Institutional isomorphism, negativity bias and performance information use by politicians: A survey experiment

Abstract

New Public Management popularized performance measurement in public organizations. Underlying performance measurement's popularity is the assumption that it injects performance information (PI) into decision-making, thus rationalizing the ensuing decisions. Despite its popularity, performance measurement is criticized. In part, this criticism results from the limited knowledge of the conditions under which PI is purposefully used by politicians. We conducted a survey experiment based on real PI with 1,240 politicians. We hypothesized that PI has a positive impact on performance information use (PIU) when PI is benchmarked with coercive, mimetic or normative pressures. Moreover, due to negativity bias we expected this positive impact to be stronger when PI signals low performance. We found that normative pressures had a positive impact on actual PIU while coercive pressures positively affected intended PIU. Negativity bias is only relevant when linked to coercive pressures and intended PIU for analysing the organization's finances.