North Carolina 2020 Redistricting

[BenjiG98]

I'm not a North Carolina expert, but this is how I would have cleaned up the court map:





Districts 1, 2, 3, 4, and 7 are unchanged.

[Torie]

I'm not a North Carolina expert, but this is how I would have cleaned up the court map:





Districts 1, 2, 3, 4, and 7 are unchanged.

What are your metrics for clean-up? For example, in Torie's world, "cleanup" means subject to the VRA, drawing a map using neutral metrics, which means minimizing splits and erosity and respecting metro are lines, and then as a tie breaker between plans that are about equal in merit, going for partisan proportionality.

In the case of NC, it would assume that the state constitution would not have a role here, because the court will have new personnel and reverse itself, or SCOTUS will emasculate it, and that the VRA will be trimmed back to not require maps that fail to hew to the metrics above, leaving only that if a series of maps hewing to neutral metrics can be drawn, then if one or more of them entails a majority minority CD, or perhaps even a performing minority CD, such a map must be selected.

Your map does not entirely hew to the neutral metrics outlined above.

[Tintrlvr]

I'm not a North Carolina expert, but this is how I would have cleaned up the court map:





Districts 1, 2, 3, 4, and 7 are unchanged.

What are your metrics for clean-up? For example, in Torie's world, "cleanup" means subject to the VRA, drawing a map using neutral metrics, which means minimizing splits and erosity and respecting metro are lines, and then as a tie breaker between plans that are about equal in merit, going for partisan proportionality.

In the case of NC, it would assume that the state constitution would not have a role here, because the court will have new personnel and reverse itself, or SCOTUS will emasculate it, and that the VRA will be trimmed back to not require maps that fail to hew to the metrics above, leaving only that if a series of maps hewing to neutral metrics can be drawn, then if one or more of them entails a majority minority CD, or perhaps even a performing minority CD, such a map must be selected.

Your map does not entirely hew to the neutral metrics outlined above.

I at least contest that your metrics are neutral, or at least that they way you measure them is neutral, in effect if not in intent.

The only by-definition neutral metric of redistricting is partisan fairness, which is unfortunately quite vague to be a particularly strong guide. The next-most neutral metric is communities of interest, but what those constitute is itself debatable, and I strongly disagree that counties (or even in some cases municipal boundaries) are an appropriate way to measure them.

[Sol]

I would not argue that partisan fairness is neutral either. All of this stuff is subjective.

[Coastal Elitist]

They should go all out on a gerrymander. It's a joke that the "fair map" the court made splits the black community of charlotte. Really shows that they don't care about the VRA unless it benefits them.

Actually the court map's split of Charlotte is ok imo since Carrabus actually has a decent growing black population; no matter what it's pretty much impossible to consolidate all of Charlotte's black population into a single district.

My bigger problem with the court map is the split of Raliegh, especially the black population. Why not just put all of Raleigh in NC-02 and much whiter Cary into NC-13?

The split of Fayetteville, the Sandhills, and Winston-Salem are also bad.

And district 9 especially just doesn't make any coherent sense.

To me, it seems like what happened is the 3 judge panel just spent an hour playing around on DRA until they got something that was "clean" and achieved partisan balance. They clearly did not have a great understanding of the state's geography as to be expected. They should have gotten a special master, but they would've had to really fastrack that process.

The court knew exactly what it was doing. They had to get a 7-7 map. Any neutral draw wouldn't have done all those splits but their goal was to get a 7-7 map and make it so best-case R's could only get 8 seats.

[Torie]

I'm not a North Carolina expert, but this is how I would have cleaned up the court map:



Districts 1, 2, 3, 4, and 7 are unchanged.

What are your metrics for clean-up? For example, in Torie's world, "cleanup" means subject to the VRA, drawing a map using neutral metrics, which means minimizing splits and erosity and respecting metro are lines, and then as a tie breaker between plans that are about equal in merit, going for partisan proportionality.

In the case of NC, it would assume that the state constitution would not have a role here, because the court will have new personnel and reverse itself, or SCOTUS will emasculate it, and that the VRA will be trimmed back to not require maps that fail to hew to the metrics above, leaving only that if a series of maps hewing to neutral metrics can be drawn, then if one or more of them entails a majority minority CD, or perhaps even a performing minority CD, such a map must be selected.

Your map does not entirely hew to the neutral metrics outlined above.

I at least contest that your metrics are neutral, or at least that they way you measure them is neutral, in effect if not in intent.

The only by-definition neutral metric of redistricting is partisan fairness, which is unfortunately quite vague to be a particularly strong guide. The next-most neutral metric is communities of interest, but what those constitute is itself debatable, and I strongly disagree that counties (or even in some cases municipal boundaries) are an appropriate way to measure them.

