Monday, November 14, 2011


Roll Call has pictures of what it considers the five ugliest House districts coming out of the current round of redistricting.

What makes a district "ugly"? Mainly, that it's not compact; to a lesser extent, the specific complaint that one or more sections of the district are connected by only a thin bridge. Districts must be contiguous (in fact, that, and rigorously equal populations are about the only strict rules out there), but in one Ohio district along Lake Erie, Democrats are arguing that "when Crane Creek State Park beach is covered during high water" the district doesn't even achieve that.

My feeling? I'm all for ugly districts. Compactness is, to be blunt, a stupid criterion for drawing district lines. Communities of interest, sure; other political jurisdictions, okay; and as far as I'm concerned, whatever political goals the majority has in mind are fine by me. Note that there's a real limit on partisan gerrymandering because while parties collectively want to win as many seats as possible, individual party politicians want very safe districts -- and safe districts require "wasting" votes won in landslides that could have been used to make some other district better for the party. Since politicians are the ones who pass redistricting maps in most cases, the most likely results are bipartisan gerrymanders, which maximize safety for incumbents of both parties at the expense of maximizing partisan gains. That doesn't always happen, but it happens quite often, even when one party controls the entire process.

At any rate, the question of district shape is really separable from the question of whether political goals should be used to write districts, although in practice reformers use the prejudice against "ugly" districts as a constraint against political motivations that they can't get the courts to knock out. Purely on their own merits, however, I see nothing at all wrong with ugly districts.


  1. Don't you think it detracts from a community sharing a common interest achieving it's goals when far-flung constituents have substantial influence?

    Maybe you're only considering the partisan issues

  2. As long as you have equal population, you're going to wind up with districts that don't match up well to traditional communities. So you wind up with choices such as which is better: connecting two small cities by a narrow ribbon (ugly), or having a compact district that contains one of those cities and surrounding suburb, exurb, and rural areas.

  3. After the 2000 census, New York lost two Congressional seats. As you suggest, the Democrats (who ran the Assembly, i.e., the lower house) and the Republicans (who ran the Senate) got together and decided to eliminate one seat from each party. Since the common conception is that Downstate (New York City and environs) is or ought to be Democratic and Upstate is or ought to be Republican, they targeted the outliers, a Downstate Republican and an Upstate Democrat. (I guess it makes the political world easier to deal with.) The Downstater was easy since a Hudson Valley Republican was retiring anyhow. For the Upstater, they picked Louise Slaughter of Rochester, the state's third-largest city. What had been Rochester's district was divided up among four neighboring rural/suburban districts (including one that now runs in a narrow strip along the Lake Ontario beach to the Buffalo suburbs). A decade later, however, Louise Slaughter is still in Congress. That's always made me wonder if jerrymandering is really as effective as everyone claims.

  4. As you may already know, Scott, a similar redistricting deal is set to occur in the same regions due to the 2010 census, in which NY again lost 2 house seats. Weiner's old district is being phased out downstate, and one more seat is being removed from the weirdly configured upstate Western NY area of Buffalo to Rochester to Syracuse:

    It's not yet clear what will happen, but there's a chance it might turn into a configuration of districts that doesn't split up the Rochester metro area in such an odd way. To do this, they may once again get rid of Slaughter's existing district. We'll see if she once again survives as an incumbent with a frequently changing district.

  5. If the geographic component of some of these districts is going to be so ludicrous as to be essentially meaningless, then why not just dump the geography and use some other demographic factor to engineer the desired outcomes?

    On the other hand, if we want to maintain a geographic component, then the geography ought to make sense.

  6. JB: ceteris paribus, compact districts should lead to voters being more connected to their district. Two reasons come to mind:
    1) constituents can tell what district they live in fairly easily
    2) constituents are more able to talk about their MC with their neigbors, because they are much more likey to live in the same district. (this 2nd one is because one definition of "compact" would involve the shortest total amount of length of district borders in the state, and the more border, the more likely you are to live within a short distance of one)

    However, I think these effects are minor. And, that ceteris paribus is HUGE. It's not like folks are RANDOMLY drawing convoluted lines; those lines that are non-compact are like that in order to achieve some goal. Remember, compactness IS a defense of lines in court, so anyone who draws lines that aren't compact is risking (albeit not too much) court intervention, so they have a reason to do so. Thus, the problem (or benefit!) isn't the non-compact lines per se, but the reason they are drawn that way.

  7. Matt,

    Yup, yup, and yup. I know that compactness is something the courts consider, but on the whole I'd probably prefer that they didn't.


Note: Only a member of this blog may post a comment.

Who links to my website?