ajay | Link to this comment | 02- 6-18 1:21 PM
horizontal rule
82

How many public-sector spaceflights have carried a $200,000 sports car?


Posted by: Moby Hick | Link to this comment | 02- 6-18 1:24 PM
horizontal rule
83

81 A really stupid and vain stunt.


Posted by: Barry Freed | Link to this comment | 02- 6-18 1:26 PM
horizontal rule
84

82: we don't know. There have been a lot of classified payloads.


Posted by: ajay | Link to this comment | 02- 6-18 1:33 PM
horizontal rule
85

And Whitey's car is on the moon.


Posted by: Cryptic ned | Link to this comment | 02- 6-18 1:39 PM
horizontal rule
86

That's why you tip the valet.


Posted by: Moby Hick | Link to this comment | 02- 6-18 1:47 PM
horizontal rule
87

|| Relevant to LB's interests? Tonga in the Olympics |>


Posted by: Mooseking | Link to this comment | 02- 6-18 1:55 PM
horizontal rule
88

On the other end of the spectrum - or really multiple spectra.


Posted by: Minivet | Link to this comment | 02- 6-18 1:56 PM
horizontal rule
89

Quote from the link in 88:

In 2014, Hughes blasted himself 1,374 feet in a $20,000 homemade rocket -- a feat he claims to have achieved without the help of science.

Thanks for keeping us out of that.


Posted by: Opinionated Science | Link to this comment | 02- 6-18 2:00 PM
horizontal rule
90

88: He's a "recent convert" to the flat earth theory, which I'm pretty sure means that he's happy to say stupid things about the shape of the world if flat earthers are willing to give him 8 grand towards his rocketry hobby.


Posted by: Seeds | Link to this comment | 02- 6-18 2:06 PM
horizontal rule
91

90: I skimmed through the video and it's six minutes of him talking with an apparent rocket setup perhaps 100 feet behind him. Ideally he's just defrauding them with a fake setup.


Posted by: Minivet | Link to this comment | 02- 6-18 2:13 PM
horizontal rule
92

87: Hrmphf. Colonizing Tongan bastards -- Samoa will never forget when they threw off the Tongan yoke!


Posted by: LizardBreath | Link to this comment | 02- 6-18 2:19 PM
horizontal rule
93

81. It worked. Pretty impressive.


Posted by: DaveLMA | Link to this comment | 02- 6-18 2:25 PM
horizontal rule
94

And when they actually take samplings of randomly drawn districts, it strongly seems to bear out the idea that clustered minorities are more easily represented than diffused minorities.

But is that about giving them any representation at all, or about representing them in proportions similar to their numbers?

And in fact, Massachusetts is 30% Republican, but they are so evenly distributed that you actually cannot draw a district that they'd be a majority of, even if you suspend compactness, etc.

I don't think this invalidates my 54 either - I think diffuseness only works as I describe when the sides are more even overall. As any majority (racial or partisan) increases, past a certain tipping point, random draw will start to favor them regardless. Similarly, in CA where we have a nonpartisan commission, Democrats got 62% of the House votes in 2016 but 39 of 53 seats (74%). (PR would take away this advantage - though it would also elevate Greens and Liberts.)


Posted by: Minivet | Link to this comment | 02- 6-18 2:33 PM
horizontal rule
95

auuugh my long answer just got deleted.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 2:59 PM
horizontal rule
96

augh I want it to be written but I don't want to rewrite it.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 3:00 PM
horizontal rule
97

what to do!


Posted by: heebie-geebie | Link to this comment | 02- 6-18 3:00 PM
horizontal rule
98

I assume that:

(a) it would be worse for Democrats in many states where they are in the minority to be evenly distributed throughout the state. For a random example, in Nebraska, a clustering around Omaha makes it easier to draw a Democratic district than it would be if the entire state was 35% Democratic.

(b) in close states (at a statewide level) a heavy packing of the Democratic vote into cities hurts Democrats. E,g,, if Wisconsin is 51% Democratic, but nearly all of these people live in inner-city Milwaukee (this isn't true, but go with it) that clustering is worse for Democrats than it would be to have the Democrats more evenly distributed throughout the State, since you can't draw more than one inner-city Milwaukee district.

Which effect predominates probably depends both on a particular State's geography and partisan balance. And you'd have to look at both effects and net them out to get a sense of the impact of clustering nationally. Is that right? (Genuine question, I don't have a very strong view about this).


