Wisconsin Gerrymandering

Joined: Nov 2002
Posts: 20353
On another topic, I'd love to peak in on an alternate universe where it was Democrats who were benefiting from multiple extreme gerrymanders to find out just how political the court is in the opinions it is coming to.

I know the Wisconsin Democrats, while nominally having an ordinary bipartisan gerrymandering process ready to go, were planning back in 2008 to write an extreme gerrymander in their favor. They rejected a Republican deal, probably given in bad faith, to write into law a neutral boundary drawing process back when everyone thought the Dems would control the government in 2010.
Oct 4, 2017 2:11 PM
0 0
Joined: Oct 2008
Posts: 6986
Esoteric AllusionOn another topic, I'd love to peak in on an alternate universe where it was Democrats who were benefiting from multiple extreme gerrymanders to find out just how political the court is in the opinions it is coming to.

I know the Wisconsin Democrats, while nominally having an ordinary bipartisan gerrymandering process ready to go, were planning back in 2008 to write an extreme gerrymander in their favor. They rejected a Republican deal, probably given in bad faith, to write into law a neutral boundary drawing process back when everyone thought the Dems would control the government in 2010.

So it's the Dems fault too, then?
Oct 4, 2017 3:09 PM
0 0
Joined: Nov 2002
Posts: 20353
Paris the Goat
Esoteric AllusionOn another topic, I'd love to peak in on an alternate universe where it was Democrats who were benefiting from multiple extreme gerrymanders to find out just how political the court is in the opinions it is coming to.

I know the Wisconsin Democrats, while nominally having an ordinary bipartisan gerrymandering process ready to go, were planning back in 2008 to write an extreme gerrymander in their favor. They rejected a Republican deal, probably given in bad faith, to write into law a neutral boundary drawing process back when everyone thought the Dems would control the government in 2010.

So it's the Dems fault too, then?

I don't know the story behind the gerrymander of every state. In Wisconsin, the Democratic leaders brought their destruction on themselves by rejecting the opportunity to squelch gerrymandering in the state. They did this because they thought they were going to be in the position Republicans ended up being. There are only two fixes to this. 1) The Supreme Court upholds the lower court ruling. 2) Scott Walker is defeated next year and we wait until 2022 to have fair maps again. That's pretty much it.

Republicans are currently much worse than Democrats at being willing to destroy political norms to increase power. But gerrymandering is an area where they are on par and both parties will maximize their gerrymandering opportunities if given the chance. It would be foolish not to. Republicans are the beneficiaries of gerrymandering, by a lot, simply by good fortune.

Nate Cohn has a piece on the math of gerrymandering (that he is billing as the sociological gobbdygook of gerrymandering) here:

Link

In it, he explains that because of developments in partisan geographical sorting and further refinements in data science, parties in 2020 are going to be able to wipe out each other in any state they fully control. States with major representation in the US House could effectively eliminate the other party from winning anything of note.
Oct 5, 2017 1:27 PM
0 0
Joined: Oct 2008
Posts: 6986
I meant for Wisconsin, specifically, not for every state. The Democrats dropped opposition to gerrymandering when they thought they were the ones who could do the actual gerrymandering, because of course they did. It's the same reason why IRV (or similar systems) will not get a fair shake--why should the party in power enact something that would reduce their power?
Oct 5, 2017 1:59 PM
0 0
Joined: May 2008
Posts: 14839
Oct 6, 2017 2:20 PM
0 0
Joined: Nov 2002
Posts: 20353
The American Sociological Association ended up responding to Roberts' comments in a letter:

http://www.asanet.org/news-events/asa-news/asa-president-eduardo-bonilla-silva-responds-chief-justice-john-roberts

It's about as cutting of a letter from a major professional academic organization as I've seen.
Oct 12, 2017 2:18 AM
0 0
Joined: Nov 2002
Posts: 20353
I'm curious about your take on this article Nim:

The Supreme Court is Allergic to Math

It's a 538 piece that argues that the Supreme Court handles empirical questions, especially ones related to math - specifically statistics - poorly. It contends that it avoids, misunderstands, and engages in sloppy reasoning consistently because judges do not or do not want to understand how to interact with empirical matters. IT uses this gerrymander case as the springboard to make that argument, but it talks about a variety of cases and refers to research supporting that point of view.

