The decentralized nature of American policing largely allows agencies to define crime, and the requisite severity of that crime, at the institutional level. As a result, local hate crime statistics are differentially constructed across space. The current research posits that official hate crime statistics most accurately reflect a rate-producing, rather than a behavior-producing, process. Focusing on anti-Muslim hate crime, a particularly understudied form of hate crime, we examine reporting differences across community factors, using the media reports as a comparison point. We analyze official FBI hate crime reports, and news reports collected by Mapping Islamophobia, employing a logistic regression to predict anti-Muslim hate crime reporting. Our findings suggest significant variation in the social construction of anti-Muslim hate crime, both across institutions and across space.