I thank Barbara Wejnert for her generous assessment of my book. This response concentrates on two crucial issues that her comments raise and that concern the political rationality of mass contention and the choice of research design.
One of my central claims is that waves of unorganized, leaderless crowd protests triggered by striking external precedents have low chances of success. Such emulative protests quickly spread across countries because discontented citizens rashly infer from the surprising downfall of a foreign ruler that their own incumbent can be toppled as well. My historical research attributes these facile inferences to cognitive shortcuts that people commonly apply, but that can cause substantial distortions in judgment. Due to these heuristics, citizens overestimate their chances of success and imitate the initial precedent, although the power constellation in their own polity is actually unpropitious. Therefore, these ill-considered challenges, which form an impressive diffusion wave, frequently fail.
This sobering message runs against the grain of recent scholarship on mass contention, which holds more sanguine views—and hopes. But my bounded-rationality theory does not depict the protesting crowds as “irrational masses,” as Wejnert claims. Making Waves avoids this term; it does not seek to revive crowd psychology à la Gustave Le Bon. Rather than acting out emotionally, protesters believe that contention has good chances of success. They intend to proceed rationally, but their beliefs are distorted by inferential heuristics. These misjudgments came to the fore again during the Arab Spring with its meager outcomes, which provided an “out of sample” confirmation for my theory, derived from the 1848 revolutions. Peaceful protests may generally be more successful than violence, as Erica Chenoweth and Maria Stephan’s study cited by Wejnert argues. But the subset of nonviolent contention that is unorganized and that arises from precipitous emulation efforts is unlikely to achieve success, as the analysis of this kind of diffusion waves in my concluding chapter shows.
By studying protesters’ perceptions and choices in depth, my book seeks to complement the “multivariate analyses” that Wejnert advocates and that she conducts in her own book. A wealth of statistical studies have proven the importance of diffusion and have established its parameters, conditions, and pathways, such as the cultural commonalities, historical linkages, media connections, and personal networks that Wejnert’s review highlights. But many scholars, including statistical researchers, now call for unearthing the causal mechanisms that drive these diffusion processes. This is my book’s main goal: It reconstructs actors’ decision making through controlled case studies, such as a comparison of emulative contention in Germany during 1848 versus 1917–19.
Finding quantitative indicators for the subtle mechanisms of inferential heuristics is practically impossible, especially for large numbers of cases. The sparse research on the unfolding of the 1848 revolutions in many settings, for instance, hinders statistical tests of my theory. While quantitative investigations are good at disentangling the structural correlates that Prof. Wejnert highlights, there is also an important place for qualitative studies, especially the historical examination of causal mechanisms that my book conducts.
