Wednesday, March 21, 2018

Blogger Candidate Forum: Is There Such A Thing As A Good Electoral Map?

Hello Everyone:

Welcome to a rainy day edition of Blogger Candidate Forum.  The Human Typist decided that she had enough of dodging an early spring storm, currently soaking Southern California, and decided to seek shelter.  We cannot have Human Typist messing up her hair or neatly applied makeup, now can we?  Alright, enough about that, shall we move on to today's subject gerrymandering and commuter patterns?  

When we last took up the subject of gerrymandering, about two months ago, we looked at an upcoming Pennsylvania State Supreme Court case that would force the state to re-draw its congressional districts so that they were more evenly distributed between Democrats and Repbulicans.  We will be returning to this later in the post.  One thing is for sure, gerrymandering is bad; it unfairly tips the political scales in one direction or another. In addition to one-sided election results, a gerrymandered map is objectionable because the bizarrely draw districts do not correspond to any and all relevant geographical data.  You do not have to take Blogger's word for it, you just have to check out to Garret Dash Nelson's CityLab article "What Would a Good Electoral Map Even Look Like?"  Excellent question when you consider this statement from Mr. Nelson, "People have emotional and political attachments to all sorts of geographic entities: jurisdictions like states and cities as well as cultural regions like the Bay Area or Appalachia."  Seriously, do any of you think that Human Typist introduces herself as a proud resident of CA-33 (her congressional district) or has its boundaries tattooed on some part of her anatomy?  A No to both questions.

To quantify just how politically unfair gerrymandering is, "we can count up partisan demographics and look at election results."  However, the more difficult task would gauging how well or poorly the map matches up with the patterns of human geography. While many states mandate the districting process respect "so-called 'communities of interest,'" how they defined and delineated remains an open question, therefore these requirements rarely carry any legal weight.

However, there is a way to follow the geographic framework of a community by studying the networks that bind places together.  One network that is particularly crucial and fairly easy to measure and map is commuter patterns.  Mr. Nelson writes, "In the study of networks, a 'community' is defined as a group of objects that have relatively dense set of connections among each other, and a relatively weak set of connections beyond."  He does concede that this definition "is much thinner than our social and political concepts of 'community,' it does have the advantage of being measurable."

Using community structure of a network, as opposed to the assigned voting districts, offer the possibility of checking whether or not an electoral preserves "communities of interest."  Mr. Nelson argues that "a good map, seen from this perspective, would keep areas that are connected by commuter webs together in the same district as much as possible."  By contrast, "a bad map would slice up commuter regions, fracturing cities and confecting electoral districts out or places that have relatively less connection with one another."

How would a network scientist measure whether or not an electoral district preserves a "community of interest?"  These scholarly women and men developed a statistic called "modularity" which gauges how good (or bad) a proposed electoral district matches up with the established structure of connections within a given network.  Think of it "as a ratio between the number of communities which stay inside of the borders of a single district versus the number of commutes which cut across district lines."  Mr. Nelson humorously named this measure the "Elbridge score (Electoral Boundary Resemblance to Identifiable Geography)" after the 19th century Massachusetts governor and gerrymandering namesake Elbridge Gerry (; Feb. 10, 2017; date accessed Mar. 21, 2018) .

Garret Dash Nelson's own methodology is fairly straightforward.  He used "data set of millions of commutes--the same one that my colleague Alasdair Rae and I used to redraw the map of the U.S. commuter 'megaregions'--to compare congressional maps."  What Mr. Nelso discovered was an enormous variation in context to how well congressional districts pair with commuter regions made up of interconnected urban, suburban, rural areas.  The five states that did well were: Kansas, Indiana, Iowa, Kentucky, and Tennssee.  Each has a fairly even population distribution at a distance from each other.  Mr. Nelson notes, "This makes some intuitive sense.  Because the population of a U.S. Congressional district is standardized at around 700,000 people, the easiest states to divide into fair districts will be the ones which are already roughly split up into commuter zones of roughly 700,000 each."

