Deciding to support violence: An empirical examination of systematic decision-making,activism,and support for political violence

Abstract

This study examines how attitudes of activism and systematic decision-making are related to support for political violence. Using unique data from a randomly selected sample of undergraduate and graduate students (N = 503), this study explores how activism, systematic decision-making, and political affiliation coincides with existing support for political violence. Among respondents, stronger support for activism and less systematic decision-making behavior was associated with support for political violence on one’s behalf. These results hold across models and suggest that in the United States, cognitive psychology and decision-making perspectives inform the decision to support political violence and in turn, should be considered in efforts to curb support for organizations which use political violence as a tactic.

Keywords

Activism decision-making political violence TRDM impulsivity

Introduction

In testimony on 23 July 2019 before the US Senate Committee on the Judiciary, Federal Bureau of Investigation Director Christopher Wray (2019) stated that “The most persistent threats to the Nation and to U.S. interests abroad are homegrown violent extremists (‘HVEs’), domestic terrorists, and foreign terrorist organizations (‘FTOs’)” (p. 2). Citing a range of factors that include increased use of encrypted online platforms, perceptions of government overreach, socio-political concerns, and racism, Wray suggested that the issues of domestic terrorism and homegrown violent extremism extend beyond individual ideologies and amount to a serious threat to national security. Adding to this concern, recent public reporting has suggested that domestic political extremists have exhibited a shortened time span between the adoption of extremist views and the perpetration of violent acts (Barrett, 2019). As a result, law enforcement agencies in the United States are placed at a distinct disadvantage as the temporal window to detect and prevent violent acts narrows.

In response to the threat of extremist violence, since 2011 the Department of Homeland Security Countering Violent Extremism (CVE) Task Force has worked alongside non-governmental organizations and community stakeholders advancing a prevention-focused model called “Countering Violent Extremism” (CVE) (Department of Homeland Security, 2017). Generally, CVE initiatives aim to raise awareness of extremist messaging and prevent the adoption of views that engender political violence based on strategies that identify at-risk communities and individuals and develop resilience and counter-narratives (Harris-Hogan et al., 2016). This approach requires both an understanding of individuals who support or adopt extremist views and how these individuals may differ systematically from comparable politically and socially involved individuals. Though recent research has focused on the former issue using quantitative and qualitative methods (Jensen et al., 2018; LaFree et al., 2018), limited focus has been placed on examining the latter.

The present study suggests that theories of systematic decision-making are a promising framework that could aid in describing systematic differences between supporters of political violence and non-violent activists, thereby filling this gap and aiding future CVE efforts. To date, research on decision-making finds that individuals who are less likely to deeply consider the consequences of their actions tend to engage in a variety of antisocial behaviors and achieve negative outcomes (Louderback and Antonaccio, 2017; Paternoster and Pogarsky, 2009). Moreover, Paternoster and Pogarsky (2009) and Frederick (2005) find that systematic decision-making is strongly linked with short- and long-term positive outcomes. Consequently, examining decision-making among individuals who support activism and political violence could prove fruitful in identifying appropriate comparison groups for CVE interventions and inform efforts to degrade support for, and potentially reduce the use of political violence (Malthaner and Waldmann, 2014).

This study aims to contribute to the empirical literature by exploring how systematic decision-making can describe support for activism and political violence in a university sample. In order to do so, novel individual-level survey data were collected on a randomly selected sample of undergraduate and graduate students at a large public university. These individual-level data, collected in the fall and winter of 2018, include scales of activism and radicalism, Thoughtfully Reflective Decision-Making (TRDM), impulsivity, and cognitive reflection.

Activism and political violence

Across the political spectrum, individuals hoping to achieve change have taken to the streets, the papers, and the Internet to garner support for their cause (Bonilla and Rosa, 2015; Walsh, 2000). However, these movements are not always peaceful, and examples of political violence abound. From the Weather Underground’s bombing campaigns throughout the 1970s (Jacobs, 1997), to a recent mass shooting in El Paso, Texas, by a White nationalist (Eligon, 2019), when individuals choose to engage in violence in service to a broader social cause, it has serious repercussions for society. Research has shown that though activism and support for political violence co-occur (Decker and Pyrooz, 2019; Moskalenko and McCauley, 2009), understanding the conditions that may inhibit activists from choosing to support violence remains an important task.

A related body of work on social movements provides several structural and socio-demographic reasons for participation in protests. Following the terrorist attacks on September 11, 2001, Santoro and Azab (2015) found that among Arab Americans in the Detroit area whose Arab identity was not particularly salient, personally experiencing repression was key in mobilization to peaceful protest. More generally, research on high- and low-risk activism suggests that biographical availability (i.e. unmarried individuals without children and not employed full-time) may differentiate participation in high-risk activism from low-risk behaviors (Wiltfang and McAdam, 1991). Contrasting this finding, Nepstad and Smith (1999) instead highlight the role of agency and social network ties in movement participation. To date, research on supporters, but not participants, of political violence in the United States has been limited. Though studies have examined detailed data relating to incidents of terrorism (LaFree et al., 2014), extremism and hate crime (Chermak et al., 2012), networks of illicit and terrorist organizations (Asal et al., 2016), and individuals who engage in criminal acts of extremism (LaFree et al., 2018), few studies of supporters exist (Moskalenko and McCauley, 2009). This is not surprising however; supporters are a difficult population to identify. As supporters of political violence do not make headlines as a result of aspired or actualized criminal behavior, there is often little information available in the public domain about the sympathizers and constituencies of violent extremist organizations. Fortunately, work in the United States describing activism and radicalism, as well as studies looking internationally at support for political violence can inform this inquiry.

