Palmetto Postmortem: Examining the Effects of the South Carolina Voter Identification Statute
| DOI | 10.1177/1065912919837663 |
| Author | M. V. Hood,Scott E. Buchanan |
| Date | 01 June 2020 |
| Published date | 01 June 2020 |
| Subject Matter | Articles |
research-article2019
Article
Political Research Quarterly
2020, Vol. 73(2) 492 –505
Palmetto Postmortem: Examining
© 2019 University of Utah
Article reuse guidelines:
sagepub.com/journals-permissions
the Effects of the South Carolina Voter
https://doi.org/10.1177/1065912919837663
DOI: 10.1177/1065912919837663
journals.sagepub.com/home/prq
Identification Statute
M. V. Hood III1 and Scott E. Buchanan2
Abstract
In 2011, South Carolina passed a government-issued photo identification (ID) statute. We examine the effects of
this law on overall turnout, as well as for minority turnout in particular. A series of difference-in-difference tests
are specified using individual-level population data on registrants with and without ID, comparing the 2010 (pre-
implementation) and 2014 (post-implementation) election cycles. The results of our analysis indicate that the voter
ID statute did dampen overall turnout. These findings comport with recent scholarship which has found evidence
that voter ID laws can lower turnout rates. The size of the effect, however, is quite diminutive. We estimate that
initial implementation of the South Carolina statute decreased turnout in the 2014 general election by 0.19 percent.
In addition, the evidence gathered on the question of racial effects indicates that there is no discernible racial impact
from the state’s voter ID law.
Keywords
voter ID, political participation, public policy, election sciences
Introduction
Our study involves the analysis of a voter ID statute
from a single state. The fact is, however, that once one
Since the 2000 presidential election debacle in Florida,
delves into the details of these statutes none of them are
much attention has been paid to the mechanics underly-
exactly the same. To that point, isolating our study to a
ing the electoral process. Over the last decade the most
single state holds any nuances between these laws con-
contentious issue has involved voter identification (ID)
stant. In addition, it is quite possible that the effects of
statutes, especially those which require the presentation
these statutes related to voter turnout may be quite minor.
of government-issued photo ID (Hicks et al. 2015). In the
Indeed, this is what we find. As such, studies that rely on
interim, much ink has been spilled opining on the relative
different methodologies such as cross-state comparisons,
merits or the potential harms related to such require-
but lack the types of data we are able to bring to bear on
ments. Yet we still know relatively little about the actual
this question, may be insufficient for the present task. To
effects of such laws once they are implemented, espe-
the extent that there is a tradeoff between the ability to
cially in regard to how they may impact overall turnout,
make causal inferences and generalizability to other state
or turnout among specific groups.
settings, we choose the former.1
In this study, we examine the effects of a voter ID
law (Act R54) implemented in South Carolina in 2013.
More specifically, we are interested in determining
Theoretical Underpinnings
whether Act R54 negatively affected turnout in the 2014
Research that focuses on voter ID laws, whether a state
general election. We do so using individual-level data
requires a photo ID or not, is relatively sparse. While
on the population of registrants in South Carolina dur-
other aspects of electoral reform have been extensively
ing the midterm before, and following implementation
of the voter ID statute. We also have an identified set of
1University of Georgia, Athens, USA
registrants thought to lack Act R54 ID. These registrants
2The Citadel, Charleston, SC, USA
will serve as a comparison group whose behavior we
can contrast to those registrants who possess ID. Finally,
Corresponding Author:
M. V. Hood III, Department of Political Science, University of Georgia,
we also collect and analyze several ancillary statistics
180 Baldwin Hall, Athens, GA 30602, USA.
related to Act R54.