One example, which can be viewed at is a map of Indiana's commuter patterns ("only the ones that both begin and end within the state") overlaid on the state's congressional districts.  Mr. Nelson points that each district "encloses a fairly self contained set of commutes, usually centered around ne of the state's major cities."  The commutes are illustrated in purple if they start and stop within the district's boundaries, orange if they cross districts.  The map does a good job of maintainin commutes within district boundaries.

You can compare these evenly divided mostly Midwestern states with the five worst states: Rhode Island, Hawai'i, New Hampshire, Idaho, and Nevada--all "characterized by heterogenous geography and difficult-to-contain urban clusters."  An example is all two of Rhode Island's congressional districts which slice through the capital city of Providence.  Downtown Boise, Idaho is located in the second district, but the majority of the city's populous western suburbs are situated in the first district.  In each instance, as the map of Rhode Island indicates, a huge portion of commuters live and work across district boundary lines.

This begs the question, "why aren't the state's that are most notoriously gerrymandered, like Pennsylvania and North Carolina at the bottom of the Elbridge scores?"  The reason is that each state is different--some are large, rectilinear, and evenly distributed and others have disparate population and urban development concentrations--that it is difficult to compare this score across state lines.  Therefore, Garret Dash Nelson proposes, "It's more useful, then, to test multiple proposed maps in one state against another.  By keeping the underlying geography of a state constant, we can see how well--or how poorly-- proposed maps fits a state's 'communities of interest' defined by commuter connections."

Returning to the State of Pennslyvania.  Comparing the Elbridge scores between the state's current gerrymandered map and the new map the state Supreme Court issued in February "suggests pretty convincingly that the current map cuts the state up into areas that don't match very well with commuter-based communities."  The gerrymandered version scored a mere 0.58; by contrast, the Pennsylvania Supreme Court Map rated a 0.65, and another proposed fair map scored 0.66.

Another way to diagram the same measurement, instead of mapping it out, is creating a matrix of a state's electoral districts, an counting the commutes between each district in each of the cell.  Mr. Nelson writes, "In a perfect (and impossible) districting scheme, all of the commutes would lie along the diagonal of the matrix, since they would begin and end in the same district."  If you check out, you can see the matrix that Mr. Nelson generated for Kansas and Nevada; two of the best and worst scorers.  He explains, "The number at each cell represents the percentage of the state's total number of commutes.  In Kansas, nearly everybody lives and works in the same congressional district.  In Neveada,..., a lot of people live in one district and work in another--a consequence of the fact that job-rich Las Vegas lies in its own NV-01, while the city's suburbs lie in NV-03 and NV-04."

Important thing to pay attention to, "there's a big conceptual leap between commuter connections and fully-constituted 'communities of interest.'"  Essentially, commuter networks do not record people who are not part of the workforce.  Further, they completely ignore some of the most significant ties that bind a community together such as: friendships, family networks, cultural ties, or ecology--then there the subjective unquantifiable elements that make places and political community more cohesive.

Attempting to pair electoral districts to the actual geography of "communities of interest" creates some complicated choices about the things we value.  For example, if you are interested in equatable housing or re-districting transit authority jurisdiction, tracing the geography of a commuter area is a logical decision.  However, if your goal is "to balance out partisanship at a federal level and make elections more competitive, then it doesn't necessarily follow that districts should follow functional geography."  Just the opposite may be true, given that Americans are increasingly concentrated in politically-likeminded geographies.

Garrett Dash Nelson uses this example: "which scenario would be preferable?  To vote in a geographically-coherent district together with lots of people with common interests, or in a district that doesn't make geographic sense but has an even mix of Deomcrats and Repbulicans?"  Think you know the right answer?  No, you do not because the answer really depends on other values like: "Are our representatives supposed to represent place-specific communities (in which case matching district boundaries to human geography is crucial) or do they just represent the interests of a national political party (in which case preserving local communities might be much less important)?"

These kinds of ideological and philosophical issues have been mostly passed over in the gerrymandering debate.  One thing is certain, we know what a bad electoral map looks like, but a good one is more difficult to figure out.