Violence and Non-Violence in Anti-Government Mobilization

Anti-Government Mobilization

Conflict research and media attention to anti-government mobilization is often skewed towards violent events, such as the current civil war in Syria. General definitions of conflict, however, emphasize the underlying incompatibilities or conflict of interests between actors rather than the specific means that actors use. Although grievances against the government can motivate violence, they can also and often do motivate other responses, including forms of non-violent direct action such as protests, strikes, and boycotts. During the Arab Spring, for example, long-standing dictators in Tunisia and Egypt were ousted through large-scale non-violent mobilization. Despite the common focus on violence, large-scale non-violent challenges to the government have been more common than violent challenges, and non-violent mobilization has often posed more severe challenges to authoritarian governments than violent campaigns.

In this policy brief, we describe what our research tells us about grievances over the government, when actors chose non-violent means to contest the government, and how this compares to the use of violence. We draw on insights from a new data set documenting claims made by dissident organizations against the government that allows for examining when we are more likely to see either violent or non-violent mobilization over these claims.

Grievances and Active Dissent against the Government

Many researchers argue that conflict over different incompatibilities tends to be driven by distinct factors and processes. For example, ethnic groups are more likely to seek independence or greater autonomy, and the risk of violence in such disputes increases when groups face greater political exclusion and marginalization and have a higher ability to challenge the government militarily (e.g., Buhaug, 2006; Cederman et al., 2013, Cunningham, 2013). However, there has been much less attention to conflicts over the central government – including claims over the political system, leader or government selection – and the specific means of mobilization that actors choose.

We argue that the key underlying motives are conceptually similar across violent and nonviolent large-scale popular mobilization against the government. Both arise under political and economic exclusion in autocratic systems, where acts of dissent inherently involve a challenge to government authority and must rely on direct action as conventional forms of politics are closed off and opposition is not tolerated. In democracies, actors have a wide range of opportunities to influence governments and policies through legal avenues and competitive elections. Indeed, the only clear case of new violent mobilization against the central government emerging in a democracy is the 1971 Marxist JVP rebellion in Sri Lanka, although some democracies see violent separatist movements by ethnic groups that reject the existing nation-state. Moreover, non-violent direct action challenging the government itself is also rare outside autocracies, and generally restricted to cases where the degree of democracy is questionable, such as the 2004 Orange Revolution in Ukraine, following contested elections widely seen as stolen.

State capacity is often defined by control of the means of violence, and there is a common folk wisdom that violent uprisings are the most serious threats to the government. However, our research actually shows that large-scale non-violent challenges to the government have been more common than violent challenges and often have a bigger impact (see Dahl et al., 2016; Gleditsch et al., 2016). Figure 1 compares the number of non-violent outbreaks of mass mobilization against the government to violent challenges and shows that the former have been more common. By contrast, violence has been more common than non-violence in mobilization for claims over territory, typically involving distinct ethnic groups. Moreover, non-violent mobilization has more often preceded concessions or reforms than violent challenges. Gleditsch et al. (2016) shows leaders have lost office twice as often under non-violent than violent challenges (see also Chenoweth & Stephan, 2011).

Mobilizing large-scale dissent is difficult in a dictatorship, and the government typically uses repression to deter or suppress dissent. Given the high risk to participants, it is not surprising that large-scale mobilization against the government remains rare, even if grievances are widespread. However, comparing cases where we see active dissent reveals important clear differences between non-violent and violent challenges.

Whereas violent challenges to the government can evade repression by going underground, non-violent challenges inherently rely on public opposition. Since individual participants must expose themselves to the government, it is a risky proposition without sufficiently high total participation and efforts to mobilize will be unlikely to impose significant costs on the government unless participation is high. Hence, mobilizations will often fail to materialize when participation cannot be expected to be sufficiently large, even if grievances against the government are widely held.

With high participation, however, large-scale non-violent dissent and refusal to comply with state authority can impose substantial costs on the government and increase the likelihood of elite defections. The risks to individual participants from repression decline rapidly with larger total numbers. It is easier to attract more participants for non-violent campaigns as no particular skills are required, and dissent does not require leaving civilian life behind to the same extent as joining a rebel group. By contrast, the feasible participation in violent dissent is normally capped at a low level, at least in the short run, since converting sympathizers into capable rebels requires arms and training.

It is generally more difficult for the government to repress large numbers in non-violent protest, and doing so leads to higher risks of international condemnation. If violence alienates some potential participants, then maximum participation under violence declines even further relative to non-violence. Thus, even under a conservative and questionable assumption that non-violent dissent is never absolutely more effective than violent dissent with the same number of active participants, non-violent dissent can still be relatively more effective than violence if a sufficiently larger number of active participants can be mobilized. Thus, dissidents with broad appeal and the ability to mobilize more participants should be more likely to use non-violent tactics, while the resort to violence often arises as an adaptation to low mobilization capacity. We show elsewhere that data on participants and proxies for resources in different types of mass mobilizations are consistent with these expectations (Dahl et al., 2016; White et al, 2015).

From Claims to Mobilization

The path from grievances over the government to manifest claims is complex, and incipient mobilizations can fail at many stages. Our research separates the process leading to large-scale mobilization into stages to understand why some countries see smaller-scale anti-regime activity while others experience large-scale mobilization though either violent or non-violent tactics, and while yet others see no anti-regime activity at all. Claims against governments and mobilization are led by organizations willing to engage in anti-regime action. We can think of organizational expression of anti-regime grievances as a precondition for large-scale mobilization, whether non-violent or violent. We have collected annual data on expression of grievances by organized dissidents for nearly 100 randomly selected countries, not including consolidated democracies. We identify these through named dissident organizations or other social organizations (e.g., unions, students, religious congregations) making public statements calling for a fundamental change to the political system (e.g., establishing democracy in a dictatorship), the composition of the central government, or the electoral process (e.g., claims of electoral fraud). Our data on governmental incompatibilities identify conflicts over government independent of mobilization, where we can examine what drives these disputes to evolve into violent or non-violent mobilization, or no large-scale dissent.

Contrary to the common claim that grievances are ubiquitous and cannot explain political action, we show that organizations expressing anti-regime grievances are not ubiquitous. We find active claims in only about 50% of the possible cases with plausible grievances, indicating that a significant share either lack sufficient grievances to result in public claims, or that the costs to publically challenge the government are so high that potential dissidents either fail to organize or choose not to do so.

Most governmental incompatibilities do not lead to either violent or non-violent mass mobilization. In our data, civil war takes place about 20% of the time when organizations are expressing anti-regime grievances, while non-violent campaigns occur about 6% of the time. However, even though violent civil wars are more frequent in terms of incidence than large-scale non-violent campaigns, this primarily reflects how civil wars tend to last much longer than non-violent campaigns, which tend to either succeed or fail relatively quickly, while some violent insurgencies have lasted for decades. When we look at new outbreaks we actually see more non-violent campaigns breaking out than new outbreaks of civil wars, especially after 2000. In our data, we find about twice as many non-violent campaign onsets over the government as civil war onsets over governmental incompatibilities. Moreover, while existing data show that civil wars appear to have become less common and generally less severe since the end of the Cold War, non-violent mass mobilization has become relatively more common.