In the previous chapter, we presented a model of the role of the courts in the 1960s redistricting process. In this chapter we test two of the key predictions of that model, those regarding bias and responsiveness.

The first section deals with the operational details of our tests. The second section shows that Democratic courts supervised plans that had more pro-Democratic bias, while Republican courts supervised plans that had more pro-Republican bias, controlling for partisan control of state government. It also shows that responsiveness tended to be higher in plans written by unified states facing friendly rather than hostile courts. Both of the bias-responsiveness predictions articulated in the previous chapter, in other words, are supported by the data.

The third section shows that there were many more Democratic than Republican courts with jurisdiction over congressional redistricting cases in the 1960s – part of the legacy of Democratic hegemony in the New Deal era. Putting the predominance of Democratic courts together with the impact of court partisanship on plan bias helps explain why Republican partisan gerrymanders in the 1960s produced such low levels of pro-Republican bias; why bipartisan Republican plans in the 1960s produced lower levels of pro-Republican bias than they had earlier; and thus why what had been about a 6% pro-Republican bias outside the South abruptly disappeared.

The fourth and fifth sections consider the extent to which judges appear to have pursued partisan interests and the role – dynamic or constrained – of the judiciary. The sixth section concludes.

operational matters

Obviously, the Democratic seat share in a state should increase as the Democratic vote share in that state increases.