Gerrymander Jeremiad

by John Goolrick

A hardy perennial in Virginia politics is the ritualistic denunciation of gerrymandering. Sure, redistricting is unfair. But none of the alternatives looks any better.

My hometown paper, despite its philosophical permutations over the years, has constantly railed editorially about changing Virginia’s system of redistricting. And in response I have always called their proposals half-baked. More

Share this article


(comments below)


(comments below)


  1. Let’s leave the partisan aspect out of redistricting for a moment –

    My favorite solution is to require districts to take some round or square geometric shape, with no precincts jutting out halfway across the state. If the math experts in the state could find some formula for a “regular” district shape, and then apply it, we’d have better redistricting.

    The problem with our current “irregularly shaped” districts is people don’t know their legislator because he or she often lives on the other side of the state. Districts should represent communities, not clusters of precincts in a snake shape.

    Also, I suppose if you require that districts take a uniform shape (roughly) and you’ll make it harder for partisans to throw republicans or democrats into incumbents’ districts.

  2. E M Risse Avatar


    That geometric shape should be “pie shaped” as we note in “The Shape of Richmonds Futue.”


    The reason you have seen no rational proposals is that no one in your local paper has yet suggested Fundamental Change in the governance sturcture so that there is a level of governance for each of the organic levels of settlemetn pattern. (Seven below the New Urban Region scale.)

  3. Paul — I have a formula. I served on Harrisonburg’s Planning Commission. We wanted someway to rule out “panhandle” lots that were designed just to get around some rule. I suggested that all you had to do was take the length of the perimeter, square it and divide by the area of the lot. A square would give you the number 16, something that is 4 times as long as it is wide would give 25, so I suggested that lots that had the above ratio greater than 30 would need special permission. It all seemed very obvious to me (I’m a physics professor at JMU.) Everybody looked at me as if I were crazy. That was too hard to calculate.

    Still, I think you could compare election districts on the same basis. They have much more “wiggly” boundaries so it might be that one would accept higher numbers than 30. Maybe 100 or so.

  4. That makes sense…the ratio of the perimeter vs the diameter would tell you how gerrymandered a district was.

    Perimeter/Diameter = Gerrymander ratio

    The higher the ratio, the more ridiculous “panhandles” the district has.

  5. Anonymous Avatar

    Bees use a hexagonal shape to most efficiently allocate space to the “districts” within their hives. Hexagonal shapes would more nearly fit the diagonal western border with less distortion. Round shapes woud require considerable distortion to make a fit.

    Regardless of the initial geometric shape chosen it would not be too difficult to write a program which would generate districts that meet the criterion of 1.) Maintaining the Geometric shape as near as possible 2.)keeping the size of the districts as near even as possible 3.) keeping the population of the districts even.

    If the resulting districts achieve the result noted by Paul that districts would no longer be predictably Republican or Democrat, then partisanship would be dealt a serious blow. It would become necessary to build concensus and coalitions in order to do more legislative damage to the citizens.

    By itself this is enough reason to forge ahead.

    Dorn’s suggestion as to how to evaluate districts that are bent out of shape offers an interesting approach that could easily be incorporated into an algorithm which would produce some proposed districting plan.

    It is entirely possible that the computer would produce a correct result which is, however unacceptable for some reason. Humans would have to retain the ability to monkey with the results to some extent.

    To solve this the computer could produce a first and second result. Legislators could monkey with the first answer so long as it did not result in an answer that was worse than the second answer. Better and worse would be determined by using the average of some figure such as dorn’s statistic.

    As usual, everyone here is on topic except Ed Risse. By his logic no answer is rational that does not result in Fundamental Change.

    The question at hand here is how to make government more responsive and less expensive. One way to do that is to spend less time arguing about how representation is divided and a geometric rationalization is as good a place as any to start.

    Ray Hyde
    Delaplane, VA

Leave a Reply