How Do Politicians Attribute Bureaucratic Responsibility for Performance? Negativity Bias and Interest Group Advocacy

Never Change a Winning Policy? Public Sector Performance and Politicians' Preferences for Reforms

“…Finally, proposals for school closures trigger intense media disputes and opposition from local communities during the preparation of the municipal budget . Exploiting insights from recent work on negativity bias in citizens' and politicians' perceptions of performance (Nielsen and Moynihan ; Olsen ), we maintain that performance information is particularly likely to affect politicians' preferences for reforms involving greater political risks (such as the risk of losing key supporters). The reason is that politically costly reforms require a stronger defense from politicians, and performance information can help provide the necessary justification (Asatryan, Heinemann, and Pitlik ; Johnsen ).…”

“…For ease of comparison across municipalities, we calculate municipality‐level percentage deviations from the national average. Similar measures of relative performance have been used in comparable contexts by, for instance, Charbonneau and Van Ryzin (), Nielsen and Baekgaard (), Barrows et al (), Olsen (), and Nielsen and Moynihan (). These percentage deviations are measured as MD k = [( M k – 50) * 100]/50, ED k = [( E k – 50) * 100]/50, and RD k = [( R k – 50) * 100]/50 (where M k , E k , and D k reflect the average scores for the three test subjects across all pupils in municipality k ).…”

“…Yet we are not interested in the good or bad quality of any particular school. Rather, in line with publicly available performance information, our analysis—and thus also our question phrasing—focuses on school performance at the municipality level (see also Nielsen and Baekgaard ; Nielsen and Moynihan , ).…”

“…The reason lies in the fact that negative information tends to have a stronger impact compared with positive information of equivalent magnitude (Tversky and Kahneman ). Recent (experimental) work on citizens' and politicians' responses to performance information likewise shows that especially information about negative performance induces stronger causal attribution of responsibility (Nielsen and Moynihan ; Olsen ). Decision makers are blamed for bad outcomes or unpopular decisions, but they do not receive equal credit for good outcomes or popular decisions (James et al ).…”

“…Clearly, the mere existence of performance data need not imply that politicians also take this into consideration. Politicians should be informed about the level of public sector performance—and believe this information to be relevant and credible (James and Van Ryzin ; Nielsen and Moynihan ; Van Ryzin and Lavena 2013). Performance information is generally provided through some form of communication, which “surrounds every policy implemented by bureaucrats” (Andersen and Jakobsen , 57).…”