Hostname: page-component-586b7cd67f-gb8f7 Total loading time: 0 Render date: 2024-11-27T20:46:35.819Z Has data issue: false hasContentIssue false
  • English
  • Français

Scaling Models for Legislative Roll-Call Analysis*

Published online by Cambridge University Press:  01 August 2014

Herbert F. Weisberg
Herbert F. Weisberg*
Affiliation:
University of Michigan
Get access Rights & Permissions [Opens in a new window]

Abstract

Guttman scaling is the usual procedure for scaling legislative roll-call votes. This paper calls attention to an alternative scaling model—the proximity model. Under this model, legislators approve a consecutive set of items on the scale, without the cumulation required by the Guttman scale. Circumstances under which proximity voting is likely are discussed. Congressional voting on the Compromise of 1850 is analyzed in detail to illustrate the proximity model and to emphasize the possibility of obtaining faulty inferences if one uses the Guttman scale model when it is incorrect. Guttman scaling has been successful for contemporary Congresses, but the proximity model is seen to underlie some issues in the early 1970s. Proximity scaling is not limited to the legislative realm; it can be used in survey analysis and in attitudinal research more generally.

Type
Articles
Information
American Political Science Review , Volume 66 , Issue 4 , December 1972 , pp. 1306 - 1315
Copyright
Copyright © American Political Science Association 1972

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Footnotes

*

The research reported here is based on a section of the author's doctoral dissertation, revised and extended under a postdoctoral fellowship from The University of Michigan's Rackham School of Graduate Studies. This work has benefited from the generous advice and comments of Professors Donald E. Stokes, Hayward R. Alker, Herbert B. Asher, Paul A. Beck, Jerome M. Clubb, Clyde H. Coombs, Warren E. Miller, Richard G. Niemi, and Jerrold G. Rusk.

References

1 In some instances it may be necessary to reverse items since, for example, a vote to recommit a motion is generally equivalent to a vote against its passage.

2 The Guttman scaling approach is detailed in Anderson, Lee F., Watts, Meredith W. Jr., and Wilcox, Allen, Legislative Roll-Call Analysis (Evanston, Illinois: Northwestern University Press, 1966), pp. 89–121Google Scholar.

3 This model is based on Coombs, Clyde H., A Theory of Data (New York: John Wiley & Sons, 1964), pp. 295–302Google Scholar. Its applicability in political science was first suggested by Hayward R. Alker, Jr. in the context of studying attitudes toward political candidates. See the discussion of a “point scale” in Alker, , “Statistics and Politics,” in Lipset, Seymour Martin, ed., Politics and the Social Sciences (New York: Oxford University Press, 1969), pp. 276, 77Google Scholar.

4 Additional procedures for scaling under this model are presented in Weisberg, Herbert F., “Dimensional Analysis of Legislative Roll Calls” (Unpublished Ph.D. dissertation, University of Michigan, 1968), pp. 94–103Google Scholar. It is also shown there that conventional correlation coefficients (and consequently factor analysis) are inappropriate under the proximity model.

5 Black, Duncan, The Theory of Committees and Elections (Cambridge: Cambridge University Press, 1958), pp. 1–11Google Scholar.

6 This discussion summarizes the results presented in Weisberg, , “Dimensional Analysis of Legislative Roll Calls,” pp. 69–94Google Scholar.

7 See, for example, MacRae, Duncan Jr., “A Method for Identifying Issues and Factions from Legislative Votes,” American Political Science Review, 59 (1965), pp. 909–926CrossRefGoogle Scholar.

8 The main source of the data is Hamilton, Holman, Prologue to Conflict (Lexington, Kentucky: Kentucky University Press, 1964), pp. 195–200Google Scholar. Supplementary data have been made available to the author by the Inter-university Consortium for Political Research.

9 A large number of Representatives did not participate in all five votes. There is reason to believe that many of the absences were intentional maneuvers, particularly the absences of several northerners on the fugitive slave bill. See Hamilton, p. 162. The winning margins on some of the bills were quite narrow, but they could have been defeated only if the absent Representatives from the North and South had united to vote against them. The closest votes were on the Utah territory bill and the Texas-New Mexico bill, the two bills which were opposed by some Representatives from both sections. Only two of the Representatives not voting on all bills had voting patterns inconsistent with the proximity scale. Treating all the absences as covert yes votes, as covert no votes, or as an intermediate category would detract from the fit of the proximity model, though not so much as to upset the basic interpretation of this model.

10 Alexander, Thomas B., Sectional Stress and Party Strength (Nashville: Vanderbilt University Press, 1967), pp. 70–73Google Scholar; Silbey, Joel H.Google Scholar, “Congressional Voting Behavior and the Southern-Western Alliance, 1841–1852” (Unpublished Ph.D. dissertation, State University of Iowa, 1963)Google Scholar; Silbey, Joel H., The Shrine of Party (Pittsburgh: University of Pittsburgh Press, 1967), pp. 108–114Google Scholar.

11 The Alexander and Silbey scales also contain votes prior to final passage and votes taken on other bills during the 31st Congress.

12 Legislators not voting on a bill opposed by the extremists of their sections have been treated as if they voted against it, the legislators not having supported the bill. This procedure for dealing with absences permits the categorization of nearly all the legislators, so Table 5 includes 89 more legislators than did Table 3.

13 Southerners were classified as pro-Compromise if they voted for either of two pro-North bills while northerners were classified as pro-Compromise only if they voted for the one pro-South bill. The southerners therefore had a greater chance to be categorized as pro-Compromise than did the northerners. However the southerners would still be found to be more pro-Compromise than the northerners if the District of Columbia bill were omitted so as to equalize the chances for each section.

14 Hamilton, , Prologue to Conflict, pp. 164–165Google Scholar.

15 The details of the scale are given in the Appendix.

16 Silbey, , “Congressional Voting Behavior …,” p. 291Google Scholar.

17 See Weisberg, , “Dimensional Analysis of Legislative Roll Calls,” pp. 94–103Google Scholar, and Leik, Robert K. and Matthews, Merlyn, “A Scale for Developmental Processes,” American Sociological Review, 33 (1968), pp. 62–75CrossRefGoogle ScholarPubMed.

18 See Weisberg, Herbert F., “L'étude comparative des scrutins législatifs,” Revue Française de Sociologie, 12 (1971), pp. 151–176CrossRefGoogle Scholar.

19 Leik and Matthews.

20 Weisberg, , “L'étude comparative …,” pp. 158–159Google Scholar.

Submit a response

Comments

No Comments have been published for this article.