On Tuesday, North Carolina's Republican-led legislature passed new state House and Senate districts after a state court struck them down earlier this month for discriminating against Democrats in violation of the state constitution. Shown at the top of this post (see here for a larger version), the new Senate map redraws the 21 invalidated districts out of 50 total. However, as we'll demonstrate in this post, many of the districts in both chambers still bear the signs of partisan gerrymandering, in violation of the court's order.
Campaign Action
Making matters worse, most Senate Democrats voted in favor of the GOP's new stealth gerrymanders of the upper chamber on Monday, undermining the cause for fair maps and giving the GOP political ammunition by praising the process itself. However, Senate Democrats unanimously voted against the new House map on Tuesday, and House Democrats almost uniformly held firm against both maps. Furthermore, the court retains final say over whether to approve the new districts, and it has already appointed a nonpartisan expert to draw the lines for them if necessary.
The court's criteria mandated that any new maps make a "reasonable" effort to draw lines that "improve the compactness" of the legislature’s districts and split fewer precincts compared to the illegal versions. Mapmakers are allowed to consider preserving the integrity of municipalities and avoid pairing incumbents in the same districts. Critically, they are prohibited from preserving the cores of the illegal districts or relying on any partisan or election data, but as we'll explain below, it beggars belief that Republicans didn't have partisanship in mind.
The court didn't prohibit using partisan data to evaluate maps after they have been drawn, which is what we have done here since legislators could not. Indeed, we calculated the 2010 census racial demographics and 2004-2016 statewide partisan elections, which you can see in this spreadsheet for the Senate map.
Under this analysis, the GOP’s Senate proposal discriminates against Democratic voters less than the map that was struck down, but it still discriminates against them. One straightforward way to assess how much this legislative map does (or doesn't) favor one party is to sort each seat in each chamber by Donald Trump’s margin of victory over Hillary Clinton and see how the seat in the middle—known as the median seat—voted.
Because the state Senate has an even number of seats, we average the two middle seats to come up with the median point in the chamber. That median seat voted for Trump by a 54-44 margin, 6 points to the right of his 50-47 win statewide. In other words, to capture a bare majority in the Senate, Democrats would have to win seats that Trump carried by 10 points (at least), even though swingy North Carolina only voted for Trump by 4 (when rounded). That's hardly much better than the illegal map's median districts backing Trump by 12 points.
North Carolina's unusual rules regarding how counties can be divided and combined among districts limited the scope of which districts could be redrawn and how, but there are still some noticeable signs of legislators putting their thumbs on the scale within counties. For instance, Senate Districts 31 and 32 in Winston-Salem shown below appear to do far more than merely consider incumbency: They go to likely unjustifiable lengths to protect Republican state Sen. Joyce Krawiec from seeing her currently gerrymandered 31st District turn from dark-red to swingy.
Indeed, the new 31st District is very uncompact, and to connect Republican-leaning white voters in eastern Forsyth County with heavily white and Republican Davie County to the southwest, the two portions are connected via a thin strip of territory that splits the suburb of Clemmons. These two districts prioritized incumbency far ahead of the higher-value criterion of compactness and the more neutral criteria of avoiding unnecessary municipal splits in Clemmons.
Had legislators drawn a more compact version of Forsyth County such as the nonpartisan map we devised below, only Winston-Salem would remain divided, and the districts go much further toward addressing the court's preference for compactness. Our version ignored incumbency, since because Krawiec was elected in an illegally gerrymandered district, accounting for her incumbency only serves to further perpetuate that illegal gerrymander.
While the Senate map does make major improvements toward remedying gerrymandering in other parts of the state such as in Charlotte, Republicans went much further toward stealth gerrymandering in the state House, where 56 of 120 districts must be redrawn. Under the new map below and its accompanying demographic and election result spreadsheet, the median districts voted for Trump by a wide 55-43, a margin 8 points greater than his statewide margin of 4 points. That means the districts still retain a large majority of the 11-point median seat advantage Republicans had under the illegal gerrymander.
We'll have much more to say about the House map soon, but districts appear particularly problematic in areas such as Greensboro, Winston-Salem, Fayetteville, Wilmington, Greenville, the Charlotte suburbs, and even smaller towns and rural areas such as Columbus County in the southeast, where Democrats and voters of color still appear cracked and packed to create GOP-friendly seats. We have proposed our own nonpartisan state House map below, and it performs better on the court's nonpartisan criteria, as we'll demonstrate in an upcoming analysis.
Importantly, the process by which the GOP-run legislature passed the new maps presents its own problems, which could lead to the court stepping in and drawing its own maps. Because this is now the third time courts have struck down legislative districts that Republicans drew this decade, the court's order mandated strict rules about where and how the remapping could be conducted. Specifically, the judges told Republicans that they were required to “conduct the entire remedial process in full public view” and make “any relevant computer screen visible to legislators and public observers.”
But the computer-generated maps from which the GOP cobbled together their proposal were not drawn “in full public view.” To the contrary: They were drawn by University of Michigan professor Jowei Chen, who simulated 2,000 district maps based on criteria similar to the court's own, half of which sought to protect incumbents and half of which didn't.
As an expert witness for the plaintiffs, Chen's maps and analysis were furnished as evidence of the GOP's extreme gerrymandering (and accepted as such by the court). They were not generated with the intent that legislators or the court would adopt one of them, especially since they didn’t take into account more subjective redistricting criteria, such as the preservation of communities of interest.
Furthermore, Republicans may very well be relying on Chen's maps to launder their partisan intent. That's because the GOP's legal team had access to the partisan statistics accompanying Chen's maps—the exact type of data the court forbade Republicans from using.
While none of Chen’s plans leaned as far toward the GOP as the party’s illegal gerrymanders did, some could provide Republicans with a significant unfair edge. Signs of the GOP's hidden intent became clearer on Monday, when attorneys representing Republican legislative leaders accidentally sent the forbidden partisan data to legislators, potentially tainting the entire process in the eyes of the court. Undoubtedly, the Republican redistricting experts who have gerrymandered everything in North Carolina from Congress down to local school boards are familiar enough with the state’s geography to have a sense of an area's partisan lean.
In addition, the computer-generated map selected on Tuesday used incumbency protection as one of its criteria, which could be problematic if the court ultimately determined this criterion was merely a pretext to grant the GOP undue partisan advantage. Notably, Senate Republicans’ counterparts in the state House didn’t use Chen's maps that took incumbency into account, but Republicans in both chambers nevertheless made amendments to both maps to protect incumbents.
The plaintiffs in this lawsuit have until Sept. 27 to respond to the GOP's maps with any objections, and as we have illustrated in this story, they absolutely must object if North Carolina is to finally have fair legislative districts for the first time this decade. In 2018, Democratic candidates won more votes than Republicans yet failed to take majorities, and if the court doesn't step in and fix the remaining gerrymandered districts, that outcome risks a repeat in 2020, when legislative elections will determine which party if any gets to control next decade's redistricting.