This dissertation studies the implications of a risky operating environment of firms for productivity analysis. The discussion is done under a new methodological framework and the conceptual gap between the models of production risk and the models of inefficiency is bridged. The thesis includes an introductory part and five research articles. Article 1 introduces the new stochastic semi-nonparametric envelopment of data (StoNED) estimation method that has been used in the thesis. Article 2 applies the StoNED method in the context of Finnish electricity distribution regulation. In 2010, the Finnish regulatory agency adopted StoNED to assess the cost efficiency of distribution companies. We compare StoNED to the previously used methods and it outperforms them because it is more flexible in taking into account the firm characteristics and its environment. Article 3 is a literature review which explores the links between production risk and inefficiency literature. It is found that the econometric concept of heteroscedasticity connects these two veins of literature. However, combining risk and the models of inefficiency may be complicated as it is challenging to distinguish the effects of inefficiency from the effects of the risky environment. In article 4, we examine the relationship between macro-productivity and corruption. Conventionally corruption is seen either as an impediment to or, under certain conditions, as a catalyst of economic growth. We, however, consider that corruption acts as a macro risk factor. That is, corruption increases the likelihood of productivity being low, but it allows the possibility of good performance also. Our analysis shows that the dispersion of productivity is larger among countries with high corruption levels. Article 5 studies the quality of service of Finnish electricity distribution firms in terms of costs of interruptions. We study how the underground cabling affects these costs and how the quality targets should be set. As expected, underground cabling decreases the level of costs. However, underground cabling does not significantly decrease the variation in these costs. We observe rather large variation of interruption costs even at relatively high degrees of underground cabling because the interruptions are costly in the underground networks mainly located in big cities. We also suggest that the quality targets should be set by using a StoNED based on a best-practice quality frontier rather than by an average of past performances of firms. The frontier is better for incentivizing firms to improve their operations and it produces more balanced targets than the average.
|Translated title of the contribution||Heterogeenisen tuotantoympäristön ja tuotantoriskin mallintaminen modernissa tuottavuusanalyysissa|
|Publication status||Published - 2014|
|MoE publication type||G5 Doctoral dissertation (article)|
- operating environment