Moskalenko and McCauley (2009) is perhaps the first effort to quantitatively examine the relationship between activism and radicalism (or support for political violence) in the United States and cross-nationally. Across three samples, the authors describe important sources of variation in readiness to support and participate in violent and non-violent political dissent using student samples in the United States and Ukraine, as well as a nationally representative sample of US adults. Moskalenko and McCauley (2009) integrate measures of activism orientation and mobilization attitudes to introduce the Activism and Radicalism Intention Scales (ARIS). Validated across the three samples using principal component analysis and focusing attitudes to distinct identity groups, the authors suggest that while attitudes and behavioral intentions for activism and radicalism are closely related, they represent distinguishable constructs. Supporting this, Decker and Pyrooz (2019) describe radicalism in an incarcerated sample using this instrument and find a similar distinction between attitudes of activism and radicalism.

Further research on activism suggests that individual-level factors are predictive of support for, and participation in risky dissent behaviors. Applying a within-survey experimental design, Kearns et al. (2018) build on McAdam’s (1986) model of high and low risk activism. Manipulating conditions of minority status, grievance, and risk, Kearns et al. (2018) find that while perception of risk influences whether a supporter will engage in dissent behaviors, individual factors including social dominance orientation and right-wing authoritarianism explained why some chose to support a cause in general as well as opting to actively participate in dissent behaviors.

Outside of the US context, empirical work has more directly engaged with individual-level factors associated with support for political violence. Examining shifts from peaceful protest to political violence in Northern Ireland in the 1970s, White (1989) explains that “support for political violence results from a conscious decision that occurs when people come to see peaceful protest as futile.” (p. 1297). Contrasted with work that emphasized the importance of broader structural factors, White’s analysis highlights individual experiences with state repression as the driving factor in what is deemed to be instrumental political violence. Interestingly, without actively invoking decision-making research, White’s findings appear consistent with its underlying principles. When individuals did not believe that there is a tangible benefit of non-violent actions, they engaged in political violence.

Research on support for political violence also speaks to an ongoing debate among researchers of political violence regarding the relative importance of extremist attitudes and beliefs as contrasted with specific behavioral pathways to violence. To this end, some experts note that among extremists, deep understanding of ideological tenets and fervent belief in a cause is neither necessary nor sufficient to result in violence (see, for example, Borum, 2011; Horgan, 2011). Noting divergence in extremist attitudes and behavior, they suggest that research should focus predominantly on behavioral pathways that produce violence.

The opposing perspective asserts that though extremist attitudes and belief are poor predictors of violent behavior, they remain an important dimension of radicalization and facilitate an understanding of justifications for violence and the broader milieu of support that extremists draw and may emerge from (Bartlett and Miller, 2012; Malthaner and Waldmann, 2014; Neumann, 2013). In this study, I suggest that understanding the mutable individual level attributes associated with support for political violence is important and can help answer Neumann’s (2013) call to “understand why certain belief systems resonate with certain populations and—correspondingly—what combination of factors explains their lack of resonance and decline” (p. 881).

While research has found that the decision to support activist movements and political violence is subject to opportunity (McAdam, 1986; Wiltfang and McAdam, 1991), political repression (Santoro and Azab, 2015), and other forces (Nepstad and Smith, 1999; White, 1989), Sageman (2017) asserts that it fundamentally is a choice. That is to say, even among passionate advocates, few choose to engage in dissent behaviors, fewer actively support illegal or violent action (Moskalenko and McCauley, 2009), and even fewer go on to participate in illegal or violent acts (McCauley and Moskalenko, 2017). In order to better understand this decision within a theoretical framework, I turn next to research describing systematic decision-making.

Systematic decision-making

Rational choice theory states that when individuals make decisions, we try to maximize the benefits while minimizing the costs of our behaviors (Bentham, 1776). Simply put, when an action is perceived to be high-reward and low-risk, we are more likely to go through with it; contrastingly, we are less likely to act on behaviors we believe to be low-reward and high-risk. Additional research has refined this theory in a number of domains examining concepts such as bounded rationality (Simon, 2000), individual preferences (Caplin and Dean, 2015), and hyperbolic discounting (Nagin and Pogarsky, 2001). Notably, this perspective has been applied to understanding political violence, albeit most often examining the behavior of organizations rather than individuals. At this aggregate level, studies find that the behavior of violent political extremists and their supporters tends to be consistent with theories of rational choice (Crenshaw, 1987; Dugan and Chenoweth, 2012; LaFree and Ackerman, 2009).

Research has also highlighted how all decision-making is not created equal. Kahneman (2003) describes dual process theory as a framework for understanding two ways that individuals make decisions on a day-to-day basis. In system 1 processing, decisions are made intuitively (see also, Tversky and Kahneman, 1981). As a result, this process is efficient and by relying on cognitive shortcuts (heuristics), it requires little acquisition of new information, limited consideration of alternatives, and no deep reflection on how far the outcome attained was from an optimum solution. System 2 processing on the other hand requires individuals to effortfully engage with a problem and deliberately consider the potential risks, costs, and benefits of a given response in order to maximize their chances at a favorable outcome. This second system is far more cognitively demanding, and thus does not describe how most individuals make most of their decisions. In some circumstances, the system that is engaged is not a critical factor. When considering support of political violence, however, I contend that relying on biased cognitive strategies may leave individuals vulnerable to persuasive misinformation and emotional arguments, and ultimately facilitating negative and anti-social outcomes.

