Redistricting aficionados sat transfixed at their computer screens last Friday, waiting for the California redistricting commission to announce the first draft of the newly drawn lines for the Golden State. The stakes are high. In 2000, the state legislature adopted a brilliant incumbent protection map. It worked; party control of a House seat switched in only one of 265 general elections held from 2002 through 2010.
To prevent this from recurring in the 2010s, California voters approved a ballot initiative last year that transferred the power to draw legislative lines from the state legislature to an independent redistricting commission. That body was instructed not to consider the homes of existing representatives, nor were they to pay attention to political demographics. Instead, the goal was to maintain reasonably compact districts and keep communities of interest together. It was expected that this would scramble the California delegation.
The commission's draft did not disappoint. It appears to create 12 open seats, placing over one-third of California's current members of Congress in districts now represented by another incumbent. According to Daily Kos Elections number cruncher Dave Jarman, only a handful of the new districts contain more than 75 percent of their former constituents. The commission even went so far as to abandon the traditional numbering system for districts, adopting descriptive names such as "East LA-Boyle Heights" and "Foothills" rather than "4th" or "33rd" district.
From a partisan perspective, it is more of a mixed bag. Several suburban Republicans saw their districts moved more toward the inner suburbs and even into the cities -- and hence made more Democratic. David Dreier, Elton Gallegly and a few others will have a choice between retiring or moving. In that sense, Republicans probably got the worst end of the deal.
But a more careful exploration of the districts reveals that what the commission really accomplished was exactly what it was supposed to do: Generate a more competitive map. A district-by-district breakdown like the one I did for Illinois is too cumbersome for a state with 53 congressional districts. So let's instead try a big-picture approach by first looking at how districts voted for Barack Obama vs. John McCain under both the current map and new map:
This chart shows the 2008 presidential results by Obama's margin in each district, relative to his national margin. This is like PVI, but allows for finer gradations. President Obama won by about seven points nationally, so a district that he won by nine points would be scored D+2, while one that he won by six points would be R+1, and so forth. Republicans were competitive in districts up to about D+8 or so in the 2010 midterms, while Democrats made similarly Republican districts competitive in 2006 and 2008.
This table gives us a sense of the overall universe of competitive seats in both very good and very bad Democratic years. The data were crunched by Democratic political consultants Redistricting Partners.
There are fewer Republican-leaning districts overall, but there is only one more Democratic-leaning district. For the most part, the commission took Republican-leaning districts and made them into districts that are tied politically. We'd expect Democrats to pick up two or three seats in this scenario in a neutral year. But notice that the number of safe Democratic seats drops significantly, and that there is an accompanying increase in the number of barely Democratic seats. And some of these seats, like that of 74-year old Democratic Rep. Lois Capps, will almost certainly be open in the next decade.
In other words, in a year that tilts toward the GOP, Republicans could see a three-to-four seat pickup for themselves, especially if there are some poorly timed Democratic retirements.
But Obama's race was arguably a unique event, taking place in the shadow of a housing collapse that devastated California and helped swing exurban counties toward the Democrats. His candidacy also generated unusually strong turnout among the state's minority groups. We should also look at the state's 2010 gubernatorial race for another marker.
Although 2010 was a very good year for Republicans, the GOP tsunami didn't reach fully to the Golden State. The party's Senate and gubernatorial nominees lost by double digits, and the gubernatorial nominee in particular was buffeted by a late-breaking revelation that she had employed an illegal immigrant. Moreover, party identification in the electorate was roughly unchanged: 42 percent Democratic and 31 percent Republican in 2010, vs. 42 percent Democratic and 30 percent Republican in 2008.
This chart just looks at the margin in the gubernatorial race in each district, because we can't use PVI when a candidate doesn't run nationally. Once again, the new map has four fewer districts that lean toward Republicans. But there are also more even districts, along with fewer safely Democratic ones. The fact that the results here so closely mirror what we saw with the presidential numbers suggests that it is correct to intuit that these district lines will produce a few more Democratic representatives in a neutral environment, but would give the Republicans more opportunities in a year like 2010.
Of course, this is just the initial draft, and the commission may make multiple changes in its next version, due in July. In particular, there will be pressure to create a few more heavily Latino districts, which should make for a map that is overall more favorable to Republicans by further "packing" Latino voters.
And we should always remember that demographics aren't set in stone; if the suburbs shift back toward the GOP, or if Latino voters begin to vote in proportion to their share of the population -- or vote more conservatively -- then the political makeup of the delegation would shift again. The relatively close nature of these district lines all but ensures that slight demographic changes could result in big shifts in elections over the course of the decade.