In January 2016, Külli Taro, a PhD student of the Ragnar Nurkse School of Innovation and Governance of Tallinn University of Technology, defended her doctoral thesis "The Attribution Problem in Performance Measurement in the Public Sector: Lessons from Performance Audits in Estonia".
Although collecting information on performance in the public sector is as old as public administration itself, performance measurement became the focal point in connection with the New Public Management reforms and the promotion of performance management in the public sector. Today, however, the debate is focused in particular on questions like "why are the results measured?" and "whether and how is performance information used?". It has appeared that efforts invested in the performance measurement process considerably surpass the actual usage of performance information.
"One important reason, why performance information is not used to the extent expected, stems from the inherent problem of performance measurement - the failure to clearly and unambiguously define the goals, i.e. outcomes, of the action of the public sector, the achievement of which should be measured. Another problem in performance measurement is related to analytical and methodological challenges in proving causality between the government activities and the measured outcomes. Even if appropriate indicators for evaluating the performance of the public sector can be found and measured, it is often very difficult to determine what contribution the specific action or programme has made to the outcome," Külli Taro explains.
The attribution problem denotes the tendency of humans known from social psychology to attribute results to measured activities, although in reality there may be no causal relationship between them. The main focus of the thesis is on the consequences of the causality attribution problem upon utilization of performance measurement results and the thesis demonstrates that the impact of the attribution problem depends on the purpose that the performance measurement is intended to fulfil and the environment where the measurement takes place.
The doctoral thesis has two goals:
- firstly, to demonstrate how the attribution problem hinders in particular the advisory purpose of performance measurement, since in this case the existence of a causal relationship between the action and the outcome is most critical.
- secondly, to examine application of performance measurement initiatives in the Estonian context and to explain why the attribution problem is amplified in new democracies with less-developed policy-making skills and traditions.
Successful performance measurement and adequate use of performance information set certain requirements to the country's political system and administrative capacity. If performance measurement is applied in a public sector, where the pre-conditions in terms of analytical and operational skills, quality control and stable policy-making have not been met, the intrinsic and analytical shortcomings of performance measurement, such as lack of unambiguous and clearly defined purposes and the causality attribution problem, will be amplified.
"Thus, the risk of making fundamental methodological errors upon performance measurement due to the attribution problem and misuse of the information received is particularly high in new democracies with immature policy-making environments," the PhD student adds.
The supervisor of the doctoral thesis was Professor Tiina Randma-Liiv (TUT).
The opponents were Jarmo Vakkuri (University of Tampere) and Assistant Professor Bram Verschuere (Ghent University in the Netherlands).
The doctoral thesis has been published in the in the digital collection of the library at http://digi.lib.ttu.ee/i/?3963
Kersti Vähi, Research Administration Office