Notably, the dual process cognitive framework has already been applied to understanding anti-social outcomes such as crime and delinquency. Paternoster and Pogarsky (2009) describe these two patterns of cognition through the lens of an individual’s ability to engage in TRDM. The authors describe TRDM as follows:

The tendency of persons to collect information relevant to a problem or decision they must make, to think deliberately, carefully, and thoughtfully about possible solutions to the problem, apply reason to the examination of alternative solutions, and reflect back upon both the process and outcome of the choice in order to assess what went right and what went wrong. (Paternoster and Pogarsky, 2009: 104–105)

Thus, TRDM reflects an individual’s tendency to engage in system two processing. Paternoster and Pogarsky view this as an individual trait that may vary within individuals over time, across individuals, and across contexts. Importantly, the authors note that TRDM is distinct from impulsivity—a factor which may short-circuit the systematic and deliberative process (Hirschi and Gottfredson, 1983; Paternoster and Pogarsky, 2009). Consequently, impulsivity may serve as a proxy for an individual’s pattern of engaging in system one processing (Paternoster and Pogarsky, 2009).

Research examining both antisocial and prosocial outcomes supports this assertion. Using Add Health data, Paternoster and Pogarsky (2009) found that adolescents exhibiting low TRDM were more likely to engage in general delinquency, heavy drinking, and illegal drug use, whereas those demonstrating higher TRDM were more likely to graduate from college, participate in the community, and engage with civic groups. Likewise, Louderback and Antonaccio (2017) found that, within university student and faculty samples, low TRDM was associated with both participation in computer-focused cyber deviance as well as the probability of cyber-victimization. Thus, across contexts research has found that systematic decision-making informs individual propensity to experience negative and antisocial outcomes.

TRDM and impulsivity are not optimal indicators of dual process theory, however. Notably, both TRDM and impulsivity are typically measured using self-report survey items rather than applied measures. Frederick (2005) provides a more direct assessment of system two thinking in the Cognitive Reflection Test (CRT) (Kahneman, 2003; Tversky and Kahneman, 1981). High performance on the CRT is strongly predictive of prosocial outcomes and is conceptually more precise since high performance on the cognitive puzzles requires effortful and deliberative consideration (Frederick, 2005). As a result, this test perhaps best approximates if an individual engages in systematic decision-making. Taken together then, TRDM, impulsivity, and performance on the CRT should demonstrate general patterns in systematic decision-making as well as more directly measured aptitude.

Theoretical model

To date, research suggests that understanding support for legal and non-violent political dissent is critical to understanding support for political violence (Moskalenko and McCauley, 2009). Moskalenko and McCauley (2009) suggest that while activism and support for political violence are related, they represent two distinct constructs whose relationship will vary due to important social and structural factors. Under autocratic governance and acute positions of identity group repression, activism and radicalism may converge more closely due to limited opportunities for free expression. In more democratic societies however, the legal and non-violent expression of political dissent may be viewed more popularly as a mechanism of social change, prompting a divergence between the two. Initial evidence supports this theoretical structure (Decker and Pyrooz, 2019; Moskalenko and McCauley, 2009). Consequently, I anticipate that cognitive, social, and political dynamics could help explain the remaining variation between activism and support for political violence.

After a review of the literature, no studies were found to have examined the role of decision-making in supporting political violence. However, drawing upon the theoretical tenets which underlie dual process theory (Frederick, 2005; Kahneman, 2003; Paternoster and Pogarsky, 2009; Tversky and Kahneman, 1981), empirical predictions emerge. To this end, individuals who are impulsive in their decision-making processes would be most likely to support political violence performed on their behalf. Contrastingly, those who report higher levels of TRDM would likely eschew political violence in favor of non-violent means of social change. Moreover, while self-reported systematic cognition should be a valuable global indicator, individuals who demonstrate the ability to engage effortfully in cognitive reflection should also be less likely to support political violence.

Focusing on these theoretical explanations for attitudes supportive of political violence, the following hypotheses emerge:

H1: Attitudes in favor of activism will be positively related to support for political violence.

H2: Indicators of systematic decision-making processes will be negatively related to support for political violence.

Materials and methods

Data

To examine these questions, I collected original survey data on a sample of undergraduate and graduate students at a large public university in the United States. Employing the student directory as a sampling frame,¹ this questionnaire was sent out to a randomly selected sample of 6,095 undergraduate and graduate students in August 2018 using Qualtrics, an online survey platform.² The questionnaire was anonymous and the consent waivers on the first page of the survey instructed participants to omit any identifying information. Upon completing the questionnaire, participants were entered into a drawing for one of 25 $20 Amazon.com gift cards. Of the 6,095 individuals invited to participate, 10.1% began (618) and nearly 9% (529) completed the survey. Using listwise deletion to omit individuals with missing data, the final analytic sample was 503.

Questionnaire and procedure

The survey instrument developed for this study included questions and scales representing validated items drawn from prior literature on activism and support for political violence (Moskalenko and McCauley, 2009), on impulsivity and decision-making (Frederick, 2005; Grasmick et al., 1993; Paternoster and Pogarsky, 2009), political affiliation, and demographic characteristics. Questions were presented sequentially beginning with demographic information, moving to measures of systematic decision-making, and ending with measures of support for activism and political violence. Generally, respondents took 10–15 minutes to complete the study.