Posted by: Robert Halford | Link to this comment | 02- 6-18 3:23 PM
horizontal rule
99

Dictate it?


Posted by: Cryptic ned | Link to this comment | 02- 6-18 3:23 PM
horizontal rule
100

Which effect predominates probably depends both on a particular State's geography and partisan balance.

I think that's right. Which means that I think proportional representation would be best, but short of that there isn't really such a thing as ideal districting, and the best we can hope for is to prohibit insanely one-sided districting.


Posted by: LizardBreath | Link to this comment | 02- 6-18 3:25 PM
horizontal rule
101

Ok, here's roughly the mini-talk I saw: Suppose your population is 40% green, 60% blue.
Case A: the 40% green are evenly distributed. (Massachusetts Republicans). Run the simulations to get a bunch of possible districtings and no district will ever be majority green.

Case B: the 40% green are streaky in their distribution. The simulations - I'm told - produce on average 20-25% representation for the green.

Case C: the 40% green are perfectly clustered in a city. Then it's relatively easy to get proportional representation for the greens.

Your 54: If the Democratic areas are 80% Democratic while the Republican areas are 60% Republican, any geographic map you could draw will disadvantage Democrats.

This is a case of B, not C. But a fair districting would certainly need to avoid districts entirely within Democratic areas - pie shaped wedges leaving the city and including some rural areas - so as not to pack the Democrats.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 3:25 PM
horizontal rule
102

I had to post that before it got erased, but I still have a question to ask back.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 3:26 PM
horizontal rule
103

I think I understand 54, but I don't see how that's quite what you mean by the last sentence of 42:

Population not being uniformly distributed is why single-member districts would tend to have unrepresentative (read reactionary) outcomes even if you eliminated gerrymandering entirely.

This seems correct - that this is a streaky scenario - but why reactionary?


Posted by: heebie-geebie | Link to this comment | 02- 6-18 3:28 PM
horizontal rule
104

Anything that overrepresents majority (that is, white) voters in the US is going to have right-wing outcomes. As a white person, I feel comfortable saying that the way my people vote is a scandal and we should be ashamed of ourselves.


Posted by: LizardBreath | Link to this comment | 02- 6-18 3:32 PM
horizontal rule
105

there isn't really such a thing as ideal districting, and the best we can hope for is to prohibit insanely one-sided districting.

Yes, this. And, to be clear, basically no one who thinks about this issue thinks that you can redistrict your way into true proportional representation (which may not even be desirable!) But you can prevent one party in baseline competitive areas from using redistricting to lock in extreme, durable majorities that take massive wave elections (or more) to undo. There's lots of work to be done on the latter issue even if we agree that we can't district-draw perfect proportional representation.


Posted by: Robert Halford | Link to this comment | 02- 6-18 3:34 PM
horizontal rule
106

98: a couple thoughts:
a. My impression is that it's easy for a (good-faith) redistricting to avoid the self-packing of the inner city. Compactness is really, really not a criterion that we should think very hard about - there are tons of good existing reasons to neglect it.

b. Which effect predominates probably depends both on a particular State's geography and partisan balance. And you'd have to look at both effects and net them out to get a sense of the impact of clustering nationally.

Yes to the first sentence, heavily. My impression of the second question is that it's meaningless to even look at this, because what you want is local data specific to the local election. That fair redistricting is really done in response to the specifics of that locale - what are the specific communities of interest, what are their turnouts, how racially polarized is the voting in this area, etc.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 3:37 PM
horizontal rule
107

Also, turnout increases when people don't feel like their vote is useless. So 40-45% is now considered a minority-opportunity district, whereas back in 1982, they thought it needed to be 50-60%.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 3:48 PM
horizontal rule
108

My takeaway is that the best you can do is ramp up the transparency and independent redistricting and public attention paid to these processes, and when it's done in good faith, it more or less works well enough.


Posted by: heebie-geebie | Link to this comment | 02- 6-18 3:50 PM
horizontal rule
109

what you want is local data specific to the local election. That fair redistricting is really done in response to the specifics of that locale - what are the specific communities of interest, what are their turnouts, how racially polarized is the voting in this area, etc.