This has been something at the front of my mind for awhile when thinking about the courts. They adjudicate questions that involve empirical judgement all the time and seem to be poor at it. Legal education doesn't seem interested in developing curriculum for it, and maybe that's too much to ask of what already is a long slog.

What really sunk this in for me was Florida vs. Harris. This was a case in 2013 that challenged the constitutionality of using police dogs as probable cause for searches of vehicles. The argument that it is unconstitutional, which I happen to agree with, is that while dogs in theory are excellent at detecting trace amounts of target substances, they also respond to the cues of their handlers. In drug search situations, police dogs have been repeatedly shown by a robust body of research to be functioning as a proxy for the hunches of police. So if you think that the hunch of a police officer isn't sufficient to be probable cause, then you should also think a dog sniff alert isn't either since that is what it is a proxy for. Therefore, the reliability of the dog alert must be established first before the alert can be considered probable cause. Otherwise, police might as well be searching people based on what their dowsing rods point to.

Now regardless of what you think of the legal merits of this reasoning, there is an empirical question at the heart of it. Do police dogs act as a reliable way to determine if contraband is likely present? That's an empirical question and the research is so easy to follow that a reasonably intelligent high schooler should be able to grasp it. The answer is "no" when it comes to drugs. Interestingly, it's "yes" when it comes to bombs. That's because the former is, with pervasive frequency, the result of dog alerts to police cueing and the latter is not. This because of differences in the circumstances in which dogs are used to alert to either form of contraband. When police want to search someone, a substantial % of the time dogs aren't that much different than dowsing rods in that they alert on what the person holding them subtly directs them to alert on.

The court ruled unanimously that dog sniff is probable cause provided the dog has completed a training program that tested it as this is sufficient to establish its reliability for probable cause purposes. This overturned a lower court decision on it. Disappointing to me, but what got me is the decision just didn't really understand the science. The testing referred to is a joke and does not simulate real situations. The unanimity stunned me into paying a lot more attention to how much judges just avoid or don't understanding empirical reasoning skills that are common among reasonably intelligent people with science / stat / philosophy education backgrounds. I don't have a solution - a few are offered in the article I linked - but it does concern me.
Oct 17, 2017 4:43 PM
0 0
Joined: Nov 2002
Posts: 20353
And in a heck of a coicidence, I just read this article:

https://www.propublica.org/article/supreme-court-errors-are-not-hard-to-find

Similar theme. It specifically talks about the dog alerted searches case.

Kennedy's reference to the Center for Immigration Studies' numbers, a notoriously dishonest anti-immigrant propaganda mill, was new to me and left me in awe. (To what should be no one's surprise, the numbers were dishonestly manufacturered.)
Oct 17, 2017 5:55 PM
0 0
Joined: Jan 2017
Posts: 82
Esoteric AllusionI'm curious about your take on this article Nim:

The Supreme Court is Allergic to Math

It's a 538 piece that argues that the Supreme Court handles empirical questions, especially ones related to math - specifically statistics - poorly. It contends that it avoids, misunderstands, and engages in sloppy reasoning consistently because judges do not or do not want to understand how to interact with empirical matters. IT uses this gerrymander case as the springboard to make that argument, but it talks about a variety of cases and refers to research supporting that point of view.

This has been something at the front of my mind for awhile when thinking about the courts. They adjudicate questions that involve empirical judgement all the time and seem to be poor at it. Legal education doesn't seem interested in developing curriculum for it, and maybe that's too much to ask of what already is a long slog.