Outcome measure: support for political violence

The dependent variable in this study is an averaged index of the five items identified on Table 1 under the heading “Support for Political Violence.” The items comprising this score reflect the degree to which respondents agreed or disagreed with statements such as “I would continue to support an organization . . . even if the organization sometimes resorts to violence,” “I would participate in a public protest against oppression of my group even if I thought the protest might turn violent,” and “I would attack police or security forces if I saw them beating members of my group.” Common across these statements is explicit support for the use of violence either by, or on behalf of the nominated identity group. While the items contributing to the index are ordinal Likert-type scales from 1 to 7, the averaged index ranges from 1 to 7 in increments of 0.2 and thus approximates more closely general support for political violence (Cronbach’s α = 0.834, mean = 2.409, standard deviation = 1.305). Like the radicalism intentions scale (RIS) applied in Moskalenko and McCauley (2009), lower values signify more consistent, or stronger disagreement with statements of support for political violence, whereas higher values reflect more, or more consistent support for political violence. A value of 4 here indicates a neutral attitude in the aggregate.

Table 1.

Activism and radicalism intentions scales.

Survey item	M	SD	% agree
Non-violent activism
I would join/belong to an organization that fights for my group’s political and legal rights.	5.209	1.605	69.78
I would donate money to an organization that fights for my group’s political and legal rights.	4.811	1.700	59.64
I would travel for one hour to join in a public rally, protest, or demonstration in support of my group.	4.555	1.835	54.87
I would continue to support an organization that fights for my group’s political and legal rights even if the organization sometimes breaks the law.	3.398	1.770	27.83
Support for political violence
I would continue to support an organization that fights for my group’s political and legal rights even if the organization sometimes resorts to violence.	2.256	1.612	11.33
I would participate in a public protest against oppression of my group even if I thought the protest might turn violent.	2.775	1.850	20.87
I would attack police or security forces if I saw them beating members of my group.	2.247	1.627	9.54
I would go to war to protect the rights of my group.	2.738	1.821	15.71
I would retaliate against members of a group that had attacked my group, even if I couldn’t be sure I was retaliating against the guilty party.	2.030	1.467	7.75

N = 503. All items are scored using a Likert-type scale ranging from 1 (disagree completely) to 7 (agree completely). % Agree indicates the percent of respondents who selected a value higher than 4 (neutral).

Activism index

Like the measure indicating support for political violence, activism is approximated using an averaged index of four items identified on Table 1 as “Non-violent Activism” (Cronbach’s α = 0.798, mean = 4.493, standard deviation = 1.359). The statements, which comprise this summary measure reflect a variety of legal and non-violent forms of activism including organization membership, willingness to donate to organizations, willingness to travel for a public rally or demonstration, and continued support for organizations which may sometimes break the law. Higher values here represent stronger, or more consistent agreement with statements of support for legal and non-violent activism, whereas lower values reflect more consistent, or stronger disagreement with statements of support for activism.

Systematic decision-making

To measure cognitive processes and performance, summary scores of three scales were included. The first of these is a four item scale measuring TRDM taken from Paternoster and Pogarsky (2009). These 5-point Likert-type scale questions assess the degree to which the respondent agrees with statements regarding their cognitive processes when solving problems (strongly disagree, disagree, neutral, agree, and strongly agree). Like in Paternoster and Pogarsky (2009), for each statement a point value is assigned from 1 to 5 up to a total maximum of 20, with lower scores indicating a less TRDM process and higher scores indicating a superior self-assessment of TRDM.

The second scale is the CRT (Frederick, 2005). The three free-response cognitive puzzles which comprise this require a deliberate and systematic approach to solve successfully. The summary score for this scale reflects the number of correct answers out of three (0–3), with higher values evidencing more active engagement of deliberative cognition and lower scores signifying more less systematic decision-making.

The final measure of systematic decision-making is the summed score of the four-item impulsivity index from the Grasmick et al. (1993) scale of self-control. The items that comprise this scale assess reflective agreement with four statements including “I often act on the spur of the moment,” “I don’t devote much thought and effort to preparing for the future,” “I often do whatever brings me pleasure here and now, even at the cost of some distant goal,” and “I’m more concerned with what happens to me in the short run than in the long run” ranging from strongly disagree (1) to strongly agree (4). Summing the four items, this scale then reflects the time-stable component of self-control most conceptually linked with TRDM and systematic decision-making (Grasmick et al., 1993). These results in a summary score ranging from 4 to 016 with higher values reflect higher impulsivity and lower values represent less impulsivity.

Demographic control variables

Other explanatory variables here include a measure of respondent age, whether the respondent was male, respondent race and ethnicity,³ and respondent political affiliation.⁴ While not of primary concern here, research on extremism has found a robust relationship between age, gender, and involvement in political violence (LaFree et al., 2018), and criminological research has long since debated a relationship between racial/ethnic minority status and involvement in crime (Matsueda and Heimer, 1987).

Analyses

Analyses began with a descriptive examination of the theoretical and control variables, including bivariate correlations with the dependent variable. Next ordinary least squares (OLS) regression models were estimated to address each of the hypotheses.⁵ As both hypotheses 1 and 2 predict directional associations, single-tailed tests were used in estimating these relationships. That is, significance is determined based upon evidence of a positive relationship between indicators of activism and impulsivity as they relate to support for political violence and evidence of negative relationships between TRDM and CRT and support for political violence.

