Direct Democracy: Protest Catalyst or Protest Alternative?

Kitschelt (1986, p. 68), for example, interprets referendums by anti-nuclear groups as a reason why protest took on an assimilative form in the USA (as opposed to confrontational strategies in the closed systems of France and West Germany). Kriesi and Wisler (1996) show that direct democracy in Switzerland induces movements to use its instruments, thus moderating their action repertoire.

We are of course aware of the fact that a sub-national analysis of Switzerland cannot completely overcome the problems of analyzing the causal effect direct democracy has on protest behavior. While internationally there are not any countries with a similarly high level of direct democracy as in Switzerland, within Switzerland we do not have the fully counterfactual outcome, i.e., no direct democracy at all (see Rubin 1974). Focusing on institutional configurations of direct democratic instruments, some cantons exhibit very few opportunities of direct democratic participation and come, compared to other cantons, very close to the counterfactual. Individuals in these cantons can therefore serve as our—non-randomly assigned—control group (Achen 1986; Campbell and Stanley 1963). Moreover, following King et al. (1995, p. 477) it is important for the evaluation of causal explanations in political science to test a given hypothesis in different contexts and confront the respective findings. Because previous research on direct democracy’s impact has largely focused on the U.S. states, adding empirical data for the Swiss case can therefore be seen as a further step toward causal inference.

Norris (2009, 639 et seq.), for instance, challenges these labels and suggests a new distinction “between citizen-oriented action, relating mainly to elections and parties, and cause-oriented repertoires, which focus attention upon specific issues and policy concerns, exemplified by consumer politics, […] petitioning, demonstrations, and protests.”

Eisinger (1973, p. 13), however, draws a clear distinction between political protest and more “costly” forms such, as political violence.

It has to be noted that our data from the Swiss Electoral Studies (Selects 2003) does not include any other items of the various forms of protest behavior mentioned above.

Institutions are of course the result of citizens’ collective action and may therefore be endogenous to individual behavior over the long-run (Foweraker and Landman 1997). We argue, however, that institutional arrangements can still be seen as exogenous framework conditions that cannot be changed by an individual in the short and medium-run; instead, they influence individual preferences and behavior patterns (see Huckfeldt and Sprague 1987, p. 1200).

This institutional approach is one of three accounts of protest behavior. Another approach centers around people’s economic situations. If an individual personally experiences economic grievances, he or she is likely to protest. In particular, relative deprivation is seen as the driving force for protest (Gurr 1970). Additionally, a cultural approach can be identified that challenges the rational cost-benefit analysis of the economic view (Chong 1991). In that sense protest is a culturally inherited form of participation, and cultural differences account for differences in political participation such as protests (Hofstede 1991). In particular, the cultural resource of social (or interpersonal) trust is found to be associated with protest behavior (Benson and Rochon 2004; Winters 2008; Valencia et al. 2010).

This line of reasoning is in accordance with the views of the so-called “Progressive era reformers” (Smith and Tolbert 2004, p. 3). Here, direct democratic processes have an educative effect on the people (Garner 1907; Bryce 1910; Weyl 1912). More recent studies of the USA provide empirical evidence for positive effects on social participation (Tolbert et al. 2003; Smith and Tolbert 2004; Tolbert and Bowen 2008; Boehmke and Bowen 2010) and conventional political participation through voting (Tolbert et al. 2001; Tolbert and Smith 2005; Tolbert and Bowen 2008; and more recently and specifically Dyck and Seabrook 2010). Furthermore, mobilization effects on independent voters to cast their ballot (Donovan et al. 2009) and positive effects on political trust (Smith and Tolbert 2004), as well as social trust (Dyck 2008), political knowledge (Tolbert et al. 2003; Schlozman and Yohai 2008), and political support (Bühlmann 2007) can be shown.

Eisinger (973, 27 et seq.) hypothesizes in fact a curve-linear relationship between openness of government and protest. In extremely closed systems, protest would be neither a viable nor a fruitful strategy. Arguably, however, such an authoritarian system that suppresses protests is not to be found in the Swiss context.

While only a special one, the most evident case would be when individuals protest against insufficient opportunities of participation or against representatives and their decisions. In direct democracies, on the other hand, such protest is not found (Opp 1996, p. 230).

Moreover, one can argue logically that it is the more stable (“sticky”) contextual condition, which causally affects the more volatile (“loose”) individual behavior, and not vice versa (Davis 1985).

Coding for thresholds and corresponding index points is described in detail by Stutzer and Frey (2000).

We are grateful to our anonymous reviewer for pointing this out.

For an overview see for example Dalton et al. (2009) or Opp (1996).

Level of education is highly correlated with personal income, but the latter contains more missing values. To avoid multicollinearity and to keep as many observations as possible, we use the level of education in our analysis.

We use data of protest events between 1999 and 2003 with at least 1,000 protesters (Année politique Suisse). In addition, there exists a highly significant and strongly negative effect (β = −11.4; SE = 3.3) of direct democracy on the number of major protest events in the Swiss Cantons (n = 26), even after introducing our (contextual) control variables (R ² = 0.58). In other words, the extent of direct democracy is related to fewer protest events in the Swiss cantons.

In this respect, Ladner (2007) finds both a greater number and increased importance of local parliaments in French and Italian-speaking cantons than in German-speaking cantons. Knüsel (1994) argues that language regions are influenced by their respective neighboring countries: The representative model of democracy in Italy and France delegates responsibility and power away from the individual to the unitary state. In German-speaking cantons, on the other hand, citizens embody the idea of a “small” state and are thus left with more power and responsibility.

In analyses not documented here, we tested further individual variables that could be connected to protest behavior, namely variables for political trust and political knowledge. These variables are not significant in our model, do not change the model estimates, and are therefore excluded. These results are available from the authors upon request.

We also tested further contextual indicators of our individual variables, namely variables for age distribution and education pattern per canton. Again, these variables are not significant in our model, do not change the model estimates, and are therefore excluded. Moreover, further analyses of potential cross-level interactions between direct democratic settings and individual accounts of protest behavior (not presented here) do not show significant effects, and thus do not support the educative reasoning. These results are available from the authors upon request.

Although our empirical results clearly favor the hypothesized negative relationship, they do not clarify the mechanism: is the decline in protest behavior due to the deliberative environment of direct democracy or is it because citizens use direct democratic votes as a valve? As evidence for the valve effect, the degree of direct democracy should also be positively correlated with individual participation in popular votes. In further analyses of the same models (not documented here), the direct democracy variable indeed exerts a highly significant and positive effect on individual participation in popular votes. These results are available from the authors upon request. With this in mind, our results seem to support the argument that direct democratic institutions indeed act as valve. As we detail below, however, more empirical investigation is needed to scrutinize the underlying causal mechanism.

With regard to other cantonal idiosyncrasies, it could also be that it is not direct democracy per se that reduces protests, but rather that direct democratic cantons are more likely to pass laws that make protests unnecessary—for example by recognizing minority rights. In order to rule out such indirect effects we correlate the direct democracy measure with data on cantonal recognition of religious minorities by Christmann (2010, p. 21). We find an insignificant and low (0.2) correlation, indicating that direct democratic cantons are not more likely to pass laws that effectively reduce or suppress protests.