What really sunk this in for me was Florida vs. Harris. This was a case in 2013 that challenged the constitutionality of using police dogs as probable cause for searches of vehicles. The argument that it is unconstitutional, which I happen to agree with, is that while dogs in theory are excellent at detecting trace amounts of target substances, they also respond to the cues of their handlers. In drug search situations, police dogs have been repeatedly shown by a robust body of research to be functioning as a proxy for the hunches of police. So if you think that the hunch of a police officer isn't sufficient to be probable cause, then you should also think a dog sniff alert isn't either since that is what it is a proxy for. Therefore, the reliability of the dog alert must be established first before the alert can be considered probable cause. Otherwise, police might as well be searching people based on what their dowsing rods point to.

Now regardless of what you think of the legal merits of this reasoning, there is an empirical question at the heart of it. Do police dogs act as a reliable way to determine if contraband is likely present? That's an empirical question and the research is so easy to follow that a reasonably intelligent high schooler should be able to grasp it. The answer is "no" when it comes to drugs. Interestingly, it's "yes" when it comes to bombs. That's because the former is, with pervasive frequency, the result of dog alerts to police cueing and the latter is not. This because of differences in the circumstances in which dogs are used to alert to either form of contraband. When police want to search someone, a substantial % of the time dogs aren't that much different than dowsing rods in that they alert on what the person holding them subtly directs them to alert on.

The court ruled unanimously that dog sniff is probable cause provided the dog has completed a training program that tested it as this is sufficient to establish its reliability for probable cause purposes. This overturned a lower court decision on it. Disappointing to me, but what got me is the decision just didn't really understand the science. The testing referred to is a joke and does not ?simulate real situations. ?The unanimity stunned me into paying a lot more attention to how much judges just avoid or don't understanding empirical reasoning skills that are common among reasonably intelligent people with science / stat / philosophy education backgrounds. I don't have a solution - a few are offered in the article I linked - but it does concern me.

I remember reading some years ago one of the dog-sniffing studies(performed in a church, with colored cards indicating both real and false drug caches...) that was completely damning of the practice, proving conclusively that handler-bias has a statistically significant effect on the ability of the dog to indicate false-positives. As you said, the results were very clear, and damning.? Their ruling that 'so long as it's been through training' either completely ignores or misunderstands these results - though i have to wonder if this was a partisan vote aimed at protecting the ability (and convictions) of police using these tactics. But really, dogs are worse than lie detectors, as at least there's an objective 'truth' that the detector is measuring. The dog will just hit on cues from a trainer regardless of what other factors may be present.
Oct 17, 2017 8:23 PM
0 0
Joined: May 2008
Posts: 14839
BattyRoy
Esoteric AllusionI'm curious about your take on this article Nim:

The Supreme Court is Allergic to Math

It's a 538 piece that argues that the Supreme Court handles empirical questions, especially ones related to math - specifically statistics - poorly. It contends that it avoids, misunderstands, and engages in sloppy reasoning consistently because judges do not or do not want to understand how to interact with empirical matters. IT uses this gerrymander case as the springboard to make that argument, but it talks about a variety of cases and refers to research supporting that point of view.

This has been something at the front of my mind for awhile when thinking about the courts. They adjudicate questions that involve empirical judgement all the time and seem to be poor at it. Legal education doesn't seem interested in developing curriculum for it, and maybe that's too much to ask of what already is a long slog.

What really sunk this in for me was Florida vs. Harris. This was a case in 2013 that challenged the constitutionality of using police dogs as probable cause for searches of vehicles. The argument that it is unconstitutional, which I happen to agree with, is that while dogs in theory are excellent at detecting trace amounts of target substances, they also respond to the cues of their handlers. In drug search situations, police dogs have been repeatedly shown by a robust body of research to be functioning as a proxy for the hunches of police. So if you think that the hunch of a police officer isn't sufficient to be probable cause, then you should also think a dog sniff alert isn't either since that is what it is a proxy for. Therefore, the reliability of the dog alert must be established first before the alert can be considered probable cause. Otherwise, police might as well be searching people based on what their dowsing rods point to.