I estimated four models shown below

\hat{β} P o l V i o l e n c e = β_{0} + β_{1} Activism + β_{2} Controls

(1)

\hat{β} P o l V i o l e n c e = β_{3} + β_{4} Activism + β_{5} TRDM + β_{6} Controls

(2)

\hat{β} P o l V i o l e n c e = β_{7} + β_{8} Activism + β_{9} CRT + β_{10} Controls

(3)

\begin{array}{l} \hat{β} P o l V i o l e n c e = β_{11} + β_{12} Activism + β_{13} TRDM + β_{14} CRT + β_{15} Impulsivity \\ + β_{16} Controls \end{array}

(4)

In equation (1), I regressed support for political violence on the summary index of activism, political affiliation, and demographic controls. In equation (2), support for political violence was regressed on the index of activism as well as the TRDM score, political affiliation, and demographic controls. Equation (3) models support for political violence as a function of activism, score on the CRT, political affiliation, and demographic controls. Finally, I regress support for political violence on activism, TRDM, CRT, impulsivity, political affiliation, and demographic controls.

Results

Table 2 presents descriptive statistics for the variables in the following analyses as well as the Pearson’s correlations between each item and the dependent variable “Political Violence Index.” In this preliminary exploration, I found significant relationships at the bivariate level between 6 of the 15 indicators and support for political violence. Notably, while the activism index was strongly correlated with support for political violence, so too were all three measures of systematic decision-making. As anticipated, individuals who report more systematic and reflective decision-making processes (TRDM) and demonstrated these abilities (CRT) were less likely to support political violence. Contrastingly, more impulsive individuals were significantly more likely to support political violence.

Table 2.

Descriptive statistics.

Variable	M	SD	Minimum	Maximum	Correlation
Political violence index	2.409	1.305	1	7
Activism index	4.493	1.359	1	7	0.458**
Systematic decision-making
TRDM	16.155	2.377	4	20	−0.103*
CRT	1.730	1.227	0	3	−0.902*
Impulsivity	7.662	2.124	4	14	0.130**
Political affiliation
Democrat	0.503		0	1	0.046
Republican	0.131		0	1	−0.028
Independent	0.366		0	1	−0.028
Demographic
Race—White	0.632		0	1	−0.002
Race—Black	0.103		0	1	0.114*
Race—Chinese	0.121		0	1	−0.021
Race—Asian Indian	0.070		0	1	−0.081
Race—Other	0.087		0	1	0.014
Ethnicity—Hispanic	0.091		0	1	0.079†
Age	23.014	6.209	18	74	−0.100*
Male	0.392		0	1	0.044

TRDM: Thoughtfully Reflective Decision-Making; CRT: Cognitive Reflection Test.

N = 503. Pearson’s correlation “Corr.” represents the bivariate correlation with the dependent variable “Political Violence Index.”

†

p < .10; *p < .05; **p < .01 (using two-tailed significance test).

Considering next the control variables, no political affiliation was related to support for political violence at the bivariate level. Among the demographic indicators however, younger respondents were significantly more likely to support political violence, as were respondents who identify as black or Hispanic.

In sum, the bivariate associations in Table 2 indicated a strong positive relationship between activism and support for political violence, but critically, I also found significant associations between all three indicators of systematic decision-making and support for political violence.

I next estimate equations (1) to (4) in order to assess the multivariate relationships. Table 3 presents the OLS regression results for these analyses, with columns 1 and 2 reflecting equation (1), columns 3 and 4 reflecting equation (2), columns 5 and 6 reflecting equation (3), and columns 7 and 8 reflecting equation (4). All models include the full set of demographic and political affiliation variables.

Table 3.

OLS regression results.

Variable name	Model 1		Model 2		Model 3		Model 4
	$\hat{β}$	RSE	$\hat{β}$	RSE	$\hat{β}$	RSE	$\hat{β}$	RSE
Activism
Activism index	0.479**	0.042	0.489**	0.042	0.479**	0.042	0.486**	0.041
Sys. decision making
TRDM			−0.070**	0.021			−0.056**	0.022
CRT					−0.100*	0.044	−0.081*	0.043
Impulsivity							0.052*	0.027
Political affiliation
Independent	0.113	0.112	0.124	0.111	0.113	0.111	0.125	0.110
Republican	0.105	0.177	0.133	0.177	0.093	0.174	0.130	0.173
Demographic
Race—Black	0.450**	0.170	0.447**	0.167	0.398*	0.166	0.411*	0.165
Race—Chinese	0.161	0.163	0.206	0.163	0.230	0.163	0.215	0.164
Race—Asian Indian	−0.126	0.183	−0.104	0.183	−0.134	0.181	−0.132	0.180
Race—Other	0.153	0.196	0.193	0.195	0.114	0.199	0.115	0.200
Ethnicity—Hispanic	0.310†	0.185	0.288	0.184	0.298	0.183	0.270	0.185
Age	−0.018*	0.007	−0.015*	0.007	−0.018*	0.007	−0.012	0.007
Male	0.440**	0.114	0.475**	0.113	0.503*	0.119	0.506*	0.115
Constant	0.357	0.273	1.338**	0.419	0.486†	0.271	0.785	0.528
R ²	0.261		0.277		0.269		0.289

$\hat{β}$ = ordinary least squares estimate; RSE = robust standard error; TRDM: Thoughtfully Reflective Decision-Making; CRT: Cognitive Reflection Test.

