Interpreting performance information: Motivated reasoning or unbiased comprehension?

Two Stata datasets and do-files for \"Interpreting Performance Information: Motivated Reasoning or Unbiased Comprehension? A Replication and Extension\" Abstract Despite a growing number of studies on how prior beliefs distort citizens' interpretation of performance information for service providers, little is known about whether prior beliefs matter equally across different services and types of providers. In this study, we provide a wide replication and extension of the experimental design used in Baekgaard and Serritzlew (2016) with three types of providers (public, non-profit, and for-profit) across two services (nursing homes and refuse collection). Based on two large-N nationally representative experiments (N = 3018 and N = 3020), we find that citizens' sector preference does indeed impact their interpretation of performance information, corresponding to the original study. However, public sector preference plays a substantially different role in the two services. Our findings strengthen the external validity of previous research and simultaneously identify theoretical boundaries to its application across various services and providers. This, we argue, underlines the importance of replicating and extending pivotal studies on performance information.