5 ms·

Math Professor Fighting Gerrymandering with Geometry | https://news.ycombinator.com/item?id=13713252 https://news.ycombinator.com/item?id=13713252 (Feb 2017) W

by j_s 9y ago

Math Professor Fighting Gerrymandering with Geometry | https://news.ycombinator.com/item?id=13713252 https://news.ycombinator.com/item?id=13713252 (Feb 2017) What convinced the Supreme Court to take the Wisconsin gerrymandering case? | https://news.ycombinator.com/item?id=14751953 https://news.ycombinator.com/item?id=14751953 (Jul 2017) Metric Geometry and Gerrymandering Group – Tufts University | http://sites.tufts.edu/gerrymandr/ http://sites.tufts.edu/gerrymandr/ a need for expert witnesses who understand the mathematical concepts applicable to gerrymandering. To meet that need, she’s spearheaded the creation of a five-day summer program at Tufts [the first in a series of regional trainings] that aims to train mathematicians to do just that [...] over 900 people have indicated their interest by signing up for a mailing list http://tufts.us15.list-manage.com/subscribe?u=3529c170e5d9b7aa8ab22ea62&id=a979bdf71d http://tufts.us15.list-manage.com/subscribe?u=3529c170e5d9b7... (re-post of mailing list link) edit: To be clear, I don't have a pony in this race; I just believe the info and opportunity is important and may reach members of its intended audience here. The July article dug into the reasoning behind the efficiency gap and pursuit of mathematical models: > The efficiency gap is an elegant way to quantify the extent of partisan gerrymandering. But more than that is needed to win a case. The challenged plan must also be more biased than other maps, inconsistent with the state’s spatial patterns, and durable in its effects. And these elements can only be established by additional studies. Another HN user shared a higher-level summary I will try to TL;DR here: > paulmd: SCOTUS turns based on which side of the bed Kennedy wakes up on (on most issues). [...] mathematicians heard [Kennedy might be open to striking gerrymandering down] And now the court cases are working their way back up to SCOTUS, only with the academic models that Kennedy has indicated he wants. https://news.ycombinator.com/item?id=14753909 https://news.ycombinator.com/item?id=14753909

trendia 9y ago
Thanks for posting this. The efficiency gap completely ignores geometry.
- jpfed 9y ago
  But why consider geometry itself to be important? Isn't geometry a mere proxy for something else that's desirable? My own suspicion is this: that districting is an act of partitioning the social graph. The weights of this graph are not necessarily "who socializes with who" but rather "to what extent do these people share common interests". It is not necessarily the case that regions characterized by people of common interests will be shaped in ways that we can express easy geometric criteria for.
  - trendia 9y ago
    Geography does matter within the Constitution. But also consider a counter example using a third party: California, which has 53 representatives. If 5% of the population votes third party, then there is an efficiency gap (since 2 - 3 representatives should be from that party), but to fix this gap would require some very strange districts. You could argue that this gap isn't very important, but to come to this conclusion you would have to use some metric other than the efficiency gap.
    - jpfed 9y ago
      There's almost always an efficiency gap. To judge whether an efficiency gap is disqualifying for a map depends on what other maps might be drawn- is this efficiency gap unusual given how the voters are distributed?
      - trendia 9y ago
        > To judge whether an efficiency gap is disqualifying for a map depends on what other maps might be drawn This is false. Efficiency has been applied by the Supreme Court to mean precise mathematical equality, rather than an approximate or relative metric [0]. For instance, if the populations of two districts is not exactly equal according to the most recent census data, then the district must be explicitly justified. Quote [1]: > But in Kirkpatrick v. Preisler,288 a sharply divided Court announced the rule that a State must make a “good-faith effort to achieve precise mathematical equality.”289 Therefore, “[u]nless population variances among congressional districts are shown to have resulted despite such [good-faith] effort [to achieve precise mathematical equality], the State must justify each variance, no matter how small.” We would expect that the Supreme Court would apply the same mathematical exactness to the efficiency gap, should they find the efficiency gap a valid measure of gerrymandering. [0] https://en.wikipedia.org/wiki/Wesberry_v._Sanders https://en.wikipedia.org/wiki/Wesberry_v._Sanders [1] http://law.onecle.com/constitution/article-1/05-congressional-districting.html http://law.onecle.com/constitution/article-1/05-congressiona...
sukilot 9y ago
Why does this project need research mathematicians to study simple statistical problems that have known solutions. It seems that they are recruiting "mathematicians just to provide an aura of authority to their work, in the one field where results can always speak for themselves.
- j_s 9y ago
  > simple statistical problems that have known solutions Can you please share the key terms allowing further research of the known solutions?
- ghaff 9y ago
  Because they’re not in fact simple problems?
weeksie 9y ago
Weird geometry is at best a flawed heuristic for gerrymandering. There are good reasons for oddly shaped districts (coastlines, natural features, "logical" constituencies in an odd layout, etc. . . .) Some of the geometric approaches gloss around this stuff but none do what the efficiency gap does: actually identify unfair "wasted" votes. The exciting thing about efficiency gap is exactly how unambiguous it is which makes it GREAT for basing a law on.

