The Effects of Election Proximity on Participatory Shirking: The Staggered‐Term Chamber as a Laboratory

• DOI: http://doi.org/10.1111/lsq.12090
• Author: Akitaka Matsuo,Kentaro Fukumoto
• Published date: 01 November 2015
• Date: 01 November 2015

KENTARO FUKUMOTO

Gakushuin University

AKITAKA MATSUO

Nuff‌ield College, University of Oxford

The Effects of Election Proximity

on Participatory Shirking:

The Staggered-Term Chamber

as a Laboratory

This study discusses a downside of electoral pressure. As elections approach,

legislators reduce their effort in legislative activities, albeit while increasing their effi-

ciency. To show this, we propose a new, natural experimental design exploiting

staggered legislative election calendars to identify the effect of approaching elections.

Two-way natural blocking improves the balance of pretreatments and an instrumental

variable approach addresses noncompliance by retirees. Our analysis of the Japanese

House of Councillors demonstrates that legislators up for election show up in the cham-

ber less often than those not facing election; however, when they do show up and speak,

they tend to speak longer.

Elections are the most important tool for disciplining incumbent

legislators’ activities, so long as these legislators intend to seek another

term. In order to determine whether this “electoral connection”

(Mayhew 1974) really works, various studies have analyzed whether

and how much, in the absence of electoral pressure, incumbents engage

in shirking or “behavior that differs from what would be observed

given perfect monitoring and effective punishment by constituents”

(Rothenberg and Sanders 2000, 316). Although scholars have not found

conclusive evidence of retiring legislators’ “ideological shirking”

through voting against their constituents’ preferences, scholars are in

agreement that retiring legislators engage in “participatory shirking”

through “reduc[ing] their effort level by voting less” (Rothenberg and

Sanders 2000; see also Bender and Lott 1996). These f‌indings imply

that electoral pressure discourages shirking and is benef‌icial to voters’

interests. By theoretically and methodologically reassessing the impact

of electoral pressure on legislators’ efforts, we claim otherwise: electoral

LEGISLATIVE STUDIES QUARTERLY, 40, 4, November 2015 599

DOI: 10.1111 /lsq.12090

V

C2015 The Comparative Legislative Research Center of The University of Iowa

pressure can also encourage participatory shirking, thereby harming

voters’ interests by obstructing their representation in the chamber.

Presence and proximity of the next election are two sources of the

electoral connection in the literature. Studying either one of these sources

leads researchers to compare different groups of legislators and different

kinds of shirking. On the one hand, the focus on the presence of the next

election (or the “last period problem”) leads to a comparison between

retirees and reelection seekers and establishes that election presence

decreases participatory shirking (Bender and Lott 1996; Herrick,

Moore, and Hibbing 1994; Lott 1990; Rothenberg and Sanders 2000;

Zupan 1990). On the other hand, the focus on the effect of the proximity

of the next election leads to the comparison between legislators up for

election and those not up for election which has led to debate as to

whether election proximity decreases ideological shirking (Ahuja 1994;

Albouy 2011; Elling 1982; Hill and Hurley 2002; Kalt and Zupan 1990;

Lindst€

adt and Vander Wielen 2011; Poole 1981; Shepsle et al. 2009;

Stratmann 2000; Thomas 1985; Wright and Berkman 1986). In contrast

to these studies discussing the positive effects of electoral pressure, we

are the f‌irst to explore a negative effect of proximity on participation:

election proximity actually increases participatory shirking because

election proximity necessitates campaign activity that uses up time

which would otherwise be spent on legislative activity (Fenno 1978).

To make a sound causal inference on the issue, we develop a gen-

erally applicable method that frames staggered-term chambers (e.g., US

Senate) as a natural experimental setup. In staggered-term chambers,

during any given pre-election period, only some members are up for

election (treatment group) while others are not (control group). At the

next pre-election period, the situation is reversed: members who were in

the control group in the previous period are now up for election and in

the treatment group, while those who were in the treatment group are, if

reelected, in the control group. Since the same legislators appear in both

the treatment and control groups in data pooled over several election

cycles, observations in both groups should share many characteristics

and differ only in their exposure to electoral pressures. The virtue of the

(natural) experimental approach is to allow scholars to worry little, if at

all, about omitted variable bias in regressing a dependent variable (effort)

on the treatment assignment (up for election or not). This is due to the

fact that in a natural experimental design, “control over confounders

comes primarily from appropriate research-design choices, rather than ex

post statistical adjustment” (Dunning 2012, 24).

Our method takes full advantage of unique features of the

staggered-term setting to implement such a design. First, unlike foregoing

600 Kentaro Fukumoto and Akitaka Matsuo