Omitted category: Female who identifies as a Democrat and self-describes as non-Hispanic and white. The R² statistic provided serves as an indicator of model fit. Sample size for all models is 503.

†

p < .10. *p < .05. **p < .01 (one-tailed significance test used for Activism and Sys. Decision Making items; two-tailed significance test used for Political Affiliation and Demographic indicators).

Activism

In support of Hypothesis 1, I found evidence of a robust positive relationship between the activism and support for political violence. Across all models, this relationship remains significant at p < .01. Controlling for all else, averaging one point higher on the activism index was associated with reporting about a half of a point higher on the support for political violence. Practically, this suggests that compared with someone who strongly disagreed with all items on the activism index, on average an individual who was very involved and invested in activism would likely be neutral or perhaps supportive of political violence.

Systematic decision-making

Turning to the indicators of systematic decision-making, I found strong support for Hypothesis 2. In models 2 and 4, I observed a significant and negative relationship between self-reported TRDM and support for political violence. Similarly, in model 3 I found a negative association between performance on the CRT and support for political violence. Likewise, in model 4 when controlling for TRDM and impulsivity the relationship remains significant at p < .05. Finally, in model 4, I found self-reported impulsivity to be positively associated with support for political violence at p < .05, even when controlling the remaining measures of systematic decision-making and attitudes toward activism.

Control variables

Examining the control variables, compared to those who identified as a democrat, I did not find either self-proclaimed independents or republicans to be more likely to support political violence. Contrastingly, compared to women, men were significantly more likely to support political violence across all models. In models 1 and 2, younger individuals were more likely to support political violence; however, this relationship was no longer significant when controlling for CRT performance and impulsivity. Finally, compared to individuals who identified as non-Hispanic and White, identifying as Black was positively associated with support for political violence, whereas identifying as Spanish, Hispanic, or Latin(x) was marginally positively associated with support for political violence before controlling accounting for measures of systematic decision-making.

Discussion

This paper presents novel empirical evidence for cognitive psychological and behavioral economic decision-making frameworks in describing support for political violence. The above findings also suggest that there remains much to learn from the study of activism and support for political violence. In this study, I examined the relationship between systematic decision-making and support for political violence among a sample of 503 students at a large public university. Due to an ongoing need for empirical studies of precursors to radicalization and support for violent extremism, this provides a valuable contribution through developing cognitive psychological frames within extant radicalization models.

These findings provide important distinguishing factors between individuals who support political violence and those who do not. In Moskalenko and McCauley’s (2009) framework attitudes toward activism are positively associated with indicators of radicalism (or support for political violence). Similarly, I find a robust relationship between activism attitudes and support for political violence across all models. Of the 45 individuals reporting some affirmative support for political violence (scored 4.2 or higher on the political violence index), only two were neutral or not supportive of activism on average (scored 4.0 or lower on the activism index). While unsurprising, these findings highlight the importance of this relationship more generally.

Dual process theory suggests that when individuals choose to engage in effortful decision-making, they are more likely to pursue action consistent with their desired outcomes and less likely to engage in criminal behavior or fall victim to cognitive misconceptions (Kahneman, 2003). Here, I find a significant and negative relationship between systematic decision-making processes and support for political violence among a biographically available population. This generally supports dual process theory, highlighting the value of research applying the constructs of TRDM and CRT.

Next, it is important to note several key limitations of this study. First, regarding the sample, the limited response rate for this study raises concerns about the representativeness of the findings. A response rate of 8.6% is underwhelming and comparing the respondents to the demographic profile of the University highlights some differences. Survey respondents were more likely to be female, and White than the student body writ large. Consequently, these findings should be interpreted with caution. Moreover, as research has found a positive association between TRDM and higher education, this sample represents the upper tail of the distribution of TRDM scores. Contrasted with a prominent non-university sample, respondents in the present study reported a TRDM score nearly 8 points higher on average than in Paternoster and Pogarsky (2009). While the Paternoster and Pogarsky (2009) sample was drawn from a much younger population (mean age between 15 and 16; 23 years old in the present study), this suggests that further research should explore these patterns within the general population to discern the relationship given a much broader distribution of age and TRDM scores. Despite this, college students are a biographically available and often politically engaged population (Wiltfang and McAdam, 1991), and thus merit careful attention when seeking to understand the links studied here.

Second, while the index of support for political violence represents five forms of support, it does not span the universe of behaviors. For example, willingness to create or share inflammatory or misleading material online is not considered. Moreover, expressions of support on social media for activism and political violence alike are not measured here. Considering these, future work should examine the role of social media as an avenue of expressing support for political violence. Finally, while survey indicators assessed global self-reported decision-making (Paternoster and Pogarsky, 2009), global self-assessed impulsivity (Grasmick et al., 1993), and effortful cognitive reflection (Frederick, 2005), these elements do not reflect the probability of engaging reflective cognition when encountering information and arguments supportive of political violence. Accordingly, vignette studies examining these specific contexts are necessary.

Third, though the present study focuses on the association between decision-making at the individual level and support for political violence, a rich body of research has commented on corresponding factors at larger units of analysis. To wit, research has highlighted the important role of interpersonal interactions (McCauley and Moskalenko, 2016), group dynamics (Sageman, 2004, 2017) social movements (Della Porta, 2018), and political opportunity (Della Porta, 2008; Malthaner and Waldmann, 2014), in understanding support for (and the use of) political violence across geographic contexts. As such, the findings presented here represent a promising additional layer of nuance at the individual level that merits further exploration in the study of political violence.

