The Impact of District Magnitude on Voter Drop‐Off and Roll‐Off in American Elections

Published date01 November 2015
AuthorJames M. Curry,Paul S. Herrnson,Jeffrey A. Taylor
Date01 November 2015
University of Connecticut
ICF International
University of Utah
The Impact of District Magnitude
on Voter Drop-Off and Roll-Off in
American Elections
This study demonstrates that multimember districts (MMDs ) complicate bal-
lots, reduce voter information, and increas e incentives for strategic voting in ways
that reduce voter participation. Using data from three sta tes that elect members of at
least one legislative chamber from both single- member districts (SMDs) and MMDs,
we test hypotheses about the impact on MMDs on ba llot drop-off (selecting fewer
candidates for an office than permissible) and rol l-off (not voting in down-ballot
races). We find support for both sets of hypotheses, with the str ongest results related
to ballot drop-off. The results have broad i mplications for voter participation, re pre-
sentation, and election administration in t he many states and localities that use
MMDs to elect public officials.
Ballot designs, electoral institutions, and the interactions
between them can result in voter confusion and errors and affect rates
of participation in American elections. Although one need look no
further for evidence than the impact of the so-called “butterf‌ly ballot”
used in Palm Beach County, Florida and the Electoral College’s unit
rule on the outcome of the 2000 presidential election, other cases
abound. Many elections feature complicated ballots, and all political
arrangements inherently possess some biases. Elections held in mul-
timember districts (MMDs) are among them, especially elections
held in districts with large magnitudes (i.e., those that elect many
off‌ice holders). Used to select off‌iceholders in many modern indus-
trial democracies, MMDs deviate from the single-member districts
(SMDs) that distinguish the most visible elections in the United
States. Nevertheless, 10 state legislatures use some form of MMD to
elect at least some representatives—as do many counties, cities,
DOI: 10.1111 /lsq.12091
C2015 The Comparative Legislative Research Center of The University of Iowa
towns, school boards, and other so-called special governments.
Prior to the Apportionment Act of 1842, many states also used
MMDs to elect members of the US House of Representatives, and
MMDs were used periodically to elect members of Congress until
they were outlawed in 1967 (Martis 1982).
MMDs inf‌luence the conduct and f‌inancing of US campaigns
and elections (e.g., Cox and Morgenstern 1995; Curry, Herrnson,
and Taylor 2013; Niemi, Jackman, and Winsky 1991; Trounstine
and Valdini 2008). However, less is known about how they inf‌luence
voter behavior. While the large MMDs used in proportional repre-
sentation systems overseas have been shown to increase voter parti-
cipation, there are reasons to believe that bloc MMD elections, such
as those used in the United States, may have the opposite effects, in
part because these elections add complexity to already complicated
ballots (Niemi and Herrnson 2003). They can lead to confusion over
the number of votes an individual can actually cast and frustrate
voters who do not understand how MMD elections work. Further-
more, they raise information costs for voters by requiring them to
learn about more candidates and increasing the likelihood that they
fail to complete a ballot.
Our study addresses this important question: how do MMD elec-
tions affect rates of voter participation? Specif‌ically, we analyze the
effects of MMDs on rates of ballot drop-off and ballot roll-off among
voters. For the purposes of this study, drop-off is def‌ined as the selection
of fewer candidates for an off‌ice than is permissible, and roll-off is
def‌ined as not voting in down-ballot races.
Using precinct-level data,
we analyze the effects of district magnitude in three states (Maryland,
New Hampshire, and Vermont), each of which uses both MMDs and
SMDs to elect members of at least one legislative chamber. This research
design is advantageous as it allows inferences to be drawn about the
effects of district magnitude more easily than if we had to compare
across states and legislative chambers simultaneously. We show that vot-
ers are more likely to drop-off in MMD elections than in SMD elections
and are even more likely to drop-off in larger magnitude MMDs. Voters
who use a ballot that includes an MMD for a state legislative election
also are more likely to roll-off than voters who are given the option of
selecting only one state legislative candidate. These results have impor-
tant implications for elections and representation in American state legis-
latures as well as in municipal and county governments. States should
carefully consider the potential consequences of district structure and
ballot design in organizing elections and make an effort to educate voters
about the challenges posed by MMDs.
628 Paul S. Herrnson, Jeffrey A. Taylor and James M. Curry

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT