Friday, February 22, 2013

Safe Seats Do Not Mean Gerrymandering

I guess we're not explaining this well enough.

Dave Weigel:
I enjoy the new, #slatepitchy argument that gerrymandering is overrated as an issue, and that it doesn't influence whether members moderate their votes or not, but sequestration's putting that to a test.
Kevin Drum:
Hold on a second. Who's saying that? The argument I've heard making the rounds lately is that gerrymandering is probably responsible for a fairly modest change in the number of House seats Republicans won last year...But who's been saying that gerrymandering has no effect on whether members feel any need to moderate their votes or their rhetoric? What have I missed?
What both of these normally astute writers are missing is that gerrymandering (drawing district lines for political reasons) is only one reason for safe seats.

Why are there safe seats? Lots of reasons, including:

* Geographic distribution of partisans. If Democrats and Republicans live in different neighborhoods, or different cities, or are differently drawn from urban, suburban, and rural areas, then you're very likely to get lots of safe partisan seats unless districts are very specifically drawn to prevent it.

* Partisan voting. As voting by party increases, any Republican in a Republican seat becomes safer, as do Democrats in Democratic seats.

* Incumbency. Even Republicans in Democratic districts can be safe if incumbency advantages are strong enough.

* Information. To the extent that, say, Republicans get most of their political information from the GOP-aligned partisan press, they're very likely to support their current Republican Member of Congress. Note that the old "neutral" press tended to be extremely incumbent-friendly to their Congressional delegations.

* Gerrymandering. Be careful, though. The kind of gerrymandering that tends to create safe seats, especially for the majority party, is bipartisan gerrymandering -- the kind in which incumbents from both parties cooperate to draft lines that protect all incumbents. See California, 2002-2010, for a famous example. Partisan gerrymandering tends to yield more seats for the majority party at the cost of each of those seats being relatively less safe. However, partisan gerrymandering, to the extent it's successful, should make the "victim" party have safe seats; the idea is to pack as many of the minority party into as few seats as possible, but to use majority voters more efficiently.

So: yes, gerrymandering has effects, although they're less impressive than many think. But when it comes to safe seats, gerrymandering on the whole won't be a major factor, and partisan gerrymanders in particular shouldn't (overall) produce more truly safe seats for the majority party. Of course, it may produce some; one of Weigel's example is a Member whose seat was made safer through partisan redistricting. On the other hand, however, perhaps without that gerrymander (and I haven't looked into what happened in this case) the seat wouldn't have been a safe Republican seat at all; maybe it would have been a Democratic seat.

Generally, this reminder: for whatever reason, people really, really, really like attributing effects to gerrymanders. So if you come across one of these, by careful: it's very possible that something else is really at work.


  1. Picking up point one, there, presumably a lot of the apparent 'structuraly' GOP tilt in the House comes from the slight inherent bias towards rural representation in the system?

  2. Every time a pundit wants to blame gerrymandering for some political ill, they should ask themselves, "Does this happen in the Senate too?" As it happens, we have loads of crazies in the Senate. Ted Cruz is trying to bring back McCarthyism. Jim Inhofe expressed support for birth certificate conspiracies until he had to backpedal to avoid political backlash, and then he blamed Obama for the spread of the conspiracy theories he had personally supported, not to mention his accusation the Obama wants to eliminate the military so that so that the Chinese can shoot missiles at us and then mount a land invasion (for some reason they apparently have some deep incentive to do that in Senator Inhofe's fantasies). Even supposedly "moderate" senators like Lindsey Graham and John McCain have been promulgating absolutely crazy Benghazi conspiracy theories, and I'm not even sure if it's an attempt to appeal to the far right of their party in primaries or if they're acting out of personal spite. Senate Minority Leader Mitch McConnell has openly admitted to making a strategic choice of deliberately undermining the functioning of his own legislative chamber. Senate Republicans as a whole have refused to ratify uncontroversial treaties because of conspiracy theories about a UN takeover. There's an unprecedented permanent and universal filibuster on all Senate activity by the minority party. And it wasn't all that long ago that we had Kyl, Bunning, and DeMint in the Senate. Had Akin and Mourdock waited until after the election to publicly air the crazy on their opinions on rape, what sort of conversation would we be having now? Would the House be seen as that different from the Senate? Of course, it is true that having safe districts means that those crazy candidates who get out of Republican primaries can actually serve in Congress, and you have a few of those absolutely insane Republican representatives in the House who have no inhibitions about mouthing insanities to any reporter willing to listen, but it seems to me that John Boehner isn't any crazier than Mitch McConnell. The biggest difference is that Mitch just needs to find five senators willing to compromise with Democrats on need-to-pass measures, while Boehner is expected to produce a majority of his crazy House caucus. The Senate is filled with crazy Republicans who are doing crazy things, none of whom got there because of gerrymandering.

    1. Isn't the Senate heavily gerrymandered by design?

      The Senate massively favours representation of voters in underpopulated rural places, and heavily discounts voters from densely populated places. In the current political system, that adds up to a structural tilt to the right, which in fact we see; it's why you still get so many conservative Democrats, now that the GOP has abandoned the middle ground altogether. In the Senate it's less party gerrymandering than a general political tilt to the conservative end, built into the allocation system; originally, as a protection for slave-holders.

  3. It sure wasn't "gerrymandering" of state lines that resulted in more than 99% of campaign spending on ads in the final 7 months of the 2012 campaign being devoted to only 10 states -- and those being the same 10 states that were the only ones that Romney and Obama visited for campaign rallies after the conventions and all being among the 13 most favored states in 2008.

    It wasn't "gerrymandering" that resulted in more than 40 of California's congressional districts being completely locked down for one party after an independent commission drew the lines? (As spelled out in this blog: )

    It's winner take all voting rules -- rules that can be changed statute, as shown for all states at FairVoting.US

    1. I mean, yeah, it wasn't gerrymandering. People with similar political views tend to group together, which produces a tendency toward naturally producing congressional districts and states that are more polarized than the country as a whole, which is different from gerrymandering. That's kind of Jonathan's point. It is true you could reduce or eliminate this effect by moving away from a first past the post voting system, but again, this is unrelated to gerrymandering.

  4. I thought the idea of gerrymandering was to create fairly safe districts for your party and ridiculously safe districts for your opponent's. So a Republican would try to create lots of districts that are, say, 60% Republican and a few districts that are 90% Democratic.

    My concern about gerrymandering is not that it makes for more extreme candidates, but just that it is less representational. (Silly me, still believing in democracy!) I don't actually see the problem with extreme candidates. As a Democrat, I would prefer my representatives to be more extreme. Isn't that the whole point? After all, we compromise to get something we really want in exchange for something we can live with (or without). The problem is not extremism; it is that the Republicans don't accept their own proposals when they are offered.

    For example: Ari Melber on Republican Nihilism

    1. If one *defines* gerrymandering to mean the manipulation of district lines to give one party an advantage it shouldn't have (using the party's share of the total popular vote in the state) then of course gerrymandering by definition is less representational. The problem is that many people define "gerrymandering" as meaning bizarre lines--yet maps with convoluted district lines are not *necessarily* less representative than maps with compact districts.

  5. This is actually a great reminder that in politics, it is easy to simplify things to such a degree as to miss out on the whole picture. Looking at just gerrymandering as the whole issues is wrong and this post actually illustrates all the different factors that go into how seats are maintained, protected, and won by the parties.

    -Ali Olomi, UCLA


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

Who links to my website?