Data innovations on protests in the United States
| Author | Cassy Dorff,Grace Adcox,Amanda Konet |
| DOI | http://doi.org/10.1177/00223433221143808 |
| Published date | 01 January 2023 |
| Date | 01 January 2023 |
| Subject Matter | Regular Articles |
Data innovations on protests
in the United States
Cassy Dorff
Department of Political Science, Vanderbilt University
Grace Adcox
Department of Political Science, Vanderbilt University
Amanda Konet
Science Institute, Vanderbilt University
Abstract
For decades, the United States has been generally excluded from cross-national quantitative datasets on the study of
collective action and political resistance. More recently, however, new data collection efforts are on the rise. These
projects specifically focus on gathering granular level information about street protests and mobilizations in the
United States. In this article, we conduct a rigorous exploratory data analysis of three contemporary protest datasets.
These data collect information about the when, where and how of contentious politics in the United States. Our
thorough data review first summarizes the key similarities and differences across the datasets. Next, we review the
regional, temporal and methodological strengths and weaknesses of each dataset both individually and in contrast to
one another. Last, we examine potential research applications of these data by demonstrating what these data reveal
about the risks of protesting in these types of events. We conclude by offering recommendations for data use and
future data collection strategies for the study of collective action.
Keywords
protest, data, collective action, United States
Introduction
In the fall of 1959, Reverend James M Lawson stood in
front of a group of students leading nonviolence work-
shops in Nashville, Tennessee. Students such as Diane
Nash and John Lewis listened to these lessons, learning
how to withstand assault and harassment from counter-
protesters, business owners and police. Within a few
years, nonviolent sit-ins, freedom rides and public
marches would come to characterize many of the
well-known events of the Civil Rights Movement in
the United States. The movement would also come
to define decades of scholarship on social movements
in the United States (Morris, 2021). Despite the atten-
tion and progress brought by the Civil Rights Move-
ment, killings of African Americans – particularly
responses to the 2013 acquittal of George Zimmer-
man, who fatally shot 17-year-old Trayvon Martin,
and more recently the 2020 police murder of George
Floyd – continue to galvanize organizers into the streets
more than 50 years after Reverend James M Lawson’s
lecture. Street demonstrations related to a broad set of
equality issues, such as access to clean water and
women’s rights, also characterize today’s modern pro-
test movements. With such a rich history and contem-
porary presence of protests in the United States, what is
the state of social science data collection efforts on
these events?
Corresponding author:
cassy.dorff@vanderbilt.edu
Journal of Peace Research
2023, Vol. 60(1) 172–189
ªThe Author(s) 2023
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/00223433221143808
journals.sagepub.com/home/jpr
Protest, as a political event, takes on many different
meanings. Protests are often thought of as ‘disruptive
collective action’ events aimed at undermining elites and
institutions (Tarrow, 1991) and as ‘confrontational
activity’ by nongovernmental actors that directly chal-
lenge government policies (Carey, 2006). For our pur-
poses, we consider protests as ‘contentious and
potentially subversive [collective] acts that challenge nor-
malized practices, modes of causation, or systems of
authority’ (Beissinger, 2002: 14). This definition
encompasses a range of protest types that we observe
in the United States. In addition, it has been broadly
applied in social science research on collective action.
1
To understand the state of academic research on pro-
test events, we present a rigorous and historical review of
protest data in the United States. We next turn to an
original analysis of three key recently developed United
States protest datasets – the Armed Conflict Location
and Event Database (ACLED), the Crowd Counting
Consortium (CCC), and the National Study of Protest
Events (NSPE) – and compare the primary features and
coverage of these data. We then demonstrate the research
uses of these data by focusing on one applied research
question: what can researchers learn from these data
about the risks of protesting in these types of events?
This is an important question for social scientists both
to better understand repression–activism dynamics but
also for studying barriers to participation more broadly.
In our analysis, we pay careful attention to the role of
different actors and reveal insights about what these data
can and cannot tell us about the inte raction between
nonviolent action, counter-protesters and state violence
in the United States context. We conclude by discussing
future research and how to triangulate these data with
other projects on related subjects – such as police brutal-
ity and legislation – in order to advance research on
collective action.
Data on protests in the United States
The study of nonviolent protest in the United States is,
in many ways, undoubtedly rich. While early scholarship
on collective action viewed social movements as non-
rational, spontaneous and largely unstructured phenom-
ena, Civil Rights Movement scholars directly challenged
these ideas to argue that movements were organized,
rational, strategic and directly shaped by both
institutional and political pressures (Lipsky, 1968;
Smith, 1968; Cantor, 1969; Ackerman & Kruegler,
1994; Morris, 1999).
2
This research set the foundation
for scholars to systematically examine protest and collec-
tive action events as a specific class of data. Event data,
typically understood as a ‘qualitative change that hap-
pens in a specific point in time’ such as an arrest, birth,
crime, or battle (Allison, 2014: 1), are very useful for
near real-time analysis and comparative analysis given the
relative ease in generalizing event definitions across cases.
Yet, as we show below, the United States has been
largely left out of cross-national event datasets on pro-
tests and nonviolent action events. Data that does exist
on similar topics in the United States is often tailored
directly to individual campaigns or movements. Impor-
tantly, campaign data are especially valuable for aggre-
gate studies on the efficacy of specific movements such as
the Black Lives Matter movement (Williamson, Trump
& Einstein, 2018; Oliver, Hanna & Lim, 2019) and the
Women’s March movement (Fisher, Dow & Ray,
2017), as well as historic campaigns including the Civil
Rights Movement (Davenport, Soule & Armstrong II,
2011; Wasow, 2020) and anti-war protests of the twen-
tieth century (Swank, 1997; McAdam & Su, 2002).
Furthermore, these data have been essential in describing
the origins (McAdam, 1996), activities (Wang & Piazza,
2016; Wang & Soule, 2016) and consequences of Amer-
ican social movements (Carey, 2006; Fassiotto & Soule,
2017). We recognize that these data are informative for
questions that look at the broader characteristics and
strategies of social movements as a whole; in this paper,
however, we are interested in examining interactions and
tactics that occur at the granular level of events. Further-
more, our focus is timely: efforts to develop a richer
understanding of protest in state-specific contexts have
proliferated in recent decades, though the United States
has been largely excluded from such efforts in recent
years (Salehyan et al., 2012; Clark & Regan, 2016;
Francisco, 2019).
Figure 1 demonstrates the relative underrepresenta-
tion of the United States in the scholarly landscape of
protest campaign and event datasets.
3
Across all high-
1
See, for examples Ulfelde r (2005) and Berg-Schloss er, Badie &
Morlino (2020). This definition is broad and suggests protest
events can be nonviolent, violent, or a mix of both.
2
Though our article is written for political scientists and primarily
focuses on political science research, we necessarily includ e select,
overlapping literature and data from related subfields in sociology.
3
Data methodology in Online appendix. Data shown in Figure 1:
SCAD ¼Social Conflict Analysis Database; MM ¼Mass
Mobilization Data Project; MMAD ¼Mass Mobilization in
Autocracies Database; NAVCO ¼Nonviolent and Violent
Campaigns and Outcomes; WiRe ¼Women in Resistance
Dorff et al. 173
To continue reading
Request your trial