Behavioral Public Performance

Abstract

A revolution in the measurement and reporting of government performance through the use of published metrics, rankings and reports has swept the globe at all levels of government. Performance metrics now inform important decisions by politicians, public managers and citizens. However, this performance movement has neglected a second revolution in behavioral science that has revealed cognitive limitations and biases in people's identification, perception, understanding and use of information. This Element introduces a new approach - behavioral public performance - that connects these two revolutions. Drawing especially on evidence from experiments, this approach examines the influence of characteristics of numbers, subtle framing of information, choice of benchmarks or comparisons, human motivation and information sources. These factors combine with the characteristics of information users and the political context to shape perceptions, judgment and decisions. Behavioral public performance suggests lessons to improve design and use of performance metrics in public management and democratic accountability.

Valuation tools and politicians’ willingness to sell public real estate: a survey experiment

“…First, this study adds to scholarly work on Public Value Governance, which explicitly calls for more empirical research into how diverse government actors deal with the assessment of (public) values for given policy issues (e.g., Bozeman 2019;Bryson, Crosby, and Bloomberg 2014;Fukumoto and Bozeman 2019;Hartley et al 2017;Nabatchi 2012;Huijbregts, George, and Bekkers 2021). Moreover, this study adds to scholarly work on Behavioural Public Administration, which explicitly calls for more empirical research into how political decision-making is impacted by different strategy and performance tools (e.g., George 2020; James et al 2020). Second, this study employs a vignette-based, randomized survey experiment among actual local politicians in Flanders, Belgium, to infer causality concerning the impact of valuation tools on political decision-making.…”

“…In addition, by means of these tools, elected officials can legitimize their decision-making behaviour (Carpenter and Krause 2012;Coicaud 2002). Importantly, these assumptions concerning the rationality ensuing from using valuation tools are strongly contested on the basis of insights from behavioural science (George 2020;James et al 2020), which indicate the existence of cognitive biases among politicians (e.g., Battaglio et al 2019;. Despite these conflicting perspectives on the impact of valuation tools, little causal evidence can be found on the decision-making consequences of valuation tools among elected officials.…”

Getting a Grip on the Performance of Collaborations: Examining Collaborative Performance Regimes and Collaborative Performance Summits

“…However, recent performance management research has found that effective performance management requires appropriate performance routines as well (Behn 2010 ; Gerrish 2016 ; James et al 2020 ; Kroll and Moynihan 2018). Performance routines such as goal‐setting processes, performance budget negotiations, and performance reviews “structure how [actors] experience their work” (March and Simon 1993 ), help actors to make sense of ambiguity (Noordegraaf 2017), and motivate them to improve performance (Moynihan and Kroll 2016 ).…”

Does Managerial Use of Performance Information Matter to Organizational Outcomes?

“…To better understand the effects of ubiquitous public sector performance systems, Moynihan and Pandey (2010) proposed treating performance information (PI) use as a “big question” for researchers to address. Over the past decade, researchers have amassed an impressive body of evidence (James et al, 2020). The next logical question is whether performance information use makes a difference to organizational performance.…”

“…However, PI use may not improve organizational performance for a number of reasons, such as goal ambiguity (Chun & Rainey, 2005; Moynihan, 2008), political constraints on managerial autonomy (O’Toole & Meier, 2011), difficulty in inferring what the data actually suggests about concrete actions (Moynihan, 2008), and cognitive biases (James et al, 2020). Hence, using PI has been found to be associated with null average effects at public organizations (Gerrish & Spreen, 2017; Hvidman & Andersen, 2014).…”