Evaluating the Quality of Changes in Voter Registration Databases

• Author: Spencer Schneider,Seo-young Silvia Kim,R. Michael Alvarez
• Published date: 01 November 2020
• Date: 01 November 2020
• DOI: http://doi.org/10.1177/1532673X19870512
• Subject Matter: Articles

https://doi.org/10.1177/1532673X19870512

American Politics Research

2020, Vol. 48(6) 670 –676

© The Author(s) 2019

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/1532673X19870512

journals.sagepub.com/home/apr

Article

Voter files are an important resource for political research and

also crucial for the integrity of the administration of elections,

as they dictate who votes. The files constantly change—but

not all changes are intentional or welcome. External intru-

sions into voter files became a salient issue in the 2016 presi-

dential election (Sanger, 2018). Media and federal officials

reported that foreign actors attempted to access voter data in

various states and that they may have tried to alter registration

data to manipulate election outcomes or undermine public

trust (Fandos & Wines, 2018; Perlroth, Wines, & Rosenberg,

2017). However, internal quality deterioration can sometimes

occur because voter files are large, dynamic, and complex.

For example, in the 2018 June primary election in California,

records for as many as 77,000 voters in the state’s system

were duplicated inadvertently by the Department of Motor

Vehicles; in the 2018 primary election in Los Angeles County,

118,000 voters were left off precinct rosters due to a merge

error (Myers, 2018; Reyes & Smith, 2018).

While election officials work tirelessly to guard against

cyberattack and human error, there are calls for independent

auditing of voter files and generally to improve their quality

(Alvarez, Ansolabehere, & Stewart, 2005; Alvarez, Jonas,

Winkler, and Wright, 2009; Ansolabehere & Hersh, 2010).

Past studies of voter data quality have been static in their

focus, as in Ansolabehere and Hersh (2014).1 However, as we

have seen, data quality can sharply change over time, present-

ing problems both to election administrators and to scholars

using the data. In this article, we present two methods that

evaluate the internal validity of voter registration data as it

changes over time, which increases assurance of voter file

quality and provides interesting data for election scholars, as a

novel source of data on election administration practices.

Collaborating with the Orange County Registrar of Voters

(OCROV), our first approach matches voter snapshots at dif-

ferent points in time and quantifies the changes to the file. As

voter files are dynamic, some rate of change is expected, due

to new registrations, residential mobility, deceased voters,

and changes in personal information. The resulting time-

series of changes can undergo statistical anomaly detection

to find anomalous changes, which represent those that depart

from the expected rate of change. While this approach can

provide notification of a sudden deterioration of database

quality, we argue that the generated audit data can also be of

scholarly interest, serving as an important source of informa-

tion on election administration, a rare window into this

important—and often overlooked—component of the demo-

cratic process. Our second approach is a duplication detec-

tion scheme which provides a list of potential duplicates with

a principled, automated approach, while minimizing cases

where the election official might accidentally delete a valid,

nonduplicate voter. Combined with voter file information on

how the registration data were generated, we show that we

870512APRXXX10.1177/1532673X19870512American Politics ResearchKim et al.

research-article2019

1California Institute of Technology, Pasadena, USA

Corresponding Author:

Seo-young Silvia Kim, California Institute of Technology, HSS, Caltech,

1200 East California Blvd., Pasadena, CA 91125-0002, USA.

Email: sskim@caltech.edu

Evaluating the Quality of Changes in

Voter Registration Databases

Part of Special Symposium on Election Sciences

Seo-young Silvia Kim1, Spencer Schneider1,

and R. Michael Alvarez1

Abstract

The administration of elections depends crucially upon the quality and integrity of voter registration databases. In

addition, political scientists are increasingly using these databases in their research. However, these databases are

dynamic and may be subject to external manipulation and unintentional errors. In this article, using data from Orange

County, California, we develop two methods for evaluating the quality of voter registration data as it changes over

time: (a) generating audit data by repeated record linkage across periodic snapshots of a given database and monitoring

it for sudden anomalous changes and (b) identifying duplicates via an efficient, automated duplicate detection, and

tracking new duplicates and deduplication efforts over time. We show that the generated data can serve not only to

evaluate voter file quality and election integrity but also as a novel source of data on election administration practices.

Keywords

voter registration data, record linkage, election integrity, election administration, data quality