| Article ID: | iaor20052707 |
| Country: | United Kingdom |
| Volume: | 40 |
| Issue: | 5 |
| Start Page Number: | 401 |
| End Page Number: | 416 |
| Publication Date: | Sep 2004 |
| Journal: | Transportation Research. Part E, Logistics and Transportation Review |
| Authors: | Boame Attah K. |
| Keywords: | performance, statistics: data envelopment analysis |
This paper uses a bootstrap data envelopment analysis (DEA) method to estimate technical efficiency scores for Canadian urban transit systems from 1990 to 1998. Bias and confidence intervals are estimated for the efficiency scores. A Tobit regression is used to analyze the sources of efficiency change. The paper finds that the original efficiency scores are biased. The bootstrap results indicate that the average technical efficiency of the transit systems is about 78%. Most transit systems experience increasing returns to scale. While higher average transit speed increases efficiency, peaking decreases efficiency, which reinforces previous DEA efficiency studies.