Abstract: Intimate Partner Violence imposes severe socioeconomic and psychological costs yet remains pervasive worldwide. Are victim-blaming norms to blame? How can they be shifted, and how can policy and behavioral changes be achieved? We address these questions in three steps: (1) identify cross-country and individual-level determinants of victim-blaming attitudes, (2) conduct a survey experiment with 3,600 Turkish citizens, and (3) collaborate with the Women’s Directorate of Istanbul Municipality (IBB) for an RCT to shift harmful norms at large. The survey experiment employed both a within- and a between-subjects design. In the latter, we make victim-blaming norms salient only by incentivizing the elicitation of participants' second-order beliefs about others --but provide no other information; in the former, we provide them with misperception-correcting information regarding these norms. We find that correcting misperceptions led to less victim-blaming attitudes, reflecting a re-anchoring effect from overly pessimistic initial perceptions (65% overestimate), but did not affect policy and behavior. In contrast, priming the salience of norms led to significant policy and behavior changes (including higher donations for victim support) by participants who think themselves to be less victim-blaming (58% in our sample) compared to their perception of society on average. These, self-enhancing, relative moral comparisons acted as “behavioral subsidies” that license action. The implication is that, while social norm change --and stigma reduction-- are long-term processes, welfare-improving policy and behavior changes are feasible even in the short-run. These insights guide the planning of the RCT.