Loosened Partisan Attachments and Receptivity to Incumbent Behaviors: A Panel Analysis, 1972-1976

Abstract

An established hypothesis explaining the increased incumbency advantage is that incumbents reaped the benefit of the reduction in partisan identification's influence on the congressional vote. Briefly, the hypothesis maintains that as partisan identification's influence waned incumbent behavior's influence increased. The increased incumbency advantage is then the result of incumbents actively attracting voters whose decision is based less on partisan considerations and more on what incumbents offer. Using the 1972-76 NES panel study, we test this hypothesis by examining incumbent behavior's influence on voters with loosened partisan identification. The results support this premise. The interaction between incumbent behaviors and loosened partisan identifi cation increases the probability of an incumbent vote, attracting loosened challenger party identifiers and retaining loosened incumbent party identifiers. Finally, the results suggest that higher levels of incumbent activity increase the probability of an incumbent vote.

The Incumbency Advantage in U.S. Elections: An Analysis of State and Federal Offices, 1942–2000

“…Kostroski (1973) observed a pattern similar to that described in Figure 3 below for the Senate and attributes it to party dealignment. Erikson (1972), Burnham (1974), Ferejohn (1977), Nelson (1978-79) and Romero and Sanders (1994) also argue that party dealignment caused higher incumbency advantages in House elections. Krehbiel and Wright (1983) argue that declining loyalty of voters explains the growth of the incumbency advantage in House elections.…”

What They Do Does Matter: Incumbent Resource Allocations and the Individual House Vote

“…1. Although slightly more supportive than other similar recursive analyses (Box-Steffensmeier et al, 2003;Cain et al, 1987;Romero & Sanders, 1994), Table 1 offers mixed evidence in support of the hypothesis that incumbent resource allocations influence the individual vote, ranging from support at traditional levels of statistical significance (staff, and the latent style score, both q = .01 two-tailed), to a narrow miss (bills sponsored and cosponsored, q = .06 two-tailed) to insignificant (number of district office, q = .11 two-tailed). Despite this spread in significance, as a whole, these results suggest that pooling elections data helps sharpen alternative hypothesis findings by ameliorating the restricted variance problem associated with individual-level tests of the resource allocation hypothesis.…”

“…Cain et al (1987) find that the 1978 vote was responsive to larger staffs, but not caseloads or trips home. On the other hand, in a panel analysis Romero and Sanders (1994) find that incumbent resource allocations (a latent variable with bills, staff, and district offices as indicants) did influence the 1976 vote, but mostly for voters with weakening partisan identification. Finally, Box-Steffensmeier et al (2003) find that the 1994 individual vote was influenced by floor speeches, bills sponsored, and the percent of bills with a local orientation, but also only at about a .05 one-tailed significance level.…”

“…Since then, Serra and Cover (1992), Serra (1994), and Serra and Moon (1994) have replicated this finding. Johannes, 1988;Parker, 1980;Ragsdale & Cook 1987;Romero & Sanders 1994). 3 Scholars suggest two methodological problems hinder establishing the incumbent resource allocation hypothesis.…”