Know Your Opponent and Know Yourself: The Political Economy of Relative Performance Evaluation

• In Event: Panel 4: Policy Learning

Abstract

Seeking to systematically construct a dynamic model for non-linear political incentive in the non-democratic context, this study examines how local politicians and bureaucrats adaptively adjust (AA) to the relative performance evaluation (RPE), which I conceptualize as relying on various reference standards to mitigate information cost and cognitive burden when evaluating performance. Utilizing a 6-year monthly panel dataset of the “Safety Index” in Zhejiang Province, China, the empirical findings reveal various dynamics of RPE-AA: (1) local politicians with Positve/Negative (higher/lower than the reference standard) RPE will “slack off”/“bounce back” in the future, in which (2) the Negative RPE tend to trigger stronger response than the Positive RPE ("negativity bias" effect), (3) the Social RPE (based on peer comparison) proves more influential than the Historical RPE (based on comparison to the past), (4) the combination of Historical-Social and Positive-Negative RPE generates more fuzzy and confusing information, and (5) the effect of RPE tends to play a continuous role but shrink rapidly after more than a month (“recency effect”). In combination with qualitative analysis based on complementary fieldwork, interviews, and cases, this article also identifies various mechanisms and alternative explanations that have been relatively unnoticed in academic debates. Based on all above, the ultimate goal of this study is to theorize the RPE as an underlying, relatively imperceptible but steadily self-renewable mechanism of precise performance information, non-linear career (dis)incentives, and ubiquitous political pressures for local governments on a daily basis, persistently triggering up-and-down AA and performance fluctuation.

Author

• Ze Cai, New York University