The Anti-Gerrymander

[Ug make hammer]

"RAG" in the tables means "Reverse Anti-Gerrymander" before I decided that was just confusing. "Reverse" was meant to capture that anti-gerrymandering is only done to maximize the number of competitive districts: setting aside voters to competitive districts comes FIRST, anti-gerrymandering is what happens to the remainder.

Another thing I didn't mention is that a gerrymander/anti-gerrymander algorithm, unchained from previous districts, can be expected to redress the injustice of being anti-gerrymandered (ie, vote dilution) for at least some voters. Redistricting could even be made a 5-yearly thing, with every second redistricting being driven by the Census.

Perhaps voters could be rotated (one decade at a time) between tossup and anti-gerrymandered districts. This really wouldn't work in 2 district, highly partisan states, however it might work well in medium-sized highly divided states. It still wouldn't require human judgement: just another trunk on the elephant in the room which is the districting algorithm.

 

[TU Curmudgeon]

Uh-uh. Don't start talking about the Presidency.



Irrelevant.



Because it's the thread subject, we're talking about the US House, if you don't mind. Because the US is not even attempting more than two parties, their House is more proportionate to the vote than the Canadian Parliament is.



Except that above, you seem to be talking about ANYTHING BUT the US House.

Well, if you add up the Presidential, Senatorial, and Representative votes that were cast in selecting the holders of those offices, then what you are telling everyone is that the US should have Democrat for its President and Vice-president (separate electoral rules constitutionally apply here), and should have a Senate and House that are 51.47 Republican and 48.53% Democrat. That works out to 51 Senator[R]s and 49 Senator [D]s as well as 224 Representative[R]s and 211 Representative [D]s.

That, of course, is a recipe for total legislative gridlock.

Quite frankly I don't actually give a damn (well, OK, I would prefer that "MY Guys" win - but I don't get all bent out of shape if they don't) about the outcome of elections that are "free, fair, open, and honest". What appears to be the consensus in the US is that the electorate doesn't give a damn about "free, fair, open, and honest" elections as long as their preferred candidate always wins.

You appear to be advocating for a slight variation on the electoral mechanics to ensure that the candidates that you want to win always do so.

HOWEVER, I do credit you for the thought that has gone into your proposal.

 

[CaughtInThe]

Check this out (I'm gonna pick on NC once again).


After the 2012 NC election...
99 Republicans
71 Democrats

House total votes...
2.1 million votes for Republicans
2.2 million votes for Democrats




House...

 

[TU Curmudgeon]

Check this out (I'm gonna pick on NC once again).


After the 2012 NC election...
99 Republicans
71 Democrats

House total votes...
2.1 million votes for Republicans
2.2 million votes for Democrats




House...

View attachment 67379490

Yes, a very good example of making sure that the election mechanics "return the right people to power" REGARDLESS of the will of the people.

It's 'The ***A*M*E*R*I*C*A*N*** Way'

 

[Ug make hammer]

Well, if you add up the Presidential, Senatorial, and Representative votes that were cast in selecting the holders of those offices, then what you are telling everyone is that the US should have Democrat for its President and Vice-president (separate electoral rules constitutionally apply here), and should have a Senate and House that are 51.47 Republican and 48.53% Democrat. That works out to 51 Senator[R]s and 49 Senator [D]s as well as 224 Representative[R]s and 211 Representative [D]s.

That, of course, is a recipe for total legislative gridlock.

For at most two years. I didn't have 2022's results to use (obviously) however if there is a national swing to Republicans as expected, they would get a fat majority in the House. My system is MORE not LESS sensitive to national swings, because there are many more districts which either side could win.

Quite frankly I don't actually give a damn (well, OK, I would prefer that "MY Guys" win - but I don't get all bent out of shape if they don't) about the outcome of elections that are "free, fair, open, and honest". What appears to be the consensus in the US is that the electorate doesn't give a damn about "free, fair, open, and honest" elections as long as their preferred candidate always wins.

You appear to be advocating for a slight variation on the electoral mechanics to ensure that the candidates that you want to win always do so.

Not at all. There is a problem I admit, that I can't explain in fairness terms WHY my system works, besides pointing out that its equilibrium point is proportional in almost every state. I can't think of a way to model election outcomes without masses of data and at least one state 'anti-gerrymandered'.

I guess the next step is to try to get hold of the gerrymandering software state governments use. I'm pretty sure it exists, because some states clearly lack the expertise to even know the purpose of gerrymandering — they seem to think it is to reduce work campaigning, for their own majority — and anyone that stupid couldn't possibly be drawing maps by hand.

HOWEVER, I do credit you for the thought that has gone into your proposal.

Not that much actually. It began as a simple idea, and despite the tables it remains quite simple: gerrymandering against the majority, in each state, creates many more competitive seats. That it also seems to produce more proportionate state contingents to the House is a lucky accident I'm still puzzled by.

 

[Ug make hammer]

Yes, a very good example of making sure that the election mechanics "return the right people to power" REGARDLESS of the will of the people.

