Monday’s trial recaps are here and here. Tuesday’s trial recaps are here and here.
Plaintiffs were scheduled to call their last witnesses on Wednesday morning and the defense to begin presenting their witnesses in the afternoon. The court has other items scheduled for Thursday and Friday. While there is a little optimism that proceedings could finish today (if formal closings are waived and submitted in writing), most are resigned to the fact that a fourth day on Monday will be needed to wrap up.
I’ll extend my diaries to a part 7 for Monday’s last witnesses and a part 8 for to cover closing arguments and overall thoughts of the proceedings. Parties have five additional days to submit additional documents to the judge (due 4/18 at 5PM) and then the judge will rule within five days after that (by 4/25 at latest). No matter the ruling, an appeal to the Supreme Court is all but certain.
Wednesday, April 6, 9AM
Overall, Wednesday’s testimony brought more pushback and objections from both sides.
First plaintiff witness is Professor Michael Smith of Emporia State University. He is an expert on Kansas politics and elections and is the first called by the plaintiffs located in Douglas County.
He describes the history of Kansas congressional seats. Kansas had 5 seats until 1993 (The 5th district was a rural and suburban Wichita seat which was combined with more SE counties when eliminated). KCK and Douglas have shared a district many times in history. The institutions of the university system being a big factor, specifically KU’s main campus in Lawrence and KU’s Academic Medical Center located in Wyandotte. I-70 links workers to residences in Douglas and Wyandotte in many industries. Johnson County (which also has KU satellite campuses) and Wyandotte have also always been in a single shared district. Lawrence has never been linked with western Kansas. As the eastern areas of the state have grown in population over the decades, those districts (mainly CD-3, but also CD-2) have become more compact and borders moved east in each cycle.
The results of the 2020 census required CD-1 (rural big 1st) to add 33,000 people as many western counties lost population. In Ad Astra, the Kansas Legislature produced a map that draws CD-1 further east by carving a strip of two counties, Jackson and then Jefferson out of CD-2. While not a very logical redrawing, that does add about 36,000 in population to CD-1. Then Ad Astra scoops into Douglas County to add just the city of Lawrence and over 90,000 in additional population. Again, CD-1 only needed about 33K in population and now Ad Astra has moved 130K into CD-1, about four times the number needed. To offset that, six counties west of Lawrence and Topeka currently in CD-1 are moved into CD-2.
The displacement of Lawrence is harmful to voting and democracy. Because Lawrence is a younger community, it’s voter participation already lags. By placing it the first, Lawrence voters will likely become even more discouraged because their votes will not matter and be drowned out in the GOP ocean of CD-1.
On cross, Professor Smith is asked about the GOP argument that placing Kansas University and Kansas State University is a benefit. A — KSU is a land grant university and it’s history is not that of a research university. It has a top animal medical program, but not a human one. It has no law school. If linking similar universities as COIs is a concern, why not link KU with Washburn University (the only other law school in the state)? Smith is asked if he realizes Ad Astra links KU with it’s satellite campus in Salina and also links Fort Hays St, with puts 3 of the 6 regent universities in the same district? A — Well, why not also link Wichita, since it has a KU medical campus? Q — Are COI legally defined? A — No, but the maps and shapes do tell a story. Q — CD-1 needs population, so how do you decide between moving Topeka or Lawrence? A — I don’t accept that choice because there are other options available. You don’t have to move either. Q — Asked again, objection, A — I really can’t answer because I would have to study more about each city to see which would be the lesser of those bad choices.
Next is a private citizen and self-employed business owner residing in Lawrence. She is familiar with CD-1 because her father grew up in western rural Kansas. She talks about how little Lawrence has in common with rural CD-1. Lawrence doesn’t have big farms. It doesn’t share the western Kansas water aquifers. It doesn’t have large meat packing plants. It doesn’t have large feed lots. When there is talk of expanding or relocating rural based ag plants to Lawrence, the citizens roundly reject those plans. They don’t have anything in common. She is afraid her vote would not make a difference in CD-1. If she was moved to CD-1, she would have a long drive to Salina to visit her representative's closest office. She wanted to attend the redistricting listening tour, but it was announced without enough advanced notice and she couldn’t attend.
