This analysis revolves around the question to what extent the use of various types of performance targets is conditional on the character of the tasks of the public institutions. The general prescription in most of the academic literature on performance management is that setting up performance targets should be focused on outcomes instead of output and activities. The reasoning is that outcomes are what the public institutions are supposed to achieve and therefore that this should be the focus of performance management. It is therefore surprising that the use of outcome targets in Danish state institutions is limited to 14 percent of the total number of performance targets despite that the official Danish governmental guides as well as all theoretical prescriptions recommends the use of outcome measures. However, a large theoretical literature also points out that many outcomes of public institutions are not easily measured and that performance management therefore occasionally needs to focus on measuring outputs and activities. This analysis focus on showing empirically why and when public institutions cannot use outcome measures and instead need to rely on output and activity measures well knowing that this is only the second best thing to do. The main argument is that the use of the different types of performance targets is dependent on the tasks of the public institutions. It is further argued following James Q. Wilson’s classical typology of production-, craft-, procedural- and coping institutions that the measurability of outcome and output of the tasks of the institutions determines the use of different types of performance targets. The analysis is made possible by a unique data set consisting of an encoding of the performance targets for 2014 for all Danish state institutions and for these institutions also a complete and weighted job description that enables concrete operationalization of the tasks of the institutions and consequent operationalization of Wilson's typology. Thus the specific research question is – to what extent is the use of various types of performance targets (outcome-, output- and activity targets) in Danish state institutions for 2014 determined by the tasks of these institutions (production, craft, procedural, coping)? The analysis shows that the use of performance goals to a large extend is dependent on the nature of the task of the institutions. Specifically, the analysis shows that the use of outcome goals to a large extend is dependent on craft institutions (with measurable outcomes). Furthermore the use of output goals are to some extend also dependent on procedural institutions (with measurable outputs). It is further argued that the conclusions of this analysis have a number of implications for performance management. First, that performance management in state institutions needs to a large extend to be based on output and activity measures because outcomes of most tasks of state institutions are not easily measured. Secondly, that the use of performance pay and other economic incentives should be used with great caution since the performance measures are not directly linked to the wished outcomes of the institution. Thirdly, this means that performance management in most public institutions cannot stand alone as a management tool, but must be seen in the context of the traditional weberian public management paradigm as well as within a broader trust based Public Value Management paradigm. Finally, it is argued that New Public Management (NPM) is not a coherent paradigm for public management and that much of the recent criticism of NPM therefore also is based on an undifferentiated conception of NPM and performance management. After 30 years, it is perhaps now time to put NPM on the shelf. Instead we need to split up NPM in the underlying paradigms. Only then can we have an informed discussion of NPM and performance management - and we risk not throwing workable performance management practices out with the bath water.
|Educations||Master of Public Governance, (Executive Master Programme) Final Thesis|
|Number of pages||43|