Examining the Context of Attitudes Toward Immigrants : A Reanalysis of the Threat/Contact Hypotheses
A survey of prior research testing the threat and contact hypotheses uncovers a wide, often conflicting, range of conclusions regarding the influence of contexts on attitudes. Through a reanalysis (verification and replication) of publicly available data we test the threat/contact hypotheses by examining views on immigration. Using different measures (self-reported contact vs. contextual), models, estimation techniques (logit vs. HLM) and samples (local, state, and national) the reanalysis provides insights on why we observe such wide variation in results in the literature. The results show that the findings are not only sensitive to model choice and specification, but are not generalizable across regions. The study suggests that contextual effects cannot be understood without a better understanding of the larger environments where these contexts are operating, confirming what Huckfeldt (1986) argued decades earlier, 'political opinions and behavior of individuals cannot be understood separate from the environment within which they occur.'