The recap of the morning opening statements was detailed here.
April 4th, 12:20PM (Three witnesses scheduled)
Monday afternoon begins with the first expert, Jonathan Rodden, a Professor of Political Science at Stanford University. He was a Fulbright student at the University of Leipzig, Germany and earned his BA at University of Michigan. Before joining the Stanford faculty in 2007, he was the Ford Associate Professor of Political Science at the Massachusetts Institute of Technology. He earned his Ph.D. from Yale. But first, an objection by the defense. In what becomes a standard maneuver for the record prior to each expert is the motion to exclude each witness individually as nonrelevant and not an expert. It is denied. Cross examination is available to dispute qualifications and findings.
Dr. Rodden analyzed the Kansas map (named Ad Astra) and also drew two maps from scratch as an exercise to discover the level of difficulty in drawing maps using different historical criteria. There is always conflict and trade offs. He drew a map using a concept of least change and a map of compactness. He did not draw one with a criteria to maintain core retention since that is not a standard priority.
What he discovered is redrawing a Kansas Congressional map is easy. Rebalancing only requires moving 3% of the population. That can be done in one of two ways. Either by expanding the 1st to include many more rural counties or a larger metro area. He offers these maps as examples, not proposals. The Least Change Map simply draws boundaries of the big 1st further east into the 2nd and further south to the Oklahoma border into the 4th. The 2nd continues to include areas from the Nebraska to Oklahoma border in the eastern part of the state. It shrinks the 3rd. The Most Compact map shrinks both the 3rd and the 2nd. The 2nd no longer extends to the Oklahoma border and the 1st expands to cover that area in the southeast part of the state. In both versions, only three county splits are needed. Neither version splits Wyandotte County. Why not split it? Well, it’s unique in that it has a united government that joins KCK leadership and county government into a single governing entity. It makes no sense to split its congressional representation. Additionally, there is simply no need.
What to make of Ad Astra? First, Voting Tabulation Districts should not be split. That is an opportunity for error in distributing ballots. It should only occur with the county splits to precisely bring population deviation to zero. The Ad Astra map instead of limiting VTD splits to 3, it splits 18. Ad Astra splits 5 cities, when none are required to rebalance populations. Minority populations are split from the Missouri border in Kansas City Kansas to Manhattan along the Kaw River or I-70 into different districts. The Native American population in NE Kansas is split into different districts. The odd shapes of the districts in the map actually highlight the splits.
Cross begins at 1:30. Why does your report criticize violating “adopted” criteria when the legislature in fact didn’t vote to approve any criteria? Answer, the criteria or guidelines have to be used as a point of comparison in analyzing both the new map and historical maps. Q — You can’t really measure the “quality” of a community of interest? Objection — properly ask questions. Sustained. Can economic groups be a community of interest? Here are several examples of business groups, like the manufacturers of SE Kansas. Aren’t commercial communities of interest valid? Object to asking for a legal conclusion. Q — Next, do numbers mean anything? How many counties did you move? Lets count. 1,2,3,... 32. A — Those counties don’t hold much in population and is still only representing moving 3% total. Q — Did you realize the League of Women Voters map moved 34 counties. No, he didn’t analyze their map. Q — Did you realize the original Ad Astra map split an Indian Reservation and the GOP reunified that reservation? A — No, it wasn’t his job to track changes made in the GOP map process, just review the final version. Q — Did you realize Kansas Senator Fransisco’s Lawrence district is mostly kept whole in CD 1? A — No, Q — Did you realize towns with similar minority populations like Emporia, Junction City are in CD 1 along with Lawrence? Objection — you are asking him to do math about minority populations of towns on the stand. Question — Is the Kaw River area a community of interest? A — It can be but there is tension with that criteria verses the tradition of keeping current districts compact. We are done at 2:45.
The next witness is Dr. Jowei Chen, a Yale and Stanford educated associate professor at the University of Michigan. His expertise is in statistics and he ran over 1,000 random simulations of Kansas maps to compare to Ad Astra. Objection made to his expertise, overruled. His conclusion was Ad Astra was a product of extreme political bias that resulted in minority populations having less representation than 94% of maps generated. He ran simulations using statewide election data from the last decade. It was only limited by one cycles results being unavailable from the state. Why is state data used when simulating potential congressional district results? It eliminates variables, like different candidates and voter motivation within different districts. Statewide races and results smooth those variances. He ran the election results in his 1,000 maps and found Ad Astra in comparison was always a statistical outlier.
How is it an outlier? In the 1,000 maps and simulations, CD-3 is almost always a Democratic voting district. In Ad Astra, CD-3 leans Republican. Even in rare election results (an example being Ad Astra having CD-3 as Democratic using 2020 US Senate results), but it was still outside the norm compared to others. The end result is while Ad Astra doesn’t guarantee a GOP victory, it does always give the GOP the most favorable result possible. It also places the voters from the north part of KCK in a “safe” GOP district 99.1% of the time, another outlier result.
Just one problem of Ad Astra is the county splits. Ad Astra splits a total of four counties, not the minimum three. It also splits 18 VTDs not 3.
