Competitiveness

[Sol]

Competitive seats = Arizona where 1 party wins all the competitive seats.

The Michigan Republican party also drew many competitive seats. Only 2 of them were won by McCain. Yet there is plenty of complaining. Curious.

McCain utterly collapsed in Michigan in 2008, presidential numbers from that year are unrepresentative of congressional voting.

Of course, that is the entire point. It would be folly to draw districts based on some arbitrary 'competitive' metric. As we have seen, when that 'competitive' metric does not provide the outcomes desired by a political party, that political party simply discards that metric (but only when it suits them).

Another example is, of course, New Jersey, where competitive seats were drawn and 1 party won all the competitive seats.

Another example is, of course, PA-06 (R+1), PA-07 (EVEN), PA-08 (D+1) PVIs based on the 2004 and 2008 elections.  I suspect these competitive seats are not what the original poster had in mind when he talked about creating competitive seats!

Ideally, this competitiveness metric would be based on the average of various national and statewide elections. For the latter, weird overperformances would not be considered- like Sebelius winning the KS gubernatorial. All the elections from the past decade would be considered, but recent ones would be more heavily weighted.

It should also be noted that I would not require AZ to draw competitive districts , although I also know it is required by the state constitution.


My apologies if this is kind of hard to understand, but:

I believe there has been some discussion of a point based system? This is my view of how competitiveness might work for such a system:

Scale of 1-10, More points=bad

VRA violation= +10 (automatic disqualification)
Contiguity violation(Poor road connections, counties barely bordering each other, geographical features) =+4 for every split
Large UCC violation= +3 for every split
Large County splits= +1 for every such split (+.5 within UCCs)
# of High erosity districts= +.5 for each high-erosity district
# of Microchops= +.5 for every split (+.1 in UCCs)

The points and things to be penalized for don't matter, what matters is this:

CoI and Competitiveness interests can negate points, except for on the VRA.

For example, a proposed map of VA has a Shenandoah valley based district. This district is very erose, incurring a +.5 penalty. However, because the Shenandoah valley is a CoI, the +.5 does not go into effect.

Competitiveness can also do the same, although it would only be implemented as a factor in more inelastic states. Competitiveness also subtracts points, in addition to negating them. For example, in Alabama, it would not only negate the violation of the Mobile and Birmingham/Montgomery UCC, but also subtract .5 points from the total for creating a competitive district.

Naturally, the numbers would need to be thought through more, but do you all like the concept of point-negating with CoI and point-subtracting with Competitiveness?

 

[krazen1211]

The VRA is certainly needed, IMO, particularly for Latinos, who can be easily diluted by non-hispanic whites due to their low voting rates.

I agree with you that Clyburn's district kind of screws SC up.

In the case of Thornburg v Gingles, it was North Carolina white Democrats who shattered various black communities.

Which is why black voters are such fans of the Republican maps, where they'll be in minority for the foreseeable future...

I am grossly confused as to what you mean by 'fans'. But, in any case, the Court in Gingles noted that White Democrats had limited blacks to 2-4% of the seats in the 1970s redistricting.  Quite a small minority.

So you're not smart enough to detect sarcasm?

I guess we have something new to add to the Miles personal insult barrage.

[muon2]

Competitive seats = Arizona where 1 party wins all the competitive seats.

The Michigan Republican party also drew many competitive seats. Only 2 of them were won by McCain. Yet there is plenty of complaining. Curious.

McCain utterly collapsed in Michigan in 2008, presidential numbers from that year are unrepresentative of congressional voting.

Of course, that is the entire point. It would be folly to draw districts based on some arbitrary 'competitive' metric. As we have seen, when that 'competitive' metric does not provide the outcomes desired by a political party, that political party simply discards that metric (but only when it suits them).

Another example is, of course, New Jersey, where competitive seats were drawn and 1 party won all the competitive seats.

Another example is, of course, PA-06 (R+1), PA-07 (EVEN), PA-08 (D+1) PVIs based on the 2004 and 2008 elections.  I suspect these competitive seats are not what the original poster had in mind when he talked about creating competitive seats!