Bearing these cautions in mind, to my knowledge these data are the first of their kind to examine attitudes toward activism and political violence while also accounting for multiple measures of systematic decision-making. Taken broadly, these findings support the importance of understanding activism and its relationship with support for political violence. This echoes prior work examining these relationships both cross-nationally and internationally using student samples (Moskalenko and McCauley, 2009), and among an incarcerated sample in the United States (Decker and Pyrooz, 2019). Indeed, the relationship between activism and support for political violence is not explained away across models, remaining highly significant. Regarding systematic decision-making, this study finds robust support for indicators of dual process theory as they inform attitudes supportive of political violence. Notably, I also find independent explanatory contributions of TRDM and performance on the CRT as they relate to supporting political violence. This reinforces the theoretical support for the distinct facets of global self-assessed systematic cognition (TRDM) and a more direct measure of cognitive reflection (CRT) that they each represent.

This study may also provide insight for CVE programming. These findings suggest that within a population supportive of political activism, more impulsive individuals and those demonstrate or prioritize less systematic decision-making tend to be more supportive of political violence, irrespective of political affiliation. Thus, when examining the role that programs have on limiting support for political violence writ large, efforts should explore the value of providing instruction that develops the ability of at-risk individuals to slow down and thoughtfully consider the immediate and far-reaching consequences of their actions. Building on research suggesting that systematic decision-making is a trait which can be improved (Paternoster and Pogarsky, 2009), this research may apply to resilience building. Second, these findings highlight the importance of raising awareness within activist networks of the both efficacy of non-violent dissent (Chenoweth et al., 2011), as well as the deleterious legal and pragmatic effects of unreflective and impulsive acts.

Future research on this topic should consider three primary avenues. First, research should further examine the relationship between systematic decision-making and support for, as well as participation in political violence. Examining these relationships using a nationally representative sample or in a population which includes individuals known to have participated in political violence could further support or refute the generality and specificity of the findings presented here. Second, future work should consider what effects relational ties between supporters and participants of political violence, as well as those who “stay home” may have on decision-making. Finally, quantitative methods should be supplemented with a qualitative examination of the experiences of supporters and perpetrators of political violence. Through in-depth interviews or other primary data collection, a more grounded understanding could be attained for the context-specific impacts of decision-making. With this as a guide, researchers could better understand the causes of political violence and aid in the crafting and implementation of strategies to prevent it.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Michael H Becker

Notes

Author biography

Michael H Becker, MA, is a Faculty Research Assistant at the Department of Criminology and Criminal Justice at the University of Maryland—College Park. His research interests include individual and group causes of participation in violent extremism, process and outcome measures in P/CVE programming, and theoretical testing.

References

Asal

Park

Rethemeyer

, et al. (2016) With friends like these . . . why terrorist organizations ally. International Public Management Journal 19(1): 1–30.

Barrett

(2019) FBI faces skepticism over its efforts against domestic terrorism. The Washington Post, 5 August. Available at: https://www.washingtonpost.com/national-security/fbi-faces-skepticism-over-its-anti-domestic-terror-efforts/2019/08/04/c9c928bc-b6e0-11e9-b3b4-2bb69e8c4e39_story.html (accessed 14 August 2019).

Bartlett

Miller

(2012) The edge of violence: Towards telling the difference between violent and non-violent radicalization. Terrorism and Political Violence 24(1): 1–21.

Bentham

(1776) A Fragment on Government. New York: Cambridge University Press.

Bonilla

Rosa

(2015) #Ferguson: Digital protest, hashtag ethnography, and the racial politics of social media in the United States: #Ferguson. American Ethnologist 42(1): 4–17.

Borum

(2011) Radicalization into violent extremism I: A review of social science theories. Journal of Strategic Security 4(4): 7–36.

Caplin

Dean

(2015) Revealed preference, rational inattention, and costly information acquisition. American Economic Review 105(7): 2183–2203.

Chenoweth

Stephan

(2011) Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict. New York: Columbia University Press.

Chermak

Freilich

Parkin

, et al. (2012) American terrorism and extremist crime data sources and selectivity bias: An investigation focusing on homicide events committed by far-right extremists. Journal of Quantitative Criminology 28(1): 191–218.

10.

Crenshaw

(1987) Theories of terrorism: Instrumental and organizational approaches. Journal of Strategic Studies 10(4): 13–31.

11.

Decker

Pyrooz

(2019) Activism and radicalism in prison: Measurement and correlates in a large sample of inmates in Texas. Justice Quarterly 36: 787–815.

12.

Della Porta

(2008) Research on social movements and political violence. Qualitative Sociology 31(3): 221–230.

13.

Della Porta

(2018) Radicalization: A relational perspective. Annual Review of Political Science 21(1): 461–474.

14.

Department of Homeland Security (2017) Countering Violent Extremism Task Force. Available at: https://www.dhs.gov/cve (accessed 14 August 2019).

15.

Dugan

Chenoweth

(2012) Moving beyond deterrence: The effectiveness of raising the expected utility of abstaining from terrorism in Israel. American Sociological Review 77(4): 597–624.

16.

