The record of civil rights roll-call votes in the House of Representatives has yet to be systematically explained or predicted. In particular, it is not clear why House members sometimes appear to have a great deal of independence from constituency, but at other times are limited by negative public assessments. I contend that the key is the variation in content of different types of civil rights bills, and construct a categorical scale which rates bills voted on in the House from 1957-1991 according to potential costs (whether actual or perceived) to white Americans. The expectation that an increase in category will result in decreased likelihood of passage, and significantly smaller supporting coalitions, is supported. While partisanship also emerges as a significant determinant of roll calls, a more complete explanation is arrived at when it is considered in conjunction with the legislative classification variable.
The legislation of civil rights has been one of the most visible endeavors the U.S. House of Representatives has undertaken. Much attention has been paid to the widely perceived transition from a body virtually held hostage by conservative southern Democrats to one which triumphantly passed the 1964 Civil Rights Act and the 1965 Voting Rights Act. Yet reconciling this historical role with dominant theories of how the House functions is problematic. Although the House is presumably the governmental institution most responsive to public demand, there has been no systematic attempt to explicate the public's role in the passage or failure of civil rights bills. Anecdotal evidence sometimes suggests that members of Congress have been opinion leaders in this area-convincing the general public that passing civil rights legislation was simply the right thing to do. Other accounts, however, portray this as an area in which the public speaks and Congress merely reacts.
Conventional wisdom's vague attribution of early legislative gains to the civil rights movement and/or the liberal nature of the times, avoids the important question of variation in success rates. Why was the movement able to achieve legislative success with some bills but not others? And why was meaningful policy still being passed in the 1990s when conventional wisdom held that the political environment had become more conservative? Furthermore, the well-accepted notion that, since the early 1960s, non-southern Democrats tend to support civil rights legislation and Republicans oppose it also does not tell the whole story. The record shows that roll-call variations are not fully explained by a simple partisan model.
I propose that a key omission in previous studies is an analysis of the considerable differences among civil rights bills considered by the House of Representatives. In particular, the variation in costs (whether actual or perceived) the legislation represents to non-southern whites, and the opposition this engenders, should be examined.1 After constructing a typology which classifies all major civil rights legislation considered by the House from 1957 through 1991 according to these varying costs, I find support for the hypothesis that as opposition among non-southern whites increases, the likelihood of passage and the size of the (non-southern) supporting coalition declines. Party emerges as an important variable as well, but the classification of legislation ultimately provides a richer explanation of the variation in rollcall outcomes than does a simple reliance on partisanship.
The need for a systematic means of classifying civil rights legislation is clearly underscored by the fact that neither of the principal models of roll-call behavior provides a satisfactory explanation when applied to the House record in this area. The distributive approach asserts that roll-call voting decisions are overwhelmingly fueled by reelection considerations (Mayhew 1974). Congressmen tend to support policies that have particularized benefits and widely dispersed costs. …