Severe state repression of civilians can backfire on a given regime and spur mass increases in mobilisation. In a recent surge of interest in the nonviolent resistance literature, scholars have suggested that if present in a given nonviolent campaign, pre-existing movement infrastructure and corresponding media institutions increase the likelihood of repression backfiring. As such, these previous studies have not answered why repression actually backfires. We still know little about how harsh repression influences political identity. At the same time, we know little of the mechanisms underlying backfire and are uncertain as to how it operates in different historical and communicational eras. Through comparative analysis and use of process-tracing methods, I analyse twelve empirical cases of harsh state repression inflicted on civilian protest movements across a wide range of historical eras. Results offer new insight to why some incidents of severe repression have a higher likelihood of backfiring than others.