Eligon

(2019) The El Paso Screed, and the Racist Doctrine behind It. The New York Times, 7 August. Available at: https://www.nytimes.com/2019/08/07/us/el-paso-shooting-racism.html (accessed 14 August 2019).

17.

Frederick

(2005) Cognitive reflection and decision making. Journal of Economic Perspectives 19(4): 25–42.

18.

Grasmick

Tittle

Bursik

, et al. (1993) Testing the core empirical implications of Gottfredson and Hirschi’s general theory of crime. Journal of Research in Crime and Delinquency 30(1): 5–29.

19.

Harris-Hogan

Barrelle

Zammit

(2016) What is countering violent extremism? Exploring CVE policy and practice in Australia. Behavioral Sciences of Terrorism and Political Aggression 8(1): 6–24.

20.

Hirschi

Gottfredson

(1983) Age and the explanation of crime. American Journal of Sociology 89(3): 552–584.

21.

Horgan

(2011) Lessons Learned Since the Terrorist Attacks of September 11, 2001. Washington, DC. Available at: https://www.c-span.org/video/?301326-1/lessons-learned-terrorist-attacks-september-11-2001 (accessed 30 November 2019).

22.

Jacobs

(1997) The Way the Wind Blew: A History of the Weather Underground. New York: Verso.

23.

Jensen

Seate

James

(2018) Radicalization to violence: A pathway approach to studying extremism. Terrorism and Political Violence. Epub ahead of print 9 April. DOI: 10.1080/09546553.2018.1442330.

24.

Kahneman

(2003) A perspective on judgment and choice: Mapping bounded rationality. American Psychologist 58(9): 697–720.

25.

Kearns

Asal

Walsh

, et al. (2018) Political action as a function of grievances, risk, and social identity: An experimental approach. Epub ahead of print 25 October. DOI: 10.1080/1057610X.2018.1507790.

26.

LaFree

Ackerman

(2009) The empirical study of terrorism: Social and legal research. Annual Review of Law and Social Science 5(1): 347–374.

27.

LaFree

Dugan

Miller

(2014) Putting Terrorism in Context: Lessons from the Global Terrorism Database. London: Routledge.

28.

LaFree

Jensen

James

, et al. (2018) Correlates of violent political extremism in the United States*. Criminology 56(2): 233–268.

29.

Louderback

Antonaccio

(2017) Exploring cognitive decision-making processes, computer-focused cyber deviance involvement and victimization: The role of thoughtfully reflective decision-making. Journal of Research in Crime and Delinquency 54(5): 639–679.

30.

McAdam

(1986) Recruitment to high-risk activism: The case of freedom summer. American Journal of Sociology 92(1): 64–90.

31.

McCauley

Moskalenko

(2017) Understanding political radicalization: The two-pyramids model. American Psychologist 72(3): 205–216.

32.

McCauley

Moskalenko

(2016) Friction: How Conflict Radicalizes Them and Us. New York: Oxford University Press.

33.

Malthaner

Waldmann

(2014) The radical milieu: Conceptualizing the supportive social environment of terrorist groups. Studies in Conflict & Terrorism 37(12): 979–998.

34.

Matsueda

Heimer

(1987) Race, family structure, and delinquency: A test of differential association and social control theories. American Sociological Review 52(6): 826–840.

35.

Moskalenko

McCauley

(2009) Measuring political mobilization: The distinction between activism and radicalism. Terrorism and Political Violence 21(2): 239–260.

36.

Nagin

Pogarsky

(2001) Integrating celerity, impulsivity, and extralegal sanction threats into a model of general deterrence: Theory and evidence. Criminology 39: 865–892.

37.

Nepstad

Smith

(1999) Rethinking recruitment to high-risk/cost activism: The case of Nicaragua exchange. Mobilization: An International Quarterly 4(1): 25–40.

38.

Neumann

(2013) The trouble with radicalization. International Affairs 89(4): 873–893.

39.

Paternoster

Pogarsky

(2009) Rational choice, agency and thoughtfully reflective decision making: The short and long-term consequences of making good choices. Journal of Quantitative Criminology 25(2): 103–127.

40.

Sageman

(2004) Understanding Terror Networks. Philadelphia, PA: University of Pennsylvania Press.

41.

Sageman

(2017) Turning to Political Violence: The Emergence of Terrorism. Philadelphia, PA: University of Pennsylvania Press.

42.

Santoro

Azab

(2015) Arab American protest in the terror decade: Macro- and micro-level response to post-9/11 repression. Social Problems 62(2): 219–240.

43.

Simon

(2000) Bounded rationality in social science: Today and tomorrow. Mind & Society 1(1): 25–39.

44.

Tversky

Kahneman

(1981) The framing of decisions and the psychology of choice. Science 211(4481): 453–458.

45.

Walsh

(2000) Organizing the scale of labor regulation in the United States: Service-sector activism in the city. Environment and Planning A: Economy and Space 32(9): 1593–1610.

46.

White

(1989) From peaceful protest to guerrilla war: Micromobilization of the Provisional Irish Republican Army. American Journal of Sociology 94(6): 1277–1302.

47.

Wiltfang

McAdam

(1991) The costs and risks of social activism: A study of sanctuary movement activism. Social Forces 69(4): 987–1010.

48.

Wray

(2019) Oversight of the Federal Bureau of Investigation. Washington, DC. Available at: https://www.judiciary.senate.gov/meetings/07/23/2019/oversight-of-the-federal-bureau-of-investigation