An extensive literature on counterinsurgency outlines a variety of measures that target the insurgency’s local support base. Civilians exposed to counterinsurgency repression should change their preferences accordingly by siding with the insurgency (Kocher et al. 2011). We contend that it is not obvious that this is the case. We outline a simple framework that isolates the effect of a heretofore unknown counterinsurgency strategy that systematically represses civilians. We look at roughly 40,000 villages in Turkey with panel data in which names of thousands of villages were changed in the 1957-1968 wave of geographical renaming. Using a two-step estimation procedure, we first show that Kurdish villages close to shared borders with Iraq and Syria carried the burden of the cultural makeover. Despite the widespread implementation of the Turkification policy, name-changing had no effect on levels of insurgent violence at the village level. As villagers continued to use old village names, cultural policies failed to go beyond administrative experiments. These findings present a contrast to the literature by showing that repression neither backfires nor facilitates violence when it fails to redistribute civilian preferences.