How hard is it to draw fair legislative districts? Guest columnist Andrew Dean, a SUNY Buffalo law student, reflects on the plans he and other students submitted in a competition to try and answer that question.
In the fall, I was on a team of six first-year law students assembled at SUNY Buffalo Law School to take on Fordham University and the Public Mapping Project’s Redistricting Competition. The challenge: using free software, redraw New York State’s Congressional Districts to be fair and sensible (based on a number of objective criteria). Between studying for final exams and going to classes, our team managed to draw a map of that beat out submissions from across the country to claim first prize.
Getting Started, a Whole Lot to Learn
Redrawing New York’ Congressional districts required us to learn a great deal quickly. Redistricting is governed by a body of caselaw and statutes that can be intimidating. Additionally, we had to master free software developed by researchers at Harvard and George Mason University. That was just the start. Once we had an idea of how to draw legal districts and physically implement them into the software, we were still stuck staring at a blank map of New York State. How could we divide New York State, in all of its diversity, into 27 Congressional Districts without destroying parts of what makes us so special?
I believe that the only way to solve big problems is to divide them into manageable chunks, and that’s exactly what we did. The first “chunk” we tackled was deciding, after hours of debate, that preserving communities was our top priority. This meant that if a region of the state read the same local newspaper, it stood a good chance of sharing a community identity. We refined this concept as we went, because fortuitously, each of the students on the team came from a different region of the state. I am a central New Yorker, so I knew that Utica, Rome and Syracuse share far more in common than do Utica and towns along the Pennsylvania border. Thus, our Congressional map joined Utica, Rome and Syracuse into a single district and gave the southern tier a single district. We did the same for every other region of the state.
Drawing the Maps
This sounds simple, but implementing the idea of community took some work. The other criteria (or “chunks”) we had to accommodate were (1) drawing districts of almost exactly 717,000 persons each, (2) complying with the Voting Rights Act, (3) making our districts as compact as possible, (4) making our districts land contiguous (meaning you could drive from one end to the other without passing through an intervening district), (5) being mindful of partisan proportionality and differentials, and (6) eliminating two districts from the state’s Congressional delegation. Finally, considering the growth in the Hispanic population, we set out to add a second Hispanic Voting Age Majority Congressional District, and also added an Asian Voting Age Plurality District in Flushing, Queens.
Over a two month span, we systematically confronted each of these criteria, drawing and redrawing districts dozens of times in the software. At one deeply discouraging moment toward the end of November, we discovered that our last Congressional district was short by over 60,000 people. We had accidentally drawn all of the first 26 districts 1-4% too large. So, we had to go back and manually redraw the entire map.
What emerged at the end of this process is something we are immensely proud of. In a matter of weeks, we had accomplished something that we thought could only be done by a handful of experts throughout the country. This is a true testament to the software, which gave us the freedom to think up an idea, discard it and start all over again until our map was perfect. Pictured at right are the old upstate Congressional Districts (top) and our proposed upstate Congressional Districts (bottom). Note that in addition to looking cleaner, our districts are superior to the old ones by every criteria put forward by the competition organizers (for example, our compactness ratios, partisan differentials etc... are far better).
Only a Few Ways to Slice the Cheesecake
Since winning the competition, people often ask me what the most important thing I learned is. At first, I gave the typical mantra of “work hard and good things will happen.” However, there is a deeper lesson here that only became clear after a few months.
Shortly after we submitted our maps back in December, Common Cause New York released its proposed Congressional maps (at left top). In March, a panel of federal judges adopted Judge Roanne Mann’s Congressional plan (at left bottom), which is now official for the State. All three of our plans share striking similarities. Although part of this can be explained by the close relationship between Common Cause and the Court, such an explanation does not account for SUNY Buffalo Law’s map being so similar.
Three groups, applying the same objective criteria independently, came up with similar maps, even though there were dozens of variables involved. This means that redistricting isn’t really hard at all.
If someone presented you with a cheesecake, and asked you to divide it into 8 equal parts, chances are you’d cut it into triangular slices (ok geometry geeks might do concentric rings that get bigger or something). Similarly, if someone presents you with a map of New York State, and tells you to follow a few rules (717,000 people per district, make districts as compact as possible, etc...) you’d probably draw something darn similar to what SUNY Buffalo Law, Common Cause New York, and the federal courts came up with.
The Real Strength of Public Mapping Software
Thus the real strength of public mapping software isn’t that it enables brilliant people to draw a virtuoso congressional map in their bedrooms. Rather, the real strength of public mapping software is that is shows the wisdom of the crowd. If different groups from across the state draw maps that are so alike, their common solutions must be right. We all cut the cheesecake into slices. Fair redistricting is within the grasp of ordinary people, as long as software does the drudge work.
OK So if Redistricting is So Easy...
If redistricting were so easy that anyone with the right equipment could come up with a fair plan, what the heck has been going on in New York for the past two centuries? This is where we must be careful. Although it would be easy to blame politicians for mucking up a straightforward problem, we must acknowledge that they have competing interests which are completely legal (and which none of the reform plans, including ours, took into account). Politicians often drawn districts to (1) protect incumbents and (2) protect partisan majorities in the legislature. Both interests are legal and have been around for centuries. Thus, although the public hates this kind of redistricting, it shouldn’t surprise us. In fact, a slight consideration of incumbency may be healthy...people may prefer consistent representation because seniority in the legislature corresponds to the funding a district receives. So, incumbency may be a valid consideration in redistricting. What the public must do, however, is guard against extreme gerrymandering. If we were to rank the things to consider when drawing a map, political incumbency should be on the list, but be somewhere far toward the bottom.
By entering the redistricting contest, we learned a heck of a lot and even won some prize money to boot. More importantly, however, is that we showed that a group of amateurs with the right tools can drawn fair Congressional districts, and that people from all over the State and country were doing the same thing. The true magic of the experience is that all of these groups came up with strikingly similar solutions to the same problem. We had crowd sourced redistricting.