We study the long-run effects of conflict on social attitudes, with World War II in Central and Eastern Europe as our setting. Much of earlier work has relied on self- reported measures of victimization, which are prone to endogenous misreporting. With our own survey-based measure, we replicate established findings linking victimization to political participation, civic engagement, optimism, and trust. Those findings are reversed, however, when tested instead with an objective measure of victimization based on historical reference material. Thus, we urge caution when interpreting survey- based results from this literature as causal.