Everyone can have different opinions on what metrics to use, and that is fine. My only suggestion was to spell them out, and then I spelled out mine as an example (they are quite close to the Muon2 rules). Defining terms is the first step to having a more productive conversation. So if proportionality is the trump card, just disclose that. I assume that when picking between proportional maps, you would prefer one that is less erose with fewer chops, but perhaps not.

[Tintrlvr]

I'm not a North Carolina expert, but this is how I would have cleaned up the court map:



Districts 1, 2, 3, 4, and 7 are unchanged.

What are your metrics for clean-up? For example, in Torie's world, "cleanup" means subject to the VRA, drawing a map using neutral metrics, which means minimizing splits and erosity and respecting metro are lines, and then as a tie breaker between plans that are about equal in merit, going for partisan proportionality.

In the case of NC, it would assume that the state constitution would not have a role here, because the court will have new personnel and reverse itself, or SCOTUS will emasculate it, and that the VRA will be trimmed back to not require maps that fail to hew to the metrics above, leaving only that if a series of maps hewing to neutral metrics can be drawn, then if one or more of them entails a majority minority CD, or perhaps even a performing minority CD, such a map must be selected.

Your map does not entirely hew to the neutral metrics outlined above.

I at least contest that your metrics are neutral, or at least that they way you measure them is neutral, in effect if not in intent.

The only by-definition neutral metric of redistricting is partisan fairness, which is unfortunately quite vague to be a particularly strong guide. The next-most neutral metric is communities of interest, but what those constitute is itself debatable, and I strongly disagree that counties (or even in some cases municipal boundaries) are an appropriate way to measure them.

Everyone can have different opinions on what metrics to use, and that is fine. My only suggestion was to spell them out, and then I spelled out mine as an example (they are quite close to the Muon2 rules). Defining terms is the first step to having a more productive conversation. So if proportionality is the trump card, just disclose that. I assume that when picking between proportional maps, you would prefer one that is less erose with fewer chops, but perhaps not.

I didn't say proportionality, I said partisan fairness. They're not necessarily the same thing. But the inherent nature of partisan fairness is that it is not clearly able to be defined by strict rules (proportionality is an attempt to define partisan fairness into a strict rule, but I don't think it's the only way to consider partisan fairness and may not be the best one). But partisan fairness is clearly the only partisan-neutral way to redistrict; every other way to redistrict can have built-in unfair partisan effects (even if not partisan intent). I know that's a bit tautological, but it's the place from which you have to start: The goal is partisan fairness, the only question is how you get there.

My view is that fair redistricting can't be determined by strict rules, and attempts to do so are failing ventures to begin with. The only good standard is know-it-when-you-see-it re: partisan fairness. That reasonable map may or may not be compact, may or may not follow jurisdictional boundaries and may or may not be proportional. Anything else inherently prioritizes some goal other than partisan fairness above partisan fairness, but, as stated above, partisan fairness is the only partisan-neutral redistricting goal. The second alternative to partisan fairness is to follow communities of interest, but that has a similar challenge, since communities of interest are often not bound by jurisdictional boundaries, occasionally are not compact and may sometimes not result in districts that are proportional.

[Sol]

Why on earth should we redistrict with partisan fairness in mind? Personally I redistrict to best represent the communities that people identify with, to the extent practicable, and I don't know why that's not valid.

[ProgressiveModerate]

They should go all out on a gerrymander. It's a joke that the "fair map" the court made splits the black community of charlotte. Really shows that they don't care about the VRA unless it benefits them.

Actually the court map's split of Charlotte is ok imo since Carrabus actually has a decent growing black population; no matter what it's pretty much impossible to consolidate all of Charlotte's black population into a single district.

My bigger problem with the court map is the split of Raliegh, especially the black population. Why not just put all of Raleigh in NC-02 and much whiter Cary into NC-13?

The split of Fayetteville, the Sandhills, and Winston-Salem are also bad.

And district 9 especially just doesn't make any coherent sense.

To me, it seems like what happened is the 3 judge panel just spent an hour playing around on DRA until they got something that was "clean" and achieved partisan balance. They clearly did not have a great understanding of the state's geography as to be expected. They should have gotten a special master, but they would've had to really fastrack that process.

The court knew exactly what it was doing. They had to get a 7-7 map. Any neutral draw wouldn't have done all those splits but their goal was to get a 7-7 map and make it so best-case R's could only get 8 seats.

I think it's a bad map but I don't think there were really malicious partisan undertones by the court as you suggest. None of the splits really exhibit partisan sorting and tbh, the splits really aren't that excessive given NC counties.

There are many ways a 7-7 (or really 7-1-6) map can fall naturally.