Email: th@uga.edu
Hood and Buchanan
493
studied, it is only in the second decade of the twenty-first
and Nielson (2017) and Barreto, Nuno, and Sanchez
century that more scholarly analysis has been brought to
(2009) do find effects for one or more racial/ethnic
bear on this issue. In part, this is because most modern
group.2
voter ID laws have only become commonplace in the last
While the bulk of scholarship relating to voter ID has
fifteen to twenty years (Gronke et al. 2019; Hicks,
relied on the Downsian approach associated with negative
McKee, and Smith 2016; McKee 2015). From 1950 to
effects, a number of recent studies have advanced the
1970, South Carolina was the only state to require voter
counterintuitive possibility that such laws may actually
ID until Hawaii did so. By 2000, 14 states had some type
have a positive relationship to turnout. On this point three
of ID requirement, though none required a form of photo-
hypotheses have developed (Highton 2017). The first
graphic ID (National Conference of State Legislatures
relates to educational campaigns launched by states to
[NCSL] 2016).
inform registrants of the requirements of a new voter ID
Following the 2000 presidential election, Congress
statute. Such efforts may be general or targeted to specific
passed election reform legislation termed HAVA (Help
groups (e.g., those who are thought to lack requisite photo
America Vote Act). Stewart (2014) indicates that HAVA
ID). A second mechanism whereby turnout may be
provisions designed to combat voter fraud provided a
increased occurs through a countermobilization effect in
platform for state action related to voter ID legislation. In
which groups thought to be more affected (e.g., minori-
2005, the Commission on Federal Election Reform, a pri-
ties, Democrats) may actually be galvanized to vote by the
vate blue-ribbon panel chaired by former President
presence of a voter ID law. The third view recognizes that
Jimmy Carter and former Secretary of State James Baker,
effects of such laws may be diminished over time as vot-
also recommended photographic voter ID requirements
ers learn of the new requirements and/or make any neces-
to protect confidence in the voting system. By 2016,
sary adjustments (see Vercellotti and Anderson 2009).
thirty-four states had voter ID requirements on the books,
On the second hypothesis, Hopkins et al. (2017) and
though some were facing court challenges (NCSL 2016).
Citrin, Green, and Levy (2014) find positive effects related
Highton (2017) has recently produced a comprehen-
to education efforts. In Virginia, the Department of
sive overview of scholarly work as it relates to voter ID.
Elections undertook a voter education campaign in which
The primary theory underlying the study of voter ID laws
registrants thought to lack ID were sent a mailer concern-
and turnout relates to the calculus of voting as expounded
ing the requirements of the statute and how free ID might
by Downs (1957). More specifically, these statutes are
be obtained. Following implementation, turnout was
viewed as a cost to a subset of voters who lack ID. A
found to have increased in precincts with higher rates of
member of such a group then would need to expend some
nonpossession (Hopkins et al. 2017). Citrin, Green, and
additional effort to obtain a valid form of ID to cast a bal-
Levy (2014) sent direct mailings to selected voting pre-
lot. Given this, one would hypothesize that the turnout
cincts in Tennessee and Virginia, including predominantly
rate of this group should be diminished (on this point of
black precincts in Knoxville and Roanoke. Registrants
view see also Rosenstone and Hansen 1993; Wolfinger
were sent targeted mailings informing them of voter ID
and Rosenstone 1980). Likewise, one can create a series
requirements in their respective state prior to the 2012
of sub-hypotheses by simply examining the characteris-
general election. These mailings were found to be associ-
tics of those registrants lacking ID. Racial gaps in ID pos-
ated with a slight increase in the probability of voting.3
session are one such example. If black registrants are
From a study of the Kansas voter ID statute Bright and
shown to have a higher ID nonpossession rate than white
Lynch (2017) also find a positive effect on turnout for
registrants, then one could logically hypothesize that
counties that engaged in specific types of advertising
turnout for these black registrants would more likely be
related to the requirements of the law.
adversely affected.
On the countermobilization hypothesis, Valentino and
Empirical research specifically analyzing the effect of
Neuner (2016) find confirmatory evidence of such an
voter ID laws on turnout can currently be characterized as
effect linked to voter ID laws. The intervening mechanism
having mixed results, although the bulk of findings in this
appears to be anger, which, in turn, is positively related to
area have showed little or no effect on turnout (e.g., see
voter turnout. This effect is not symmetrical, however,
Alvarez, Bailey, and Katz 2010; Ansolabehere 2009;
from a partisan standpoint. As the authors state, the “anger
Burden et al. 2014; Erikson and Minnite 2009; Gillespie
that Democrats experienced about voter ID laws was also
2015; Hood and Bullock 2012; Mycoff...
To continue reading
Request your trial