Ideally, this competitiveness metric would be based on the average of various national and statewide elections. For the latter, weird overperformances would not be considered- like Sebelius winning the KS gubernatorial. All the elections from the past decade would be considered, but recent ones would be more heavily weighted.

It should also be noted that I would not require AZ to draw competitive districts , although I also know it is required by the state constitution.


My apologies if this is kind of hard to understand, but:

I believe there has been some discussion of a point based system? This is my view of how competitiveness might work for such a system:

Scale of 1-10, More points=bad

VRA violation= +10 (automatic disqualification)
Contiguity violation(Poor road connections, counties barely bordering each other, geographical features) =+4 for every split
Large UCC violation= +3 for every split
Large County splits= +1 for every such split (+.5 within UCCs)
# of High erosity districts= +.5 for each high-erosity district
# of Microchops= +.5 for every split (+.1 in UCCs)

The points and things to be penalized for don't matter, what matters is this:

CoI and Competitiveness interests can negate points, except for on the VRA.

For example, a proposed map of VA has a Shenandoah valley based district. This district is very erose, incurring a +.5 penalty. However, because the Shenandoah valley is a CoI, the +.5 does not go into effect.

Competitiveness can also do the same, although it would only be implemented as a factor in more inelastic states. Competitiveness also subtracts points, in addition to negating them. For example, in Alabama, it would not only negate the violation of the Mobile and Birmingham/Montgomery UCC, but also subtract .5 points from the total for creating a competitive district.

Naturally, the numbers would need to be thought through more, but do you all like the concept of point-negating with CoI and point-subtracting with Competitiveness?

 

Points are fine, but from what I've seen it is virtually impossible to get agreement on the relative weight of the points. If a few of us on the site can't get there, I doubt a legislative body would either.

Instead I think the better tactic is to keep each category separate with its own score. This results in multiple top plans, but that's not a bad outcome. It allows a final decision-making body the chance to weigh local subjective considerations when selecting from a set of fair plans.

[barfbag]

Gerrymandering has never been a problem and shouldn't be now.

[The world will shine with light in our nightmare]

Gerrymandering has never been a problem and shouldn't be now.

You're joking, right?

[Miles]

I guess we have something new to add to the Miles personal insult barrage.

Well, sometimes I fell nice enough to give you the negative attention you come here to solicit; you should be thanking me.

[krazen1211]

Gerrymandering has never been a problem and shouldn't be now.

Yes, of course. Sandra Day O'connor noted that gerrymandering was self-limiting.

[Sol]

Competitive seats = Arizona where 1 party wins all the competitive seats.

The Michigan Republican party also drew many competitive seats. Only 2 of them were won by McCain. Yet there is plenty of complaining. Curious.

McCain utterly collapsed in Michigan in 2008, presidential numbers from that year are unrepresentative of congressional voting.

Of course, that is the entire point. It would be folly to draw districts based on some arbitrary 'competitive' metric. As we have seen, when that 'competitive' metric does not provide the outcomes desired by a political party, that political party simply discards that metric (but only when it suits them).

Another example is, of course, New Jersey, where competitive seats were drawn and 1 party won all the competitive seats.

Another example is, of course, PA-06 (R+1), PA-07 (EVEN), PA-08 (D+1) PVIs based on the 2004 and 2008 elections.  I suspect these competitive seats are not what the original poster had in mind when he talked about creating competitive seats!

Ideally, this competitiveness metric would be based on the average of various national and statewide elections. For the latter, weird overperformances would not be considered- like Sebelius winning the KS gubernatorial. All the elections from the past decade would be considered, but recent ones would be more heavily weighted.

It should also be noted that I would not require AZ to draw competitive districts , although I also know it is required by the state constitution.


My apologies if this is kind of hard to understand, but:

I believe there has been some discussion of a point based system? This is my view of how competitiveness might work for such a system:

Scale of 1-10, More points=bad

VRA violation= +10 (automatic disqualification)
Contiguity violation(Poor road connections, counties barely bordering each other, geographical features) =+4 for every split
Large UCC violation= +3 for every split
Large County splits= +1 for every such split (+.5 within UCCs)
# of High erosity districts= +.5 for each high-erosity district
# of Microchops= +.5 for every split (+.1 in UCCs)

The points and things to be penalized for don't matter, what matters is this:

CoI and Competitiveness interests can negate points, except for on the VRA.

