The Most Important Redistricting Case in 50 Years
In a pair of cases decided in 1964, the Supreme Court of the United States famously established the “one person, one vote” test. This meant that all congressional districts would be required to have the same number of people, while state legislative districts must have roughly the same number. The consequences of those decisions were both immediate and far-reaching. A wave of mid-decade redistricting swept the country, as virtually every congressional and legislative district had to be, at a minimum, tweaked to account for population discrepancies. Rural districts in particular lost representation, while the depopulation of urban centers helped usher in the rise of the suburbs in Congress.
Last week, the Supreme Court shocked watchers by agreeing to hear a case that could have consequences of a similar magnitude. In 1966, in a follow-up to the Reynolds v. Sims decision, the court had held that states did not necessarily need to use persons as the basis for their representation schemes. Since then the court has at times been asked to adopt various different metrics. It generally resisted these entreaties, although Justice Clarence Thomas has, at times, urged the court to take up these cases.
So most were caught off guard when the court decided to take up Evenwel v. Abbott. The plaintiffs in that case asked the court to clarify that only citizens should be counted for purposes of drawing legislative districts. The “why” of this is a bit complex, but it grows out of a (superficial, in my mind) tension between the 14th Amendment, which apportions voting districts on the basis of population, and the Voting Rights Act, which requires that states ensure there are a sufficient number of citizens of voting age in a given group to enable that group to elect a candidate of its choice.
If the court were to find for the plaintiffs – and it seems unlikely that the court would have gratuitously taken up this case, absent a circuit split, if there weren’t some substantial support for the plaintiffs’ position – it would mean that, once again, virtually every legislative and congressional district in the country would have to be redrawn (although this would not, as some have suggested, affect apportionment – i.e. the number of seats allocated to each state). This would occur at a time when Republicans control a record-high number of state legislatures and a majority of state governments. Republicans would be able to update their maps to account for changes in political orientations in their states since the previous round of redistricting.
But this would have implications for Democratic-controlled states as well. Consider that in 2012, counties with high citizen populations were more likely to vote for Mitt Romney (the t-stat is 9.047). Of the 35 states with four or more congressional districts, there was a statistically significant, positive correlation between the share of county residents who were citizens and the share of voters who cast ballots for Mitt Romney in 18 of the states, most of which are among our largest: California, Colorado, Florida, Illinois, Indiana, Kentucky, Maryland, Michigan, Minnesota, Missouri, Nevada, New York, Ohio, Oregon, Pennsylvania, Texas, Virginia and Wisconsin.
By ruling that states had to ensure an equal number of eligible voters in districts -- rather than persons -- the court would force even Democratic map makers to push districts out of heavily Democratic areas and into Republican areas. Just 78 percent of the residents of heavily Democratic Queens County are citizens, and 83 percent of similarly Democratic Kings County residents are citizens. In neighboring Nassau County, however, 91 percent of residents are citizens. In other words, 20 percent of the population in districts in Queens and Brooklyn would no longer count. To populate these districts, line drawers would be forced to push districts into Nassau County. Democrats would have a choice between weakening the Queens/Brooklyn districts, or making the remaining Nassau County districts more Republican.
To see how this plays out in practice, consider the lawsuit brought in 2012 challenging the apportionment of the New York State Senate. The current map places nine senate districts on Long Island, 26 districts in New York City, and 28 districts upstate. These districts are of similar populations.
But if we look at citizens of voting age (or CVAP), we see some pretty wide disparities. The average CVAP for the Long Island districts is 215,436 persons. The average CVAP for the New York districts is 191,133 persons. The average CVAP for the upstate districts is 217, 759 persons. There are other ways to look at the data, but the upshot is that a successful lawsuit would probably move two senate seats out of reliably Democratic New York City and into upstate New York. New York’s congressional districts are similarly apportioned between New York City, Long Island, and Upstate, so adopting CVAP would likely move a congressional district out of the city and into the swing areas upstate.
There are other examples here: Harry Enten and Dave Wasserman note that only 41 percent of the residents of California’s 34th District (downtown Los Angeles) are adult citizens; this district would probably have to be combined with most of a neighboring district (probably the 40th, where only about 40 percent of the population is adult citizens) to generate a full Hispanic-majority district. This would probably result in a new district placed in swingier areas of the state. Republican David Valadao’s Central Valley district is majority non-adult citizen; Democrats would love to weaken him but would have a hard time doing so while also protecting Jim Costa’s Fresno-based district. Weakening Loretta Sanchez’s district would be almost unavoidable. The five districts that abut the Rio Grande River in Texas have high non-citizen populations; one would likely be eliminated and, if Republicans have their way, transferred to the heavily Republicans suburbs of one of the major cities.
You get the point. As Enten and Wasserman point out, of the 50 districts with the lowest share of adult citizens, 82 percent are represented by Democrats, while Republicans represent 38 of the 50 districts with the highest share of adult citizens. Redistricting would probably move five or 10 House seats toward the Republicans, with proportional gains likely in the state legislatures.
I do think it is unlikely that the court will rule for the plaintiffs, but then again, I wouldn’t have thought that it would take this case up without a lower court forcing its hand. If it does take the case up, it could have serious consequences for the next president’s term.