No matter what, any fair map will have at least 5 D leaning seats: Charlotte, Raliegh, Durham, Black-Belt, and Greensboro (w/ maybe Salem).

A fair map also by default has about 6 R leaning seats: Asheville + 2 more Appalachia seats, the coastal seat, the Wilmington/SE Seat, and the central seat.

After that, you are left with 3 seats that could be drawn to be competative. Firstly, the 2nd Charlotte seat. Charlotte is hard because neither splitting it down the middle or packing it into a single district is great, but I'd argue as things stand cracking it is the better of the 2 choices, creating a D leaning seat. Here the court chose the cracking option.

Next, you have the Fayetteville/Sandhills seat which I still think you can technically draw to be narrowly Biden, but it seems like Clinton-Trump or Trump-Trump configs are far more common. It's also posssible to make a majority minority seat or very close to it at least but the court basically cracked this community down the middle which favors Rs.

Finally, the suburban Raleigh seat. The big decision here is what happens with Durham-based NC-04. One option is you can pair Durham/Chapel Hill with with Cary which practically makes a lot of sense. However, when you do this, you get weird leftovers in southern Wake + Alamance County which become weird to deal with. The other option is to connect Durham and Orange Counties with some surrounding counties with reasonably high populations to create a black opportunity seat of sorts, and then create a suburban seat with the Southern half of Wake County and some combination of Johnson, Harnett, and Chatham counties. Personally this one is tough because if you look at the district individually, Orange-Durham-Cary makes more logical sense from COI but in terms of how it affects the overall map, the rural pairing of NC-04 makes more sense.

If the court was really drawing with partisan intents, they could've easily made 6 safer by making it more like the old config, or 13 more D by doing a better partisan sort in Wake, or not all Randolph County to NC-09, but they did not.

NC is overall just a weird state to redistrict with 14 seats and no matter the map something has to be a bit weird/awkward cause there are simply more COIs than districts, especially in the eastern part of the state where people often talk about how one community has to be cracked.

[Torie]

Here is a map that follows the Torie rules, which closely resemble the Muon2 rules. I ignored all partisan data when drawing it. The map give the Pubs one extra seat when I checked after the fact, but it should be 7-7 within a couple of election cycles.

I am not trying to persuade anyone of the merits of my approach. That effort would be utterly futile. It is an attempt at description for those that might be interested, and nothing more. This place is not one that is open to persuasion in general. It never has been. And that is OK.

Cheers.



Here is an alternative that avoids a tri-chop of Mecklenburg. I am not which map scores higher. They have their competing virtues and flaws.

https://davesredistricting.org/join/d1dd6217-3446-4208-b344-d58281a953f9

[If my soul was made of stone]

men will literally complain that no one listens to them right before posting a map with a three way split of mecklenburg county and a bunch of completely random lines instead of going to therapy

[cvparty]

men will literally complain that no one listens to them right before posting a map with a three way split of mecklenburg county and a bunch of completely random lines instead of going to therapy

tbf splitting it into three does make sense in a vacuum, it just doesn’t work very well with the current set of district numbers/population. when NC gains a 15th seat and charlotte grows more, it’d be perfect

[Torie]

Does "partisan fairness" have the same definition as pornography - you know it when you see it?

At the end of the day, one needs some tangible metrics to work with, that try to minimize subjectivity, that make it much more difficult to reasonably adjudicate matters.

And now we have the poli sci math nerds, who use fancy mathematics, geometry and statistics to try to offer up at least a veneer of objectivity. However, under that particular hood, this particular set of eyes finds an ample supply of smoked up mirrors. They have been running circles around the legal class, and I find that somewhat annoying. Lawyers should always be at the top of the pyramid, not math nerds. 

[Sol]

men will literally complain that no one listens to them right before posting a map with a three way split of mecklenburg county and a bunch of completely random lines instead of going to therapy

Really nothing wrong with a three-way split of meck, the geography kind of asks for it.

[President Punxsutawney Phil]

I dislike splitting Mecklenburg in three, but it's a choice I'm willing to make in certain circumstances. Depends on parameters and what I'm doing elsewhere.

[ProgressiveModerate]

Snip

Outside of the triangle and that weird NC-02, I actually quite like your first map. The 2nd and 3rd begin to have some weird pairinbgs

[Torie]

Yes, Cary in one CD and Raleigh in the other works out very well hewing slavishly to neutral metrics based on governmental subdivisions, without getting into racial percentages at all. That is precisely what my map 2 did.



But then I changed the map again, although not in Wake County.

