This paper addresses the gap between normative expectations of the right to protest in liberal democracies and the continued practice of repressive protest policing. Empirical literature has identified three types of factors explaining repressive policing: macro, societal-level factors, meso-level factors relating to the police organization, and micro-level factors pertaining to specific events. Yet these factors provide only a fragmentary understanding of the phenomenon. In this paper we put forward a novel three-tiered methodology of scaled reading, which is able to examine all these explanations together. We used scaled reading to analyze the protocols of the Or Commission of Inquiry, which investigated lethal clashes between Israeli police and the Arab minority in October 2000.

Through large-scale algorithmic topic modeling, we found that all types of explanations of repressive policing co-exist within the October events. The mid-scale analysis revealed that no type of explanation exclusively belongs to a specific group of actors. The small-scale reading of most representative documents demonstrated that this co-existence of mechanisms is also present within single testimonies. Together, our findings challenge existing empirical categories and illuminate repressive policing as a non-linear, non-binary, non-causal, and non-unitary phenomenon. These insights help make sense of the phenomenon’s persistence in deeply-divided societies.