Aaltodoc publication archive (Aalto University institutional repository)
School of Business | Department of Economics | Economics | 2014
Thesis number: 13691
Estimating technical efficiency in Finnish industry: A stochastic frontier approach
|Title:||Estimating technical efficiency in Finnish industry: A stochastic frontier approach|
|Year:||2014 Language: eng|
|Department:||Department of Economics|
|Index terms:||kansantaloustiede; economics; kansantalous; national economy; teollisuus; industry; teknologia; technology; suorituskyky; efficiency; tehokkuus; effectiveness; ekonometria; econometrics; Suomi; Finland|
» hse_ethesis_13691.pdf size:2 MB (1733381)
|Key terms:||technical efficiency; industry; econometrics; stochastic frontier analysis; stochastic nonparametric envelopment of data; SFA; StoNED|
The objective of this thesis is to estimate the technical efficiency of Finnish industry in various sectors using stochastic frontier models. Furthermore this thesis provides a comparison between parametric and nonparametric estimation techniques utilized in the efficiency analysis literature. We estimate the technical efficiency for each sector separately, both as a cross-section and using panel data techniques. We discuss the various extensions to the models considered and identify further topics of research based on our estimation results.
METHODOLOGY AND DATA:
This thesis utilizes stochastic frontier estimation techniques to estimate technical efficiency for various sectors of industry. Production frontiers are estimated both parametrically and nonparametrically. The two main approaches to frontier estimation utilized in the thesis are Stochastic Frontier Analysis (SFA) and Stochastic Nonparametric Envelopment of Data (StoNED). The dataset we utilize is a cross-sectional panel dataset of companies operating in various sectors of the Finnish industry. Technical efficiencies are estimated with both cross-sectional and panel data methods, with various specifications.
We find that the mean efficiency in most sectors is quite high in most models. The estimated efficiencies are sensitive to the model choice. Parametric and nonparametric estimates are found to be similar and of the same magnitudes for the majority of sectors considered. We find significant variability in technical efficiencies over time, particularly in models where firm-specific heterogeneity is controlled.
Electronic publications are subject to copyright. The publications can be read freely and printed for personal use. Use for commercial purposes is forbidden.