by j_s 9y ago

trendia 9y ago

Thanks for posting this. The efficiency gap completely ignores geometry.

jpfed 9y ago
But why consider geometry itself to be important? Isn't geometry a mere proxy for something else that's desirable? My own suspicion is this: that districting is an act of partitioning the social graph. The weights of this graph are not necessarily "who socializes with who" but rather "to what extent do these people share common interests". It is not necessarily the case that regions characterized by people of common interests will be shaped in ways that we can express easy geometric criteria for.
- trendia 9y ago
  Geography does matter within the Constitution. But also consider a counter example using a third party: California, which has 53 representatives. If 5% of the population votes third party, then there is an efficiency gap (since 2 - 3 representatives should be from that party), but to fix this gap would require some very strange districts. You could argue that this gap isn't very important, but to come to this conclusion you would have to use some metric other than the efficiency gap.
  - jpfed 9y ago
    There's almost always an efficiency gap. To judge whether an efficiency gap is disqualifying for a map depends on what other maps might be drawn- is this efficiency gap unusual given how the voters are distributed?
    - trendia 9y ago
      > To judge whether an efficiency gap is disqualifying for a map depends on what other maps might be drawn This is false. Efficiency has been applied by the Supreme Court to mean precise mathematical equality, rather than an approximate or relative metric [0]. For instance, if the populations of two districts is not exactly equal according to the most recent census data, then the district must be explicitly justified. Quote [1]: > But in Kirkpatrick v. Preisler,288 a sharply divided Court announced the rule that a State must make a “good-faith effort to achieve precise mathematical equality.”289 Therefore, “[u]nless population variances among congressional districts are shown to have resulted despite such [good-faith] effort [to achieve precise mathematical equality], the State must justify each variance, no matter how small.” We would expect that the Supreme Court would apply the same mathematical exactness to the efficiency gap, should they find the efficiency gap a valid measure of gerrymandering. [0] https://en.wikipedia.org/wiki/Wesberry_v._Sanders https://en.wikipedia.org/wiki/Wesberry_v._Sanders [1] http://law.onecle.com/constitution/article-1/05-congressional-districting.html http://law.onecle.com/constitution/article-1/05-congressiona...

sukilot 9y ago

Why does this project need research mathematicians to study simple statistical problems that have known solutions. It seems that they are recruiting "mathematicians just to provide an aura of authority to their work, in the one field where results can always speak for themselves.

j_s 9y ago
> simple statistical problems that have known solutions Can you please share the key terms allowing further research of the known solutions?
ghaff 9y ago
Because they’re not in fact simple problems?

weeksie 9y ago

Weird geometry is at best a flawed heuristic for gerrymandering. There are good reasons for oddly shaped districts (coastlines, natural features, "logical" constituencies in an odd layout, etc. . . .) Some of the geometric approaches gloss around this stuff but none do what the efficiency gap does: actually identify unfair "wasted" votes. The exciting thing about efficiency gap is exactly how unambiguous it is which makes it GREAT for basing a law on.