For example, a proposed map of VA has a Shenandoah valley based district. This district is very erose, incurring a +.5 penalty. However, because the Shenandoah valley is a CoI, the +.5 does not go into effect.

Competitiveness can also do the same, although it would only be implemented as a factor in more inelastic states. Competitiveness also subtracts points, in addition to negating them. For example, in Alabama, it would not only negate the violation of the Mobile and Birmingham/Montgomery UCC, but also subtract .5 points from the total for creating a competitive district.

Naturally, the numbers would need to be thought through more, but do you all like the concept of point-negating with CoI and point-subtracting with Competitiveness?

 

Points are fine, but from what I've seen it is virtually impossible to get agreement on the relative weight of the points. If a few of us on the site can't get there, I doubt a legislative body would either.

Instead I think the better tactic is to keep each category separate with its own score. This results in multiple top plans, but that's not a bad outcome. It allows a final decision-making body the chance to weigh local subjective considerations when selecting from a set of fair plans.

Ooh, I like that a lot! Does my approach to Competitiveness and CoI sound ok?

[muon2]

I'd build CoI into the score for geographic integrity. County and muni chops are a form of CoI, so UCCs and MCCs should go there too. UCCs and MCCs get chopped and it adds to the chops.

Competitiveness, which I measure as the polarization of the plan, is its own score. So is the skew of the plan from a fair distribution between the parties. These measures can be applied to the set of fair plans to weed out politically undesirable results.

[Miles]

There are 9 EVEN PVI districts and none were drawn by Republican legislatures. Just pointing that out.

[krazen1211]

There are 9 EVEN PVI districts and none were drawn by Republican legislatures. Just pointing that out.

Even PVI districts as of 2011-2012 include:

NJ-02 - drawn by Republicans
PA-07 - drawn by a Republican legislature
PA-08 - this is actually D+1. drawn by a Republican legislature
WI-07 - drawn by a Republican legislature
NH-01 - drawn by a Republican legislature
OR-05 - drawn by a Republican branch of the legislature

[Miles]

There are 9 EVEN PVI districts and none were drawn by Republican legislatures. Just pointing that out.

Even PVI districts as of 2011-2012 include:

NJ-02 - drawn by Republicans
PA-07 - drawn by a Republican legislature
PA-08 - this is actually D+1. drawn by a Republican legislature
WI-07 - drawn by a Republican legislature
NH-01 - drawn by a Republican legislature
OR-05 - drawn by a Republican branch of the legislature

It's curious why you're ignoring the current set of PVI's that I was referring to.

You must be logged in to read this quote.

Maybe I should have specified Republican trifecta to cover OR-05, but I thought it was a compromise.

[krazen1211]

There are 9 EVEN PVI districts and none were drawn by Republican legislatures. Just pointing that out.

Even PVI districts as of 2011-2012 include:

NJ-02 - drawn by Republicans
PA-07 - drawn by a Republican legislature
PA-08 - this is actually D+1. drawn by a Republican legislature
WI-07 - drawn by a Republican legislature
NH-01 - drawn by a Republican legislature
OR-05 - drawn by a Republican branch of the legislature

It's curious why you're ignoring the current set of PVI's that I was referring to.

You must be logged in to read this quote.

Maybe I should have specified Republican trifecta to cover OR-05, but I thought it was a compromise.

Republican legislatures that passed the required One Man One Vote redistricting in 2011 or 2012 could only use the PVI present at the time. That is the cook PVI figures I posted. At the time, the districts listed held an even or worse than even PVI for the Republican party.


Your PVIs would only be valid for states that passed redistricting subsequent to the 2012 elections.

[Miles]

Republican legislatures that passed the required One Man One Vote redistricting in 2011 or 2012 could only use the PVI present at the time. That is the cook PVI figures I posted. At the time, the districts listed held an even or worse than even PVI for the Republican party.


Your PVIs would only be valid for states that passed redistricting subsequent to the 2012 elections.

Ok, well its still true that none of the current EVEN districts were drawn by Republican trifecta. You may try to talk around that, but thats the fact.