Friday on My Mind: Re-Assessing the Impact of Protest Size on Government Concessions

• Author: Charles Butcher,Jonathan Pinckney
• DOI: http://doi.org/10.1177/00220027221099887
• Published date: 01 August 2022
• Date: 01 August 2022
• Subject Matter: Articles

Article

Journal of Conflict Resolution

2022, Vol. 66(7-8) 1320–1355

© The Author(s) 2022

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/00220027221099887

journals.sagepub.com/home/jcr

Friday on My Mind:

Re-Assessing the Impact of

Protest Size on Government

Concessions

Charles Butcher

1

and Jonathan Pinckney

2



Abstract

Do more protesters on the streets make governments likely to grant their dem ands?

Several studies link protest size and government concessions. Yet existing research has

limitations: many studies suffer from potential endogeneity due to potential protesters

joining protests when they anticipate that concessions are likely, causal mechanisms are

often unclear, and many of the most rigorous event-level studies are limited to

Western democracies. We reexamine this relationship in a non-Western sample using

a novel instrumental variable approach, using Fridays as an instrument for exogenous

variation in protest size in predominately Muslim countries. We perform two analyses:

one using the NAVCO 3.0 dataset, and the second using the Mass Mobilization in

Autocracies Dataset (MMAD). In both analyses exogenous variation in protest size

negatively affects the likelihood of concessions. Larger protests are less likely to receive

government concessions. We suggest these surprising results point to the importance

of unanticipated protests that produce new information about regime stability to

motivate government concessions.

1

Department of Sociology and Political Science, Norwegian University of Science and Technology,

Trondheim, Norway

2

Program on Nonviolent Action, United States Institute of Peace, Washington, DC, USA

Corresponding Author:

Jonathan Pinckney, c/o United States Institute of Peace, 2301 Constitution Ave, Washington, DC, 20036

USA.

Email: jpinckney@usip.org

Keywords

conflict, domestic politics, civil resistance, protest, social movements, repression,

instrumental variables

Introduction

Does the number of participants in a protest affect the likelihood that the government

will give in to its demands? The intuitive answer, frequently invoked in popular

discussions, is yes. For example, media descriptions of the 2018-2019 revolution in

Sudan emphasized how the presence of “millions”on the streets was key in forcing

President Omar al-Bashir to step down from power.

1

Extensive scholarly work supports

the intuition that larger protests more effectively force government concessions relative

to small protests (Amenta et al. 2010;Chenoweth and Stephan 2011;DeNardo 1985;

Lohmann 1994;Madestam et al. 2013;Tilly 1995). Large protests may be more

materially costly for the regime as more people engage in noncompliance, reducing the

resources at its disposal. Large protests are also hard to repress. While dispersing a

crowd of a thousand people may be manageable, dispersing a crowd of a million is

nearly impossible. As East German security chief Erich Mielke said when ordered to

repress the peaceful anti-Communist movement in East Germany: “We can’t just beat

up hundreds of thousands of people”(Whitney et al. 1989, 27). Indiscriminate re-

pression, which may be the only repressive option available to a regime facing mass

protests (Klein and Regan 2018), risks “backfire”in which violence against unarmed

civilians causes moral outrage and leads to even more people protesting (Hess and

Martin 2006;Sutton et al. 2014).

Large protests are also thought to signal broad dissatisfaction and increase the

perceived costs of governance for the incumbent in the future (Levento˘

glu and

Metternich, 2018;Kuran 1991;Tilly 1978). Faced with these challenges, govern-

ments may seek to stem large protests by either immediately giving in to some protester

demands, or at least promising to give in to such demands in the future. While other

factors, such as the identity of protesters or their specific goals, will certainly affect the

likelihood of government concessions, it seems clear that ceteris paribus, more

protesters in the streets will lead to more concessions from opponents.

Yet there are at least three key reasons why this link may not be so straightforward.

First, correlations between protest size and regime concessions, such as those that can

be found in Chenoweth and Stephan (2011),Chenoweth and Belgioioso (2019),Klein

and Regan (2018) and Butcher et al. (2018), may be complicated by endogeneity.

Signals of regime weakness, indicating that concessions are forthcoming, have a well-

attested positive effect on protest participation, as weakly-committed individuals

bandwagon onto movements on the verge of success (Lohmann 1993;Kuran 1991).

Thus, large protest sizes are almost certainly indicative of the widely-shared expec-

tation of government concessions, making the observed relationship between protest

Butcher and Pinckney 1321

size and concessions endogenous and the true relationship much more difficult to

determine.

Second, existing theories bundle “disruption”and “signalling”mechanisms that

have different implications for the link between protest size and government con-

cessions (McAdam and Su 2002). If protests create change through disruption and

short-term economic costs (Klein and Regan 2018) then social movements should work

to create as much (nonviolent) disruption as possible. But if large protests generate

concessions by signalling new information about the relative strength of opposition to

the government, then activists should focus on building broad coalitions that signal

widespread dissent.

Third, while there is an important literature on the impact of social movements and

nonviolent action campaigns across a wide range of countries (see e.g. Chenoweth and

Stephan 2011;Kadivar and Caren 2015;Kadivar 2018;Schock 2005), most event-level

studies of the effectiveness of protest that enable us to link individual protest char-

acteristics to outcomes have been limited to Western democracies. In particular, this

literature is dominated by studies of the United States (Walgrave and Vliegenthart

2012). Yet we have little reason to believe that the distinct protest field in the United

States will translate even to other developed democracies, much less to protest

movements in developing countries or repressive authoritarian regimes.

In this paper we examine these issues through a novel study of the relationship

between protest size and government concessions. Our empirical strategy rests on an

instrumental variable never previously used (to our knowledge) to measure exogenous

variation in protest participation: whether a protest took place on a Friday in a pre-

dominately Muslim country. We argue this instrument predicts larger protests well and

isolates the effects of protest size from the anticipation of government concessions.

Fridays have a special role in the social and religious lives of many practicing Muslims

as the day mandated in the Qur’an for attending public prayers at mosques. This

provides a “focal point”for potential dissidents to overcome collective action problems

by drawing on the close-proximity social networks provided by the mosque envi-

ronment (Ketchley and Barrie 2020). Overcoming these collective action problems

enables dissidents to generate higher levels of protest turnout. Fridays are also unlikely

to be directly correlated with government concessions for unobservable reasons. In so

far as observable protest characteristics might correlate with Friday protest, such as

domination by Islamist actors or centralization in major cities, these alternate avenues

can easily be controlled for. Any remaining effect of Fridays on government con-

cessions is therefore highly likely to run through the increased mobilization facilitated

by Friday prayer attendance.

We perform two separate analyses using this instrumental variable. The first uses

data on protest and concessions from 14 predominately Muslim countries in the

Nonviolent and Violent Campaigns and Outcomes (NAVCO), version 3.0 data

(Chenoweth et al. 2018). The second combines data on protest in 34 predominately

Muslim countries from the Mass Mobilization in Autocracies Dataset (MMAD)

1322 Journal of Conflict Resolution 66(7-8)