Princeton Gerrymandering launches "overnight express" map analysis service

[Jennifer Bremer]

This is a potential gamechanger! Sam Wang and the team at the Princeton Gerrymandering Project has developed a program that will provide rapid analysis of official draft maps produced at the state level, including congressional and legislative maps. They will provide a scorecard rating the map on competitiveness, geographic funny-business (county splits, non-compactness), and partisan fairness, all with A-F letter grades and additional information. With the right data from us (see below), they can turn a map around in a day and it's all free of charge! Click here for the details.

This program makes use of the new map analysis methodology (massive montecarlo simulations) that generates thousands or even millions of maps, applies past precinct-specific election data to all of them to see how they perform, and then compares how the map submitted stacks up against this suite of maps on key metrics. Is the map submitted biased one way or the other? Is it more or less compact, etc.? Is its performance pretty typical of how this "suite of maps" performs or is it an extreme outlier, way out in left or right field? 

What they need from redistricting activists is submission of official draft maps as they are released. Ideally, they need not just a copy of the map itself, but the set of shapefiles that go along with it (shapefiles are computer-readable files that enable quantitative analysis of a geographic shape, such as a district). Technically, what they want are the block-equivalency file for the newly-drawn districts (identifying the Census blocks in each district) and the shapefiles, ideally in geojson format. Don't worry if you don't know what these are--any GIS person will recognize them. 

If you cannot get the legislature/commission to cough up the shapefiles, they can still do the analysis by generating approximations of the shapefiles from a pdf of the map (or whatever), but this is more time-consuming and not quite as accurate. Our mantra should be, "transparency in 2021 means providing draft maps with shapefiles"!

This development is fantastic! It will enable the League and our allies on the ground to get a very rapid idea of how gerrymandered a map is and respond accordingly.

Jennifer      

Dr. Jennifer Bremer

Bremer....@gmail.com
mobile:  +1 301 955 6333

twitter: JBremerDevt

[C. Norman Turrill]

I know that some states have “compactness” is one of the criteria for drawing maps. In Oregon is not one of them, and in my opinion, it should not be. I can think of many reasons why compactness should not be a criterium and why districts might naturally have odd shapes. Compactness is somewhat arbitrary and goes against other criteria that are important, like adhering to geographic and political boundaries. For example, imagine a district that extends up into a mountain valley. The ridges on the sides of the valley are natural geographic boundaries, and the people that live in that valley are a natural community of common interest. Therefore, this district might have an odd shaped finger in its boundary. Thanks for listening.

Norman Turrill

Governance Coordinator

League of Women Voters of Oregon (lwvor.org)

Making Democracy Work® for All

--
You received this message because you are subscribed to the Google Groups "LWV Redistricting" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lwv-redistrict...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/lwv-redistricting/CAJMWQ8T9_9KQXA_4xiCZso5EuMAD64t_7eCw2t2Bjw9RHmYagQ%40mail.gmail.com.

[Donna Hornberger]

In Michigan we have a new Independent Redistricting Committee. The list of requirements for their drawing districts is in order of importance and compactness is dead last on that prioritized list.


Sent from Yahoo Mail for iPad

To view this discussion on the web visit https://groups.google.com/d/msgid/lwv-redistricting/9E43E28B-603E-4337-B841-448FD95EABE2%40mac.com.

[Jennifer Bremer]

Thanks for your interest! I'll try to answer all the questions to the best of my ability in one email. 

Data: They will definitely be uploading all of the data as soon as it's released (which means tomorrow, plus a week or so to shift it from the "legacy" format to the current layout). Princeton has a lot of computing power. 

Local redistricting (county/city/etc.):  I don't think that they will be able to do local governments, unfortunately. This is a nationwide system so they will already be trying to respond to 150 different map-making exercises (one congressional, one state house, and one state senate for each state). Given multiple maps for most of these, that's a lot to take on. 

