Managing Public Performance Through Budgetary Incentives: Appropriate Regardless the Consequences?

Abstract The current budgetary crises can be seen as a test case for New Public Management (NPM) reforms that have been adopted worldwide during the last two decades. One of these tools, performance based budgeting (PBB), is the subject of this paper. Although the roots of Performance Based Budgeting can be traced back as far as the early 20thcentury, PBB gained worldwide popularity in the 1990s and early 2000s as part of the NPM agenda. NPM introduced a set of recipes that were meant to transform the public sector towards more result orientation and efficiency. It is not clear yet what the post NPM era will look like exactly, but it looks like NPM’s heyday is well behind us. In the meantime, many traces of these reforms still dominate today’s public sector landscape. It can easily be argued that PBB has become and remained so popular more because of the promises that it holds than the results that can empirically be attributed to its introduction. Regarding PBB, Allen Schick once noted that: governments that don’t manage for results will not budget for results *Schick 2003+. Expanding on Schick’s observation, the very type of result orientation that PBB was intended to achieve may turn out to be its main unarticulated premise. In other words, did PBB indeed modify public organizations and their steering relationships? Or did it merely codify existing behavior in those cases that report success? This paper presents a theoretical framework and method to assess this question. Lending from neo-institutional, more precisely principal-agent theory, alternative explanations for the use of performance information will be tested in international cases that share successful PBB implementation. The incentives PBB creates strongly rely on the logic of consequence. The alternative explanations sought follow the logic of appropriateness by focusing on the concepts such as of path dependency, cultural appropriateness and cognitive frames.