Myths about redistricting

Kevin Drum makes a point about the overstated effects of redistricting that I've been meaning to address:

Is gerrymandering responsible for the fact that it's virtually impossible nowadays to defeat an incumbent in the House of Representatives? Reporters and pundits seem to accept this without question, but academic research suggests otherwise. For example, Alan Abramowitz, an Emory political science professor who's studied the decline in competitive seats, recently published a paper concluding that redistricting has had "little to do with the recent decline in competition in House elections. Other developments, such as the growing financial advantage of incumbents and increasing partisanship in the electorate, appear to be more responsible." He figures that only 12% of the decline in marginal districts has been a result of redistricting.

The idea that redistricting is responsible for non-competitive House races is pervasive. For instance, Juliet Eilperin is a national reporter for the Washington Post who frequently makes this argument. In last month's Atlantic, she writes that "Thanks to today's expertly drawn congressional districts, most lawmakers represent seats that are either overwhelmingly Republican or overwhelmingly Democratic." She's made similar claims in her book, the Boston Globe and on NPR's "Fresh Air".

However, as Drum notes, the evidence for this claim is weak. In addition to the Abramowitz finding, John N. Friedman and Richard T. Holden of Harvard have released a working paper suggesting that redistricting has actually reduced incumbent re-election rates (see their New Republic Online article about it). There are many factors contributing to the increase in incumbent security; redistricting is at most a small component of the change.

One problem is more basic -- when people tend to live near other people with similar political views, it creates safe districts. You can't draw many competitive House districts in rural Texas or New York City. Bruce Reed, the DLC official who blogs on Slate, doesn't seem to understand this:

Yet even after poring over this week's bleak poll numbers, Karl Rove isn't completely crazy to imagine his party holding onto the House in November. Democrats aren't likely to win the popular vote by seven points, let alone 17. But what's really keeping Rove's dark hopes alive is the Safehouse that Jack and Tom Built—the firewall of safe districts that could enable the Republican party to survive what would otherwise be a China-syndrome political meltdown.

If congressional districts were truly representative, a party that won a seven-point victory in the popular vote would walk away with a 7 percent edge in the 435-member House of Representatives, or roughly a 30-seat majority. For Democrats, that would represent a pickup of around 45 seats...

Rigged districts defeat the very reason we have a House of Representatives in the first place. The founders wanted one chamber that would be held accountable to the popular will every two years. When the Electoral College is wrong, at least it's a wrong the framers intended.

What Reed is talking about is a system of proportional representation in which legislative seats are allocated in proportion to each party's national vote. That's decidedly not what the Founders had in mind. Given that people distribute themselves unequally across the country, our single-member districts will inevitably fall short of the "ideal" proportional representation outcome regardless of whether gerrymandering has taken place.

Comments

[Seth]

our single-member districts will inevitably fall short of the "ideal" proportional representation outcome regardless of whether gerrymandering has taken place.

Never moreso do we fall short than DC, the last bastion of taxation without voting representation in the US.

Wyoming = 1 congressman, 2 Senators. Population 500,000
DC = 1 congresswoman (non-voting), no Senators. Population 550,000

[Ben]

Where in the Constitution does it say anything about single member districts? All it says are that Congressmen are selected by the states, right? So there's no reason a state couldn't have proportional representation for its Congresspeople, or multi-member districts, or any number of other potentially interesting voting mechanisms.

[Brendan Nyhan]

Good point - the requirement for single-member districts is statutory and not in the Constitution, but Madison did write in the Federalist Papers about single-member districts (see here for more).

[kammanl]

Delay apparently got what he wanted in Texas by redistricting. Would he have got it without redistricting? Doubtful.

[Thomas Nephew]

Moreover, even though Madison wrote about single member districts, and even if he had managed to enshrine them in the Constitution itself instead of in his "users manual"... that doesn't mean the effects of districting and redistricting are overstated. It suggests instead those effects are a feature, not a bug. And it would suggest to at least some that such "features" are at odds with what they want from their political system. Just because the Founders had something in mind doesn't prove it's a good thing they had in mind.

Of course reality will always fall short of the ideal. The question is, how far can it fall short before setting off alarm bells? It's a political choice to say "really, really far" (or worse yet to just redefine the ideal), and it's not one that's very complimentary to or solicitous of your fellow citizens.

Re statistical evidence on gerrymandering: the key, I think, is comparing the Senate (States can't be gerrymandered) to the House (CDs can be gerrymandered). Of late, the Senate is actually less friendly to incumbents than the House, IIRC.