Several NC legislative districts struck down

[muon2]

It will be interesting to see where this line of cases ends up. They seem to suggest that a Dem partisan gerrymander is ok because there's no standard to determine how much gerrymandering is too much. However a Pub gerrymander may not be ok since in many states there is a disparate effect of racial and ethnic minorities which vote consistently Dem. Pub gerrymanders that pack Dems cannot avoid packing minorities, but the reverse is not true. If carried to the extreme it will create an asymmetrical view of the parties by the courts, which would be a marked departure from the cases of the 1990's.

[muon2]

It will be interesting to see where this line of cases ends up. They seem to suggest that a Dem partisan gerrymander is ok because there's no standard to determine how much gerrymandering is too much. However a Pub gerrymander may not be ok since in many states there is a disparate effect of racial and ethnic minorities which vote consistently Dem. Pub gerrymanders that pack Dems cannot avoid packing minorities, but the reverse is not true. If carried to the extreme it will create an asymmetrical view of the parties by the courts, which would be a marked departure from the cases of the 1990's.

How NC republicans will be able to gerrymander NC during the next session will clearly depend on the NC judicial election this year.

It will also depend on SCOTUS. These decisions are basically asking the court to overturn the Gingles standards for a VRA section 2 (Thornburg v Gingles 1986) and replace them with a much looser set of standards. The standards require districts with 50%+1 of a minority and racial bloc voting, so drawing such districts (as clarified by the Shaw decisions of the 1990s) have been a safe harbor for racial gerrymandering. In fact if a Pub plan didn't have such districts in the past they would be successfully challenged, but now they can be challenged if they do have them. That'll be the conundrum for the Pub mapmakers.

[muon2]

There's no question that the NC map is an egregious gerrymander. But past SCOTUS rulings would have considered them to be partisan gerrymanders, not racial gerrymanders. The WI case may well set the standard for partisan gerrymanders in court, and if so the NC gerrymander would probably fall under that standard. But this NC case is really testing a new theory about racial gerrymanders, not established law. If the Dems were seeking to overturn it on established law the case would have been decided at the beginning of the decade.

[muon2]

There's no question that the NC map is an egregious gerrymander. But past SCOTUS rulings would have considered them to be partisan gerrymanders, not racial gerrymanders. The WI case may well set the standard for partisan gerrymanders in court, and if so the NC gerrymander would probably fall under that standard. But this NC case is really testing a new theory about racial gerrymanders, not established law. If the Dems were seeking to overturn it on established law the case would have been decided at the beginning of the decade.

The legislative district boundaries were precleared by the USDOJ. Now the courts are saying that even though they could not have got preclearance without the 50%+1 districts, they shouldn't have done so.

The districts were also challenged in the state courts and upheld. That decision was appealed to US Supreme Court, which remanded the case to the NC Courts to reconsider their decision in light of the Alabama decision. The NC Supreme Court did so, and once again upheld the legislative district boundaries. The federal case was permitted to go forward only because there were independent sets of plaintiffs. I don't see how the SCOTUS can stay out of this considering that the NC courts were acting under their direct instructions.

So is it also your sense that this is a new theory about what constitutes racial gerrymandering?

[muon2]

This "soft Gingles" test seems to mostly eliminate the first prong. Under Gingles minority districts are not required unless there is a potential district of 50%+1 minority VAP. In Bartlett v Strickland the opinion said that a state did not have to consider crossover voters in drawing districts, and that section 2 only applied if there was a potential 50%+1 minority district. This "soft Gingles" would seem to require that Strickland be significantly altered and subordinated to a secondary position after other redistricting factors.

This "soft Gingles" also seems to heavily modify the third prong. The racial block voting analysis used by the state looks like standard fare used successfully in many cases for the last two cycles. The experts use various statistical means to show clear differences in preference between the white and black population. The Court here seems to rely on the fact that crossover voters exist to substantially change what has been meant by racial bloc voting. They point to the fact that as long as enough whites in an area support the candidate of choice by the blacks then overall bloc voting doesn't matter since given the crossover, the white majority can't defeat the preferred candidate of the black minority.

Both these changes seem to me to create a substantially new view of the VRA than the one that has been around since the 1990's.

[muon2]

This "soft Gingles" test seems to mostly eliminate the first prong. Under Gingles minority districts are not required unless there is a potential district of 50%+1 minority VAP. In Bartlett v Strickland the opinion said that a state did not have to consider crossover voters in drawing districts, and that section 2 only applied if there was a potential 50%+1 minority district. This "soft Gingles" would seem to require that Strickland be significantly altered and subordinated to a secondary position after other redistricting factors.

This "soft Gingles" also seems to heavily modify the third prong. The racial block voting analysis used by the state looks like standard fare used successfully in many cases for the last two cycles. The experts use various statistical means to show clear differences in preference between the white and black population. The Court here seems to rely on the fact that crossover voters exist to substantially change what has been meant by racial bloc voting. They point to the fact that as long as enough whites in an area support the candidate of choice by the blacks then overall bloc voting doesn't matter since given the crossover, the white majority can't defeat the preferred candidate of the black minority.

Both these changes seem to me to create a substantially new view of the VRA than the one that has been around since the 1990's.

It is not clear that the North Carolina did any racial bloc voting analysis.

In their finding that racial considerations predominated, the court pointed to numerous instances where Strickland was invoked as the rational for drawing 50%+ districts. Essentially, the map-drawers treated the words "VRA", "50%+", and "Strickland" as synonyms.

The court also pointed out that Strickland itself had noted that all the Gingles test had to be present.

I'm not saying the federal court in North Carolina is proposing a soft Gingles test, but that is what the Democrats want. Otherwise, on what basis is there a requirement for ordering race-based districting?

I assume that the US Supreme Court will take up this case.

The NC Supreme Court had previously upheld the legislative districts. That had been appealed to the US Supreme Court, which accepted the case for purposes of remanding it back to the NC courts, for them to consider the Alabama case. The NC Supreme Court then once again upheld the legislative districts. Perhaps the US Supreme Court could take up both cases, and uphold both decisions on a 4:4 vote.

I read the opinion (beginning on p. 138) as saying that there was an analysis of racially polarized voting offered by two separate experts for the state. Both found that there was racially polarized voting in that the white population preferred different candidates than the black population. The opinion however discounts this, since the Court found that racially polarized voting is insufficient to satisfy the third Gingle prong. To satisfy the third prong one must also show that the polarization allows the white majority to "usually beat the minority's preferred candidate." That additional test was not done.

The tricky part is that additional test from Gingles requires knowledge of the district in question. One can't tell if the white majority will usually beat the black-preferred candidate without knowing how much of which white population is in the district (eg. are they college students or farmers). It certainly allows for sub-50% districts to meet section 2, but it feels like it requires them when there is enough white crossover. This is the part that seems to run counter to Strickland.

It would seem that the Court wants a process by which one creates districts using generally acceptable redistricting principles without regards to race. Then one should test the totality of the plan to determine if there is a section 2 violation, and redraw as needed if a section 2 violation is detected. Then there would need to be further iterations until no violation is apparent. Generic testing of individual counties as done by NC would not be useful.

[muon2]

So you can't gerrymander if it involves persons of color, even if it is solely because they are Democrats, except to the extent if a "compact CD" is out there that would elect a candidate of their choice. That standard won't hold, if otherwise gerrymandering remains legal.

As I heard at a conference last week, that is exactly the standard the Dems are seeking. It makes Pub gerrys nearly impossible if any racial groups are impacted, yet preserves Dem gerrys when racial minorities happen to benefit. It creates a very asymmetric playing field for 2020 if it holds.