Obsessive as I am, I kept playing with the NC map, totally ignoring race and partisan numbers, and finally came up with the below after it seemed no more desirable tweaks were in play that would give me pleasure (including. e.g., trying to minimize the population involved with county chops). And then I peeked at the partisan spoils just for fun. This map just happens to be very close to proportional. Trump barely carried NC-13 (the green CD), which is trending Dem at warp speed (Trump 2020 +1.7%, Trump 2016 + 7.5%, which has to be one of the larger swings in the nation). So as of 2024, it may well be a 7-7 map. Hewing to county and city lines may overall help the Pubs, but hewing to MSA’s in general does not, given the current party coalitions. The ying and the yang.



https://davesredistricting.org/join/a8077809-bfa9-496f-857b-ef7cb938f7c2

[Sol]

What does that map look like if you combine the Winston and Greensboro MSAs for redistricting purposes, Torie?

I ask since it's very much a Bay Area situation where what's functionally one metro area is kept as two due to the Census Bureau being a bit conservative on these things. (This is true of the Triangle too).

[Torie]

What does that map look like if you combine the Winston and Greensboro MSAs for redistricting purposes, Torie?

I ask since it's very much a Bay Area situation where what's functionally one metro area is kept as two due to the Census Bureau being a bit conservative on these things. (This is true of the Triangle too).

Almost nothing, since NC-05 entirely covers the Winston MSA, and then spills into the Greensboro MSA, and NC-06 takes in the balance of the Greensboro MSA. Combining the two MSA's would increase the score of the map without changing it in other words.

The thing about using MSA's, particularly in a state like NC that has so many multi-county MSA's, is that it kind of forces the lines of the map, limiting one's choices, so the subjectivity component is minimized. I tend to pretty slavishly honor them, absent the map getting too erose and looking that way to the eye (aesthetics of a map really matter in the public square), or it avoids a large and ugly county chop in favor of but a micro-chop of a county. That is why I split the Morgantown MSA into halves, with two counties in NC-10, and two counties in NC-08.

[Sol]

What does that map look like if you combine the Winston and Greensboro MSAs for redistricting purposes, Torie?

I ask since it's very much a Bay Area situation where what's functionally one metro area is kept as two due to the Census Bureau being a bit conservative on these things. (This is true of the Triangle too).

Almost nothing, since NC-05 entirely covers the Winston MSA, and then spills into the Greensboro MSA, and NC-06 takes in the balance of the Greensboro MSA. Combining the two MSA's would increase the score of the map without changing it in other words.

The thing about using MSA's, particularly in a state like NC that has so many multi-county MSA's, is that it kind of forces the lines of the map, limiting one's choices, so the subjectivity component is minimized. I tend to pretty slavishly honor them, absent the map getting too erose and looking that way to the eye (aesthetics of a map really matter in the public square), or it avoids a large and ugly county chop in favor of but a micro-chop of a county. That is why I split the Morgantown MSA into halves, with two counties in NC-10, and two counties in NC-08.

Ah, I was probably being a little circuitous. What I was trying to get at is, what does the map look like if you draw a W-S-Greensboro district?

[Torie]

What does that map look like if you combine the Winston and Greensboro MSAs for redistricting purposes, Torie?

I ask since it's very much a Bay Area situation where what's functionally one metro area is kept as two due to the Census Bureau being a bit conservative on these things. (This is true of the Triangle too).

Almost nothing, since NC-05 entirely covers the Winston MSA, and then spills into the Greensboro MSA, and NC-06 takes in the balance of the Greensboro MSA. Combining the two MSA's would increase the score of the map without changing it in other words.

The thing about using MSA's, particularly in a state like NC that has so many multi-county MSA's, is that it kind of forces the lines of the map, limiting one's choices, so the subjectivity component is minimized. I tend to pretty slavishly honor them, absent the map getting too erose and looking that way to the eye (aesthetics of a map really matter in the public square), or it avoids a large and ugly county chop in favor of but a micro-chop of a county. That is why I split the Morgantown MSA into halves, with two counties in NC-10, and two counties in NC-08.

Ah, I was probably being a little circuitous. What I was trying to get at is, what does the map look like if you draw a W-S-Greensboro district?

Yes, indeed, if your question is how much is the map messed up by virtue of creating another safe Dem CD by combining Winston-Salem with Goldsboro. In response to that inquiry, see below.

[ProgressiveModerate]

If Republicans win supermajorities and flip the court in 2022, could they make it so justices are elected by districts? Gerrymandered court districts fall pretty naturally if you create a Charlotte and Raliegh based seats and let the other 5 seats sort themselves out.

[AustralianSwingVoter]

Republicans won both state supreme court seats, so now have a 5-2 majority. NC is in for some brutal mid-cycle redistricting.

[Vern]

So will we be getting new maps?

[Nyvin]

Republicans won both state supreme court seats, so now have a 5-2 majority. NC is in for some brutal mid-cycle redistricting.

They have a 4-3 majority, one of the judges running was an R already.