After a 10:30 break, the last morning expert for plaintiffs is Dr. Loren Collingwood an Associate Professor at University of New Mexico. His expertise includes research on racially polarized voting (RPV) and Ecological Inference Theory (EI). He studies election data to draw conclusions about micro behaviors of voters. I have nearly four pages of notes for the next two hours and there are too many details to transcribe here. The defense objects to the witness. They are very skeptical of his field of study (major flaws and unreliable). The judge allows the expert, but that was followed by sharp questions and objections for over an hour.
Dr. Collingwood discusses the concept of RPV, where minority groups have a preferred candidate that is often different from white voters. Those differences can occur at any electoral level. He studied precinct level data for Kansas with election data from the last decade (when available). As an example of RPV, white voters in CD-3 backed Donald Trump by 55%, while black voters in CD-3 supported Hillary Clinton by 95% and Latinos preferred Clinton by 83%. All minorities as a group supported Clinton by 85%. This is a RPV gap, but actually not as bad as in other CDs in Kansas in this presidential race.
2018 was the only example of no RPV in a Kansas district when both white and minority voters in CD-3 preferred Laura Kelly for governor. CD-3 has characteristics of a crossover district, which is increasing rare in the US. Under Ad Astra, it no longer would rate as a crossover district. Over the last decade, the minority preferred candidate (nearly always a Democrat) won 7 of 9 races studied. Using the new Ad Astra map, the new CD-3 would have elected a minority preferred candidate in only 2 or 3 races. To show the lack of crossover, analyzing CD-2 showed less favorable results, with the minority preferred candidate only winning once (Laura Kelly in 2018). Removing Lawrence voters from CD-2 in Ad Astra would make chances of a minority preferred candidate winning virtually impossible.
While CD-2 racial demographic percentages are numerically unchanged between the old and new congressional maps, clearly the new white voters moved into CD-2 and their preferences (RPV habits) in Ad Astra are much less likely to support minority preferred candidates. White voters in Lawrence vote much differently than white voters in more rural areas. Although the minority population percentages don’t change, the new white voter patterns acts as minority voter dilution.
On cross, “EI” is described as highly debated and not an accepted standard. They list all the social science fields where it is described as unreliable. Dr. Collingwood points out that American elections is specifically one of the areas it is strongly endorsed. Also, “highly debated” means exactly that. People are discussing it and trying to refine it. In academic circles, highly debated doesn’t mean it’s controversial or wrong.
The defense tries to tie RPV into the Voting Rights Act and enforcement in Kansas. There are objections sustained about asking the Dr. for legal conclusions. The defense attorney tries asking three different ways about VRA measurements and each is stopped as inappropriate. They eventually just tell the judge that the Dr. needs to answer questions about his report and brightline rules referenced. Well, you need to ask about that specifically. The defense is very frustrated — if the Dr. can’t answer questions in his report as they pertain to the Kansas constitution, what is he doing here? The judge asks why they don’t just admit into evidence what they are trying to link. They quote a different article by Collingwood that appears to discredit EI and RPV methodologies. Again, it appears they are deliberately cherry picking phrases out of context in an attempt to discredit him. It fails.
The defense tries to dispute CD-3 status as a crossover district. They asked in his assumptions about minority preferred candidates if he uses polls or primary election data. No. They ask if white voters in CD-3 support Sharice Davids, isn’t it likely that the support is not based on who minorities support, but the fact that she has the endorsement of business groups like several Chambers of Commerce. He studies election results, and didn’t study variables like endorsements which would be very difficult to actually quantify or measure.
There is more out of context questions about Dr. Collingwood’s “support” for the Princeton Gerrymander Project. He didn’t say that in his deposition. They have to pull the text to confirm he wasn’t aware of that project and as a separate question he said Princeton is a good university. That was not an endorsement of the project in question. The defense asks to admit that data, and another objection is sustained. The attorney proceeds to read in the results, which results in another objection and request to strike. Is defense council attempting to self-publish?? Collingwood is not an expert on the project in question, he has not studied it, and evidence admission is denied.
On redirect, plaintiffs reestablish the strengths of RPV and results in Kansas. They add while racial and partisan gerrymandering is highly correlated, they are different and distinct. While all Democrats are treated badly in Ad Astra, minority Democrats fair far worse than white Democrats.
That’s an end to a contentious morning. The plaintiffs rest. The afternoon brings a motion to dismiss and then two defense witness that don’t go so well… Those details later in Part 6.