Abstract
In recent years, the question of what form a revolutionary uprising will
take — armed or unarmed — has been raised more and more often. This
is because, as shown by numerous studies, revolutionary nonviolence can
explain why the uprising failed or succeed to lead to democracy. In the
recent decades the likelihood of revolution being nonviolent appears to
have significantly increased, but it is still not clear why this
tendency is observed. Moreover, there are only a few quantitative
cross-national studies on this topic, in which the authors tried to
explain the apparent pattern. However, none of them considered political
factors separately. This paper tests the hypothesis that a country’s
level of democracy can inhibit the spread of revolutionary violence. By
applying logistic regression to the NAVCO database, we analyze 400
episodes and conclude that, in general, the more democratic the
political system, the more likely the revolution take a nonviolent form.
Nevertheless, the various revolutionary events could be of a rather
different nature, and it is further shown that the level of democracy
matters only for sociopolitical revolutions, while for separatist
revolutions it does not play a significant role. In other words,
democracy can stop effectively violence only in revolutionary episodes
caused by sociopolitical (but not ethno-nationalist) grievances.