Sunday, October 6, 2013

Fifty plus One

Recent commentary on the current government shutdown and the fast-approaching debt ceiling crisis have often cited gerrymandering as the proximate cause for the rise of a bloc of conservative congressmen pushing the Republican conference to confrontation. As often, the media conventional wisdom badly misses the mark.

Sizable Republican gains in the 2010 midterms at the state level paved the way for Republican redistricting in swings states like Ohio, Michigan and Pennsylvania. Despite Obama winning each of these states by varying margins, the House delegations are overwhelmingly Republican (12-4 in Ohio, 9-5 in Michigan, and 13-5 in Pennsylvania). In 2012, total votes cast in House elections for Democrats exceeded Republicans by 1.7 million, yet Republicans held the House 234-201.

While gerrymandering in these and other states limited Democratic gains in the House, demographics and sorting are largely to blame for the differences between the overall vote totals and the division of seats in the House. Democratic voters are increasingly concentrated in urban areas, and, therefore, even under a "fair" redistricting, Republicans would likely outperform their overall vote totals given that Republican votes are more geographically diffuse (see this blog post from John Sides).

However, the gerrymander has very little to do with the dynamics driving the current standoff in Washington. Indeed, since the object of a gerrymander is to spread your voters across several districts, a gerrymander should generate more marginal districts with lawmakers who have stronger incentives to appeal to moderates and members of the opposite party. Many of the Republican House members expressing reservations about the shutdown hail from gerrymandered districts (see this roll call).

Much of the analysis surrounding the shutdown has also suggested that because of increased polarization of congressional districts, political pressure this time is less likely to force a resolution compared with the last government shutdown in the 1990s (see this article from National Journal). As economists like to emphasize, think marginal not average. That the average Republican district has become redder matters less than the number of marginal Republican districts. To the extent that a government shutdown benefits Democrats in 2014, these gerrymandered districts may give Democrats a better chance to take control of the House. In a less gerrymandered world, Republicans majority would likely be narrower but more durable. By broadening the playing field, a gerrymander gives Democrats a better opportunity to exploit any backlash and retake the House. Their prospective majority (like the 2008 majority) would be precarious, but all it takes is 50 + 1 to run the House.

See this blog post by Nate Cohn making a similar point.