Now regardless of what you think of the legal merits of this reasoning, there is an empirical question at the heart of it. Do police dogs act as a reliable way to determine if contraband is likely present? That's an empirical question and the research is so easy to follow that a reasonably intelligent high schooler should be able to grasp it. The answer is "no" when it comes to drugs. Interestingly, it's "yes" when it comes to bombs. That's because the former is, with pervasive frequency, the result of dog alerts to police cueing and the latter is not. This because of differences in the circumstances in which dogs are used to alert to either form of contraband. When police want to search someone, a substantial % of the time dogs aren't that much different than dowsing rods in that they alert on what the person holding them subtly directs them to alert on.

The court ruled unanimously that dog sniff is probable cause provided the dog has completed a training program that tested it as this is sufficient to establish its reliability for probable cause purposes. This overturned a lower court decision on it. Disappointing to me, but what got me is the decision just didn't really understand the science. The testing referred to is a joke and does not ?simulate real situations. ?The unanimity stunned me into paying a lot more attention to how much judges just avoid or don't understanding empirical reasoning skills that are common among reasonably intelligent people with science / stat / philosophy education backgrounds. I don't have a solution - a few are offered in the article I linked - but it does concern me.

I remember reading some years ago one of the dog-sniffing studies(performed in a church, with colored cards indicating both real and false drug caches...) that was completely damning of the practice, proving conclusively that handler-bias has a statistically significant effect on the ability of the dog to indicate false-positives. As you said, the results were very clear, and damning.? Their ruling that 'so long as it's been through training' either completely ignores or misunderstands these results - though i have to wonder if this was a partisan vote aimed at protecting the ability (and convictions) of police using these tactics. But really, dogs are worse than lie detectors, as at least there's an objective 'truth' that the detector is measuring. The dog will just hit on cues from a trainer regardless of what other factors may be present.


John Oliver had a segment relatively recently about all the junk science that has been used to lock people up. Bite mark analysis is the latest to be shown to be absolute junk but people are serving enormous terms because of it. Hair analysis has been used giving likelihoods 100,000,000x stronger than the science behind it.

there is a National Commission on Forensic Science that was set up to advise the Justice Department on how to avoid bad science, or was ? Attorney General Jeff Sessions shut it down in April.


http://theweek.com/speedreads/728186/john-oliver-takes-law-into-hands-fight-junk-forensic-science-since-trump-wont


https://theintercept.com/2016/09/07/white-house-report-concludes-that-bite-mark-analysis-is-junk-science/
Oct 17, 2017 8:47 PM
0 0
Joined: May 2008
Posts: 14839
While this has no bearing on gerrymandering, this is my current "legal thread" and this is insane to me:
https://www.huffingtonpost.com/entry/prosecutorial-misconduct-fake-subpoenas_us_59e4f2e5e4b03a7be58281c4?ncid=inblnkushpmg00000009

I mean my first assumption was that "well that guy isn't going to have fun in jail or enjoy life much without his admittance in the Bar." I had no idea misconduct of this nature was ok for DAs.
Oct 17, 2017 9:33 PM
0 0
Joined: Nov 2002
Posts: 20353
BattyRoy

I remember reading some years ago one of the dog-sniffing studies(performed in a church, with colored cards indicating both real and false drug caches...) that was completely damning of the practice, proving conclusively that handler-bias has a statistically significant effect on the ability of the dog to indicate false-positives. As you said, the results were very clear, and damning.? Their ruling that 'so long as it's been through training' either completely ignores or misunderstands these results - though i have to wonder if this was a partisan vote aimed at protecting the ability (and convictions) of police using these tactics. But really, dogs are worse than lie detectors, as at least there's an objective 'truth' that the detector is measuring. The dog will just hit on cues from a trainer regardless of what other factors may be present.