Yes, of course you are right. It's often hard to figure out what is a real "cluster" that creates a necessary form of packing (based on being a community of interest) and what is an artifical cluster you can work around to avoid packing. And the size of the state and the number of representatives matters a lot for this -- for example, if you have three representatives in Nebraska, it may be hard to avoid having one representative for Omaha and two for the rest, creating clustering. If you have 30, assembling the relevant clusters and communities of interest looks pretty different.


Posted by: Robert Halford | Link to this comment | 02- 6-18 3:50 PM
horizontal rule
110

My takeaway is that the best you can do is ramp up the transparency and independent redistricting and public attention paid to these processes, and when it's done in good faith, it more or less works well enough.

Yes. And it doesn't even have to be "independent" -- just not wildly partisan.


Posted by: Robert Halford | Link to this comment | 02- 6-18 3:51 PM
horizontal rule
111

104 speaks for me.


Posted by: Minivet | Link to this comment | 02- 6-18 3:54 PM
horizontal rule
112

81: So is this actually reducing cost/payload, or is it Musk trying swing his dick around?


Posted by: Minivet | Link to this comment | 02- 6-18 3:57 PM
horizontal rule
113

104 should have started with "I'm not a racist but...."


Posted by: Moby Hick | Link to this comment | 02- 6-18 5:33 PM
horizontal rule
114

OT: He says he got the idea from France, but the Red Square vibe is screaming.


Posted by: Moby Hick | Link to this comment | 02- 6-18 5:45 PM
horizontal rule
115

It's hard to resist the conclusion that everyone is going doing all this shit because they finally have a good excuse (well, of course no self-respecting general would want a military parade, that'd be gross, but if the CiC demands it, well, I guess we have no choice, now, do we?).


Posted by: F | Link to this comment | 02- 6-18 5:55 PM
horizontal rule
116

You are so negative. Don't you understand its a "celebration of the men and women who give us freedom"?


Posted by: Spike | Link to this comment | 02- 6-18 5:56 PM
horizontal rule
117

Anyway, between Trump and Quentin Tarantino, if being a white man is graded on a curve, I could be a much bigger piece of shit than I have been and still pass.


Posted by: Moby Hick | Link to this comment | 02- 6-18 5:58 PM
horizontal rule
118

OT: Wynn stepping down from running his company.


Posted by: Moby Hick | Link to this comment | 02- 6-18 8:12 PM
horizontal rule
119

86 is outstanding. That kind of quiet and quick work, mostly done in the shadows, without any fanfare or foofaraw, is why Moby is the GOAT.


Posted by: Von Wafer | Link to this comment | 02- 6-18 9:05 PM
horizontal rule
120

I'm willing to give him a military parade so long as 30% of the women soldiers and 20% of the men are black. And 15% overall Hispanic. And maybe, just like at the Republican convention, the networks can spend their time finding participants of color.


Posted by: CharleyCarp | Link to this comment | 02- 7-18 12:52 AM
horizontal rule
121

That kind of quiet and quick work, mostly done in the shadows, without any fanfare or foofaraw, is why Moby is the GOAT.

He's the commenter Gotham needs, not the one it deserves.


Posted by: ajay | Link to this comment | 02- 7-18 3:51 AM
horizontal rule
122

I got a snow day, because that's what I deserve.


Posted by: Moby Hick | Link to this comment | 02- 7-18 6:46 AM
horizontal rule
123

It must be for me because there doesn't seem to be any snow.


Posted by: Moby Hick | Link to this comment | 02- 7-18 7:06 AM
horizontal rule
124

Oh man I actually gerrymandered* my city's council boundaries once (within the boundaries of the rules here) and it was surprisingly boring. But this old dude came round to my hosue with a pocket calculator and a really big map and some street directories & we did it all on the kitchen table and it was pretty low-tech but it worked. None of our maps were crazy through, it was just that where you two possible plausible options we always wanted the one that helped us, basically.

Anyway my city used to have an at-large system but then we won back in the 70's and drew up districts so we wouldn't get wiped out in three year's time and it mostly worked.

We also had a city-wide at-large vote for the regional authority and our machine took 3/4 of the city positions at a canter, which just goes to show that organisation is really powerful.

(* was our side's lead on the redistricting and got 95% of what we wanted, and we won the municipal election following).


Posted by: Norm Kirk | Link to this comment | 02- 7-18 5:48 PM
horizontal rule