Our argument from the beginning of the book has been that partisan bias can only be explained in political terms. In short, we argue that pro-Republican bias can be found only where the Republicans control the districting process. Furthermore, the degree of partisan bias we observe has increased very significantly in the 2010 districting round following Vieth v. Jubelirer (2004), even in states where state legislatures had already drawn biased districts in the previous round.
In this chapter, we test these claims. In Chapter 3, we saw that the degree of partisan bias increased sharply in the districting round following the 2010 Census. We also identified the states where this bias is greatest. In the previous chapter, we considered whether this increase in bias could be explained by demographic factors, such as the urban concentration of Democratic voters or the need to draw majority-minority districts. We found that these factors cannot explain the pattern of bias we observe. Here we return to the story we laid out in Chapter 1. In the 2000 districting round, there were various complaints of partisan gerrymandering. In response to one of these complaints in the case of Pennsylvania, the Supreme Court in Vieth v. Jubelirer (2004) gave a clear signal that the courts would not intervene in partisan gerrymandering cases. Following this, state governments pursued partisan advantage in the 2010 districting round to a far greater degree than they had in the previous round.
If this story is accurate, there are two things that we would expect to observe. First, the pattern of partisan bias should follow a political logic. We should find partisan bias in those states where the state government has the ability to pass partisan districting plans and has an incentive to do so. That is to say, we would expect to find bias where one party has control over the entire districting process and where drawing biased districts brings it an advantage (we will see that this is not always the case). Second, we would expect bias to increase between 2000 and 2010 in states where one party has control. That is to say, we expect bias to increase in states where the Republicans controlled the process in both 2000 and 2010, not just in states where the Republicans gained control in 2010.