It was a 9-0 decision. Even Sotomayor joined the majority. It wasn't partisan. The evidence on dog-sniffing is quite strong from a variety of angles. Statistical analysis of outcomes, controlled testing, and underlying theoretical knowledge about animal behavior all point to the same thing. Dog handlers bias alerts based on what they want to happen such that dog alerts are not reliable and can function as a proxy of police hunches. The reason explosive-detecting dogs, though not nearly as good as their reputation, tend to do better because the suspicion environment is different.

What got me about that ruling is it completely blew a rudimentary understanding of the training and testing process. And it's not like an accurate understanding of the issue wasn't available. The case was made thoroughly in the briefs. The judges, Kagan most obviously, either didn't understand it or ignored it. This is far from the only issue with rulings getting facts wrong, but this one shook me because a uninamous decision missed something that could function as a test of a high schooler's critical thinking ability when it comes to science. As I've paid attention to it more, this is a major problem. This isn't completely surprising. You need to know almost squat about science, logic, or math to get through law school if you can cram past some basic classes and aptitude tests.

One solution proposed in both my links is staffing SCOTUS judges (and maybe others) with fact-checkers and/or advisers who can explain empirical issues to them and clean up arguments so there's no more dumb-ass assertions about proving a negative. But I think that idea is tricky as it places a lot of power in who those people are and opens up the question about who gets to decide who gets the jobs. Maybe it's not so different from clerking, but I'm not positive about that. I'd like to hear it explained more.
Oct 18, 2017 1:00 AM
0 0
Joined: Dec 2001
Posts: 41347
I think there's a lot to respond to, but I'll say this: what the ProPublica article points out is a real problem, but one that's slightly different than you might think, if you aren't versed in federal court law. Yes, the Supreme Court has gotten facts wrong, and they know it, which is why they almost never rely on facts or use facts outside the record. Federal district courts are where fact-finding happens. These are the courts where there is an actual trial, with actual evidence introduce, actual expert testimony, and actual cross-examination. On appeal, court of appeal review matters of law afresh, but do not generally question the facts from the lower court record. The Supreme Court isn't in the business, and shouldn't be, of doing a bunch of independent research, for the very reason that it didn't get properly presented, testified about, and cross-examined. When it comes to the Supreme Court's role in determining law, they almost always assume the facts as presented to them are true.

So the problem isn't that the Court is ignorant or incapable of assessing facts. It's that our criminal justice system widely relies on dubious science. This is a major problem that goes beyond dogs. This includes eye-witness testimony and fingerprinting. These are dubious metrics and the entire basis for most convictions.

But check it out:

If a person is convicted in District Court based on, say, eyewitness testimony, and the testimony isn't challenged as a matter of law, but as a matter of dubious fact, on any appeal, neither the Circuit Court or the Supreme Court is going to even consider that the eyewitness testimony was dubious. This is a base-level criminal justice problem.
Oct 18, 2017 4:06 PM
0 0
Joined: Jun 2002
Posts: 10871
NimChimpskyI think there's a lot to respond to, but I'll say this: what the ProPublica article points out is a real problem, but one that's slightly different than you might think, if you aren't versed in federal court law. Yes, the Supreme Court has gotten facts wrong, and they know it, which is why they almost never rely on facts or use facts outside the record. Federal district courts are where fact-finding happens. These are the courts where there is an actual trial, with actual evidence introduce, actual expert testimony, and actual cross-examination. On appeal, court of appeal review matters of law afresh, but do not generally question the facts from the lower court record. The Supreme Court isn't in the business, and shouldn't be, of doing a bunch of independent research, for the very reason that it didn't get properly presented, testified about, and cross-examined. When it comes to the Supreme Court's role in determining law, they almost always assume the facts as presented to them are true.