Compactness/other criteria: They will be providing standard metrics, as I understand it. I don't know how much capacity they have to track each state's specific systems, but they may be able to add something on request. Can't speak for them on that, though. Local analysts may also be able to add that.

The main thing that we as activists need to do on the front end is to put pressure on the legislature (or other map-drawing authority) to make the shapefiles public for each map as it is released. These are the files that enable a computer to read and analyze a set of districts or any set of geographic features (county boundaries, city blocks, etc.). When google maps shows you a city boundary, it is most likely reading from some type of shapefile to do that. 

Princeton has a program ready to approximate the boundaries from a pdf or other image if shapefiles are not available, but it is a lot better (quicker and more accurate) if they can upload the newly-drawn districts as shapefiles. If the official map drawers are using any of the major software packages to draw the maps (Maptitude, ESRI, etc.), rather than sharpies, then that program automatically produces shapefiles as part of the output, so it's just a question of whether the powers that be will fork them over. Shapefiles were actually invented by ESRI, by the way, back in the 1970s, I think. 

Jennifer

Dr. Jennifer Bremer

Bremer....@gmail.com
mobile:  +1 301 955 6333

twitter: JBremerDevt

On Wed, Aug 11, 2021 at 9:43 AM Sue Engelhardt <jae...@verizon.net> wrote:

Jennifer -  is this something we could use for Pasquotank?  Their first public hearing is 9/13.

Sue

-----Original Message-----
From: Jennifer Bremer <bremer....@gmail.com>
To: lwvnc-red...@googlegroups.com
Sent: Tue, Aug 10, 2021 3:56 pm
Subject: Princeton Gerrymandering launches "overnight express" map analysis service

This is a potential gamechanger! Sam Wang and the team at the Princeton Gerrymandering Project have developed a project that will provide rapid analysis of official draft maps produced at the state level, including congressional and legislative maps. 

For each map, they will provide a scorecard rating the map on competitiveness, geographic funny-business (county splits, non-compactness), and partisan fairness, all with A-F letter grades and additional information to make it easy to tell the story and react to the map. With the right data from us (see below), they can turn a map around in a day and it's all free of charge! Click here for the details.

This program makes use of the new map analysis methodology ("massive montecarlo simulations" or similar terms) that generates thousands or even millions of maps, applies past precinct-specific election data to all of them to see how they perform, and then compares how the map submitted stacks up against this "suite of maps" on key metrics. Is the map submitted biased one way or the other compared to what you would expect in that state? Is it more or less compact, etc.? Is its expected electoral performance pretty typical of how this "suite of maps" performs or is it an extreme outlier, way out in left or right field? So far, courts are buying this analysis, at least in NC. 

What they need from redistricting activists to work their mathy magic is submission of official draft maps as they are released. Ideally, they need not just a copy of the map itself, but the set of shapefiles that go along with it (shapefiles are computer-readable files that enable quantitative analysis of a geographic shape, such as a district). Technically, what they want are the block-equivalency file for the newly-drawn districts (identifying the Census blocks in each district) and the shapefiles, ideally in "geojson format." Don't worry if you don't know what these are--any GIS person will recognize them. 

If you cannot get the legislature/commission to cough up the shapefiles, the PGP team can still do the analysis by generating approximations of the shapefiles from a pdf of the map (or whatever), but this is more time-consuming and not quite as accurate. Our mantra should be, "transparency in 2021 means providing draft maps with shapefiles"!

This development is absolutely fantastic! It will enable the League and our allies on the ground to get a very rapid idea of how gerrymandered a map is and then to respond accordingly.

Jennifer      

Dr. Jennifer Bremer

Bremer....@gmail.com
mobile:  +1 301 955 6333

twitter: JBremerDevt

--
You received this message because you are subscribed to the Google Groups "LWVNC Redistricting" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lwvnc-redistric...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/lwvnc-redistricting/CAJMWQ8Qr4ne-w14YB1-9pqn94t95asTRQXZmo4LNQ9NBXGhdig%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.