When Ad Astra is compared on a racial grading graph to the other 1,000 simulations, the sims mostly fall into a cluster at a 29%-30% grade (a statistical norm), with a smaller cluster around 25%. Ad Astra is an outlier at 22%. Every measure shows Ad Astra reflects an extreme racial bias.
Cross examine begins at 4:45. Q — What is the racial makeup of Ad Astra’s CD-2? A — I don’t grade statistically using that type of breakout. The models are simply compared with each other. Q — Why not show that? Objection, asked & answered. Q — Do you always testify on the side of plaintiffs? A — I don’t testify on the behalf of anyone. I’m showing models and numbers. Q — Do you always classify states / maps you testify against as extremely biased? A — Only if their particular map is extremely biased as a result of modeling scores. Q — Do you and Dr. Rodden travel together in testifying against states? Objection. The questions are an attempt to frustrate the witness. The defense attorney has also taken to call the witness, not by “Dr.” but as some mashup of his name (Jowei pronounced similarly to “Joey”) Chen. It comes out repeatedly as “Choy”. In a plaintiff objection, they ask the defense attorney to refer to him properly by his name and title: Dr. Chen. The defense attorney seems legitimately surprised at his mistake and offers an apology to the courts and witness. ‘Good thing we aren’t in front of a jury’ he then muses.
Splits aren’t really that bad. Q — Isn’t the difference between splitting 3 and 4 counties just one. A — Answer, no, that increase of “just” one county is an increase of 33%.
The questions continue about the Ad Astra building process. Do you know why Pawnee County was split? Would it surprise you it was to keep an Indian Reservation together with another county? Dr. Chen did not study that. Does your simulations try to account for race? No. Does your simulations try to account for social communities of interest? No. Does your simulation account for core retention or otherwise preserve state districts. No. Again, the simulation maps are random, based only on the most basic rules. Population has to be even and districts have to be contiguous. Even without accounting for race or other demographics, the random maps are always to the left of Ad Astra. In other words, Ad Astra always places in the high 90%+ rating at racially segregating voters compared to any random group of maps. In random maps, there were over 400+ versions that also split Wyandotte County. Even those random splits graded better at protecting against racial dilution than Ad Astra.
5:20, Last afternoon witness, attorney and Kansas State Senator Ethan Corson (D) from northern Johnson County and a member of the House and Joint Redistricting Committee.
Q — When did you find out about the fall 2021 listening tour about redistricting? A — When it was announced to the public. Democrats weren’t consulted about the schedule. If I had been, I would have told the committee not to schedule the Johnson County stop on the Shawnee Mission School District’s first day of school, and certainly not at 3:00, precisely when school is released. There were many complaints from parents that could not attend. The entire tour was ill advised because US census data had not yet been released. We are asking for public comment, when the public doesn’t know what the population changes are and what potential changes would be needed.
Q — How would you characterize the actions of GOP members of the committee during the tour. A — Disrespectful. At the Johnson County stop, most GOP members, including the Senate President spent their time scrolling on their phones instead of listening to constituents. (A photo of several senators at the Overland Park meeting is displayed showing exactly that.) Q — So what was the purpose of the tour? A — I think it was just a box checking exercise. We made stops and collected testimony. That’s all. Senator Wagle had already promised a 4-0 map. I don’t think there was anything we could do as a group to stop that, no matter what people asked for.
People wanted the metro area (Wyandotte and north / central Johnson) to be kept together. Instead of compacting the fastest growing CD in Kansas, the new map grew CD-3 by adding all Miami, Franklin and Anderson. Q — Do people in the populous areas of Wyandotte and Johnson consider themselves neighbors? A — Absolutely. People cross the county line all the time. They live on one side but work on the other. They travel from Johnson to KU medical center in Wyandotte. Q — What about Johnson and Anderson, are they neighbors. A — No. Those areas in Franklin and Anderson are very rural and really separate. Q — Do you consider them part of the metro? A — No, and I don't think they do either. People move down there to get away from the city and be in a more rural community.
As a committee, we weren’t going to listen to voters. Once introduced, the map passed quickly, within 72 hours. Then override vote included thuggish tactics to secure the last few votes.
Cross. Q — You don’t like this map because you’re a Democrat. Objection, ask proper questions. A — I don’t like it because it’s bad for democracy. Q — Did you ever meet Senator Wagle? A — No. Q — She didn’t work on this map? A — Not that I know of. Q — Didn’t other Democratic Representatives (Clayton) on the committee also want to keep Johnson County together. Yes, but in context with other factors and only if possible in terms of COI. Q — Isn’t county splitting a legislative call? A — Yes, but within following guidelines that generally are against that. Q — Are “guidelines” in the state constitution? Are they law? Are they statute? A — They are not.
Redirect. Q — If you follow historical precedent and the goal of keeping core areas together, would you expect CD-3 to grow or shrink geographically in size. A — Shrink to become more compact. Q — When Representative Clayton put forward her plan to keep Johnson County together (Map named Mushroom Rock), was she completely serious or was there some sarcasm in the proposal. A — There was some sarcasm. Q — Did the GOP members vote for Mushroom Rock that importantly met this GOP stated goal of keeping Johnson County together? A — No, they voted against it.
6:30PM Finish. More experts on Tuesday.