So the problem isn't that the Court is ignorant or incapable of assessing facts. It's that our criminal justice system widely relies on dubious science. This is a major problem that goes beyond dogs. This includes eye-witness testimony and fingerprinting. These are dubious metrics and the entire basis for most convictions.

But check it out:

If a person is convicted in District Court based on, say, eyewitness testimony, and the testimony isn't challenged as a matter of law, but as a matter of dubious fact, on any appeal, neither the Circuit Court or the Supreme Court is going to even consider that the eyewitness testimony was dubious. This is a base-level criminal justice problem.

When Justices know they're deciding based on false "facts", do they ever mention this in their decisions? Or even hint at it?
Oct 18, 2017 5:14 PM
0 0
Joined: Dec 2001
Posts: 41347
Paquito
NimChimpskyI think there's a lot to respond to, but I'll say this: what the ProPublica article points out is a real problem, but one that's slightly different than you might think, if you aren't versed in federal court law. Yes, the Supreme Court has gotten facts wrong, and they know it, which is why they almost never rely on facts or use facts outside the record. Federal district courts are where fact-finding happens. These are the courts where there is an actual trial, with actual evidence introduce, actual expert testimony, and actual cross-examination. On appeal, court of appeal review matters of law afresh, but do not generally question the facts from the lower court record. The Supreme Court isn't in the business, and shouldn't be, of doing a bunch of independent research, for the very reason that it didn't get properly presented, testified about, and cross-examined. When it comes to the Supreme Court's role in determining law, they almost always assume the facts as presented to them are true.

So the problem isn't that the Court is ignorant or incapable of assessing facts. It's that our criminal justice system widely relies on dubious science. This is a major problem that goes beyond dogs. This includes eye-witness testimony and fingerprinting. These are dubious metrics and the entire basis for most convictions.

But check it out:

If a person is convicted in District Court based on, say, eyewitness testimony, and the testimony isn't challenged as a matter of law, but as a matter of dubious fact, on any appeal, neither the Circuit Court or the Supreme Court is going to even consider that the eyewitness testimony was dubious. This is a base-level criminal justice problem.

When Justices know they're deciding based on false "facts", do they ever mention this in their decisions? Or even hint at it?


For purposes of adjudication, the record as presented - the facts found to be true by the District Court - are assumed to be true. This is for a good reason; the Supreme Court doesnt receive evidence or hear testimony. But the Court can theoretically do whatever it wants, and has reversed things on a clear error basis. If something was obviously wrong and the fulcrum of a decision, the Court could reverse. But this is rare for obvious reasons. Facts are, for better or worse, what the adversarial trial court determines. There are some cases where the Court weighs into facts as a matter of law - cases alleging Brady violations, for example. But again, this is a niche area with a different standard of review.
Oct 18, 2017 11:22 PM
0 0
Joined: Oct 2008
Posts: 8183
I haven't read this thread and am not well versed enough in Wisconsin politics for my opinion to matter, but Mother Jones' spanking new issue goes into how voter suppression in Wisconsin not only gave the state to him but possibly won the election for Trump.
Oct 19, 2017 7:02 PM
0 0
Joined: Nov 2002
Posts: 20353
boozeI haven't read this thread and am not well versed enough in Wisconsin politics for my opinion to matter, but Mother Jones' spanking new issue goes into how voter suppression in Wisconsin not only gave the state to him but possibly won the election for Trump.

I haven't read that article, but I know from the analyses out there that it is extremely likely the various voter supression policies under Walker swung the state to Trump. His victory margin is within a reasonable lower bound of lost Dem votes due to that. What's crazy is the upper bound is enough to have cost Feingold a Senate seat too. 100k+ net lost Dem votes is absolutely plausible. We will never know about that for sure, though.

I don't know if this applies nationally.
Oct 19, 2017 7:17 PM
0 0