It's 'The ***A*M*E*R*I*C*A*N*** Way'

The Canadian way is so much better ...

2021​	​	Vote​	Seats​	S/V​
Liberal​	​	32.62%​	47.34%​	1.45​
Conservative​	​	33.74%​	35.21%​	1.04​
Bloc Q​	​	7.64%​	9.47%​	1.24​
New Dem​	​	17.82%​	7.40%​	0.42​
Green​	​	2.33%​	0.89%​	0.38​
People’s​	​	4.94%​	0.00%​	0​

 

[Ug make hammer]

Check this out (I'm gonna pick on NC once again).


After the 2012 NC election...
99 Republicans
71 Democrats

House total votes...
2.1 million votes for Republicans
2.2 million votes for Democrats




House...

View attachment 67379490

That map was successfully challenged in court, and a new map was used for the next (2014) election.

With incumbent advantage in most seats, and a significant statewide swing (53.22% to R's), Republicans held 10 of the 13 seats.

I'm just seeking to make the point that unfair maps don't necessarily LOOK unfair, and winning or losing the popular vote is just one factor in whether the results are proportionate.

We see 10/3 as mostly fair when the winning side has a 6.6% popular vote advantage, ONLY because we're accustomed to thinking in Winner Take All terms. It's actually not that fair at all: the winner should have 6-7% more seats, not 77% more!

 

[TU Curmudgeon]

The Canadian way is so much better ...

2021​	​	Vote​	Seats​	S/V​
Liberal​	​	32.62%​	47.34%​	1.45​
Conservative​	​	33.74%​	35.21%​	1.04​
Bloc Q​	​	7.64%​	9.47%​	1.24​
New Dem​	​	17.82%​	7.40%​	0.42​
Green​	​	2.33%​	0.89%​	0.38​
People’s​	​	4.94%​	0.00%​	0​

As I said previously, the elections (as are the elections in the US) are NOT decided by having everyone vote for a political party and then taking the national total in order to divide up the seats.

The same result could well happen with your "every district is as close to a 50/50 split between Republican voters and Democrat voters.

As an example, if the Republicans received 50.001% of the vote in every electoral district while the Democrats received 49.999% of the vote in every electoral district, then the Republicans would win 100% of the seats in the House of Representatives and 100% of the seats then being contested in the Senate. If the Senate had started out being 50 Republicans and 50 Democrats, it would end up being around somewhere between 50 Republicans and 50 Democrats and 80 Republicans and 20 Democrats (depending on whose Senate seat was up for grabs).

Not only that, but you would end up with almost 100% of the electoral districts in the US looking like the classic "salamander" that earned "Gerrymandering" its name.

 

[Ug make hammer]

As I said previously, the elections (as are the elections in the US) are NOT decided by having everyone vote for a political party and then taking the national total in order to divide up the seats.

They are in proportional representation systems. That those also get third parties elected, and quite often so many of them that government has to be formed by a coalition, is more relevant to Canada's broken system than to the US's two-party system.

The same result could well happen with your "every district is as close to a 50/50 split between Republican voters and Democrat voters.

Except that I have said from the start that compactness etc built into the districting algorithm, would be traded off against margin in the winnable seats. I was a bit unclear about whether "winnable" districts should be in a 5% or 10% range, but that's only because I don't know what gerrymandering algorithms are capable of turning out. To be conservative let's say 10% (eg D45/R55).

As an example, if the Republicans received 50.001% of the vote in every electoral district while the Democrats received 49.999% of the vote in every electoral district, then the Republicans would win 100% of the seats in the House of Representatives and 100% of the seats then being contested in the Senate.

Why do you keep bringing up the Senate? My proposal has nothing to do with the Senate, and it would be grossly unconstitutional to try to "redistrict" whole States for the Senate.

If the Senate had started out being 50 Republicans and 50 Democrats, it would end up being around somewhere between 50 Republicans and 50 Democrats and 80 Republicans and 20 Democrats (depending on whose Senate seat was up for grabs).

Not only that, but you would end up with almost 100% of the electoral districts in the US looking like the classic "salamander" that earned "Gerrymandering" its name.

I didn't think you were so shallow as to care what maps look like. See the example above, where the visible salamanders of 2012 NC were eliminated by a court order, but Republicans didn't lose a single district in 2014.

Salamanders are the visible evidence of a gerrymander, but that doesn't make them wrong. You have to appeal to "communities of interest" or similar to justify a gerrymander, but to justify an anti-gerrymander you only have to appeal to equal voting power in as many cases as possible.

Finally let me say that the states which are close (over the preceding decade) barely need anti-gerrymandering at all. North Carolina for instance would have 12 winnable districts with only one (moderately) anti-gerrymandered district. AND, the small percentage of people whose votes are drowned out or diluted (the one AG district) would very likely be in winnable districts the next decade. Stability of districts suits only incumbents, I would make no allowance for it at all.