I read a bit of the book. There’s some framing around it by the authors and popular press that I don’t think quite matches up with the data. In the end, it is true that movements that are predominantly nonviolent tend to win more often, but there is often a violent element that plays a role.

(Apologies for any mistakes in my transcription from the book.)

To quote it:

Our central contention is that nonviolent campaigns have a participation advantage over violent insurgencies, which is an important factor in determining campaign outcomes. The moral, physical, informational, and commitment barriers to participation are much lower for nonviolent resistance than for violent insurgency. Higher levels of participation contribute to a number of mechanisms necessary for success, including enhanced resilience, higher probabilities of tactical innovation, expanded civic disruption (thereby raising the costs to the regime of maintaining the status quo), and loyalty shifts involving the opponent’s erstwhile supporters, including members of security forces.

Which sounds like nonviolent campaigns win, right? Reading on shows it’s not quite that simple.

It is appropriate here to briefly define the terms to which we will consistently refer in this book. First, we should distinguish violent and nonviolent tactics. As noted earlier, there are some difficulties with labeling one campaign as violent and another as nonviolent. In many cases, both nonviolent and violent campaigns exist simultaneously among competing groups. Often those who employ violence in mass movements are members of fringe groups who are acting independently, or in defiance of, the central leadership; or they are agents provocateurs used by the adversary to provoke the unarmed resistance to adopt violence (Zunes 1994). Alternative, often some groups use both nonviolent and violent methods of resistance over the course of their existence, as with the ANC in South Africa. Characterizing a campaign as violent or non-violent simplifies a complex constellation of resistance methods.

It is nevertheless possible to characterize a campaign as principally nonviolent based on the primacy of nonviolent resistance methods and the nature of the participation in that form of resistance.

Later in the chapter:

As one might expect, there are several good reasons why social scientists have avoided comparing the dynamics and outcomes of nonviolent and violent campaigns, including their relative effectiveness. First, the separation of campaigns into violent and nonviolent for analytical purposes is problematic. Few campaigns, historically, have been purely violent or nonviolent, and many resistance movements, particularly protracted ones, have had violent and nonviolent periods.

Classifying any given movement as strictly nonviolent would not have worked. The data is too messy for that.

Even the definition of success is tricky:

Success and failure are also complex outcomes, about which much has been written (Baldwin 2000). For our study, to be considered a “success” a campaign had to meet two conditions: the full achievement of its stated goals (regime change, antioccupation, or succession) within a year of the peak of activities and a discernible effect on the outcome, such that the outcome was a direct result of the campaign’s activities (Pape 1997). The second qualification is important because in some cases the desired outcome occurred mainly because of other conditions. The Greek resistance against the Nazi occupation, for example, is not coded as a full success even though the Nazis ultimately withdrew from Greece. Although effective in many respects, the Greek resistance alone cannot be credited with the ultimate outcome of the end of Nazi influence over Greek resistance alone cannot be credited with the ultimate outcome of the end of Nazi influence over Greece since the Nazi withdrawal was the result of the Allied victory rather than solely Greek resistance.

In fact, they classify basically all resistance (nonviolent or violent) to the Nazi regime as a failure. There are a few exceptions, such as the Rosenstrasse Protest. However, the Nazis fell predominantly due to the actions of the armies of other nation states, not the local resistance groups (though they certainly played a role in helping, such as funneling intelligence to the Allies).