The Arab Uprising, and specifically the ongoing Syrian crisis, has got me thinking about a number of things that are related to the future of the Middle East. But for now, my thoughts have converged on one topic – the relationship between (violent) repression and dissent. The question I have been pondering, which I believe will have direct relevance to socio-political events in the Arab world for generations to come, is the following:

Why do dissidents sometimes respond to physical state repression by increasing their protest behavior and at other times respond by decreasing their protest behavior? The answer seems simple, right? Well, after doing some basic searching, I realized that it is a little bit more complicated than I thought.

Scholars of political repression have produced and developed a good number of systematic studies on the conditions under which national governing elites resort to repressive action to counter and/or deter actual or latent dissident behavior (Boudreau, 2004; Davenport 2004 & 2007; Della Porta and Reiter, 1998; Earl, 2003; Ekiert and Kubick, 1999; Ferrara, 2003; Francisco, 2004). Yet, despite this sizeable body of academic research, theoretical work and empirical analysis on the strategic interaction between the agents of repression (governments) and the agents of dissent (protestors) are lacking, creating an important gap in the literature.

The relationship between repression and dissent is important, I think, for at least two reasons. First, it is closely tied to one of the major debates in the literature on violent political conflict: many rational choice explanations, including the resource mobilization/political process school (McAdam, McCarthy, Zald 1996), suggest that repression will reduce dissident activity whereas the relative deprivation approach (Gurr, 1970) suggests that repression will increase dissident activity. The first group contends that repression raises costs to collective action whereas the second contends that repression will increase people’s sense of relative deprivation. Second, the literature is plagued by inconsistent empirical findings connected to the contrary theoretical expectations. As such, it provides a puzzle for scholars interested in evaluating general explanations of political phenomena by confronting them with systematically gathered evidence.

I am more curious about political phenomena after governments initiate physical repressive action against dissidents to crush collective action (obviously, the causes of initial dissident behavior would be relevant to any systematic study as well). What are the crucial factors that influence dissidents’ response? I suspect that a careful investigation of this sequence of events could challenge some of the findings of several prominent repression studies that have shown empirical support for the domestic democratic peace, i.e., that democracy decreases state repression (Davenport, 2007).

In the repression literature, there are two major explanations for the observation that repression sometimes deters and at other times spurs dissident activity. The first explanation, presented by Lichbach (1987), suggests that dissidents view nonviolent and violent protest activity as substitutes and select the type that best achieves their goals, depending on state repression and concessions. Lichbach presumes that because dissidents are interested in maximizing the shift in policy, they will pursue the most effective protest activity. Hence, if the state responds to violent protest behavior with repression (as opposed to accommodation), then dissidents will abandon violent protest behavior in favor of nonviolent protest behavior. Similarly, if the state repressed nonviolent protest behavior, then the dissidents will respond with violent protest behavior.

The second explanation, offered by Gupta, Singh, and Sprague (1993), suggests that context (i.e., the type of regime) explains the difference in responses. Here, dissidents are believed to choose between economic and political activity (rather than nonviolent and violent protest), and protest behavior is a function of government coercion, regime type (democratic or autocratic), group identity, and benefits from economic activity. Gupta, Singh, and Sprague find that in democracies repression is positively (and linearly) associated with both nonviolent and violent protest behavior, but in autocracies repression has an inverted –U relationship with both nonviolent and violent protest behavior.

Building on Rasler (1996), I thought of a third explanation which focuses on timing (i.e., short-run vs. long-run) effects and concessions by the state. I believe that it is important to distinguish between short-run reactions to repression and long-run reactions to repression (again, I am only interested in the effects ofphysical repression and not those types of repression that indirectly constrain political behavior). Examples of physical repression during a protest, strike, or demonstration include individual or mass arrests, torture, beatings, disappearance, imprisonment, and individual assassinations or mass killings.

I would expect that in the short run, dissidents perceive physical repression as a cost and, hence, decrease their protest behavior. Yet, as time goes by (it could take weeks or months), grievances become more acute and lead to a lagged spur to new protest activity. Thus, I believe that a single act of physical repression has both a negative “instantaneous effect” on protest activity and a positive “lagged effect” on protest activity. But we should also consider not only repression, but also concessions by the state. Thus, concessions, often in revolutionary contexts, could well spur further protest. So in sum, in the short-run, physical government repression decreases protest behavior, whereas in the long-run, physical government repression increases protest behavior.

People usually rebel if they become convinced that dissent will achieve the collective good (Muller and Opp 1986; Finkel, Muller, and Opp 1989). If the value of the collective good is combined with a high expectation of success, people are likely to participate in mass actions. The factors that are likely to increase the expected value of a collective good are individual assessments about whether their participation will make a difference in achieving the public good, and expectations that group action will be successful. Government concessions to highly visible groups enhance their perceived influence and increase the probability that individuals will join them for mass action (Muller and Opp, 1986). Thus, government concessions increase protest behavior.

In order to develop more precise empirical tests and more refined causal relationships between repression and dissent, we need to focus on sequences of interactions. A number of scholars have taken interest in studying sequences of behavior (Abell, 1993; Abbott, 1992: Dixon 1988: Heise 1989; Schrodt 1990: Schrodt and Gerner 1997) and several have done so with a particular focus on the repression-dissent nexus (Davies and McDaniel, 1996; Khawaja, 1993, 1994, 1995; Olzak, 1992; Poe et al, 1996: Snyder, 1976; Tilly, 1985).

Whether it makes sense to construct one’s theories by thinking about the sequences of interactions among actors depends entirely on the questions one asks. I personally am interested in examining dissident responses to physical state repression by paying close attention to processes and sequences of actions located within constraining or enabling structures (it should be noted that Lichbach’s model, mentioned above, relies primarily on sequential analysis).

I would probably design the statistical analysis by specifying an action-reaction equation where dissident activity is a function of the dissident’s previous behavior, previous government action, and a lagged variable of past government action, but instead of using the regression techniques used in the repression studies I referenced above (which rely on aggregate analysis), I would follow Dixon (1988) who makes a case for the superiority of using a logit estimator.

Let’s say I work on this study and go about operationalizing and measuring my variables and conduct some statistical tests, if I find in my results that there is strong empirical support for changing effects of repression over time, then I would say that we need to examine dynamic relationships between governments and challengers over time a littel bit more carefully. These interactive relationships between government and non-government actors may have more explanatory power than the type of the regime under study, be it a democracy or a non-democracy, thus challenging the findings of the domestic democratic peace. So how much does it really matter that the Syrian government is a brutal, authoritarian, killing machine? Ok, maybe it still matters big time (these guys are nuts!). But you get my point. It might have to do with a little bit more than the fact that Damascus is not Athens.