Measuring and explaining the production efficiency of Spanish universities using a non-parametric approach and a bootstrapped-truncated regression

The identification of environmental factors that explain differences in efficiency is essential for improving the results of public universities. A two-stage, semi-parametric approach with the single and double bootstrap procedure (Algorithm #1 and Algorithm #2) proposed by Simar and Wilson (J Econo...

Full description

Saved in:
Bibliographic Details
Published in:Scientometrics 2020-02, Vol.122 (2), p.825-846
Main Author: Salas-Velasco, Manuel
Format: Article
Language:English
Subjects:
Algorithms
Computer Science
Data envelopment analysis
Economic models
Efficiency
Environmental factors
Environmental impact
Inference
Information Storage and Retrieval
Library Science
Nonparametric statistics
Operations research
Regression analysis
Statistical analysis
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The identification of environmental factors that explain differences in efficiency is essential for improving the results of public universities. A two-stage, semi-parametric approach with the single and double bootstrap procedure (Algorithm #1 and Algorithm #2) proposed by Simar and Wilson (J Econom 136(1):31–64, 2007) was used in this article for making valid inferences about the impact of environmental factors on university efficiency. A data envelopment analysis (DEA) efficiency estimator was used in the first stage to estimate technical efficiency scores for Spanish public universities. It is common to explore the determinants of (in)efficiency in a second stage. To provide valid inference, Simar and Wilson (2007) suggested a parametric bootstrap of the truncated regression (Algorithm #1). Alternatively, they recommended a bootstrap procedure to obtain bias-corrected technical efficiency scores used in the second-stage truncated regression; valid inference can be obtained by using a second bootstrap procedure applied to the truncated regression (Algorithm #2). Under both algorithms, three environmental factors were statistically significant predictors of efficiency. Our results confirmed that universities with a higher percentage of academics with tenure, outgoing Erasmus students, and state grantees tend to be less inefficient.
ISSN:0138-9130
1588-2861
DOI:10.1007/s11192-019-03324-4