Productivity Change and the Effects of the Enhancement of the Mass Transportation Programme on the Bus Transit System in Taiwan

Abstract

The main objective of this article is to analyse the changes in productivity of the Taiwan Bus Transit System (TBTS) before and after the execution of the ‘Alternatives for Promoting the Development of the Public Transportation Sectors’ (APDPTS), which is designed to provide a better operating environment for the public transport sector in the provision of public transit services. We use a decomposition of the Malmquist productivity index to locate the sources of productivity growth, namely technical change and efficiency change. The former is further decomposed into an output bias, an input bias and a magnitude term to test neutrality. The latter is also decomposed into changes in pure technical efficiency and scale efficiency. In our case, between the pre‐ and post‐APDPTS periods, the efficiency increased only marginally. As a result of a five‐year enhancement programme, the technical regress slowed down slightly, and inward neutral shifts of a transformation frontier were interpreted as being the main contributor to technical regress. In particular, evidence of biased technical change was found, i.e., the effects of output capability increased slightly, but the efficient use of inputs declined during the post‐APDPTS period.

Notes

1. According to the Review Report of the ‘Alternatives for Promoting the Development of the Public Transportation Sectors (2002)’ published by Institute of Transportation, Ministry of Transportation and Communications, the tax/fee exemption has reduced the total cost of bus system by about 9.73% during the APDPTS period.

2. The average urban trip length did not change during this period, and thus the number of passengers was selected to represent the output of urban bus services.

3. Due to the unavailability of data for part‐time workers and hours worked by staff for each bus company, ‘full‐time workers’ was used as a proxy for ‘total labour’. It should be noted that some of the bus transit systems might have used part‐time workers in the provision of bus transit services. Ignoring all part‐time workers or treating them as full‐time workers grossly overestimates/underestimates the contribution of labour in the analysis. Hours worked is a better measure of the contribution of labour to productivity and efficiency than just the total number of employees.

4. To test if the served areas were changed by the selected samples, we used the ratio of urban bus vehicle‐kilometres to highway bus vehicle‐kilometres as the indicator. It, however, failed to indicate the existence of the different served areas among the selected samples using non‐parametric test (see the details in Appendix 1).

5. Although the 1997–2001 period was characterized by rising unemployment and declining stock values, trends in economic activity and unemployment did not affect staff recruitment during this period.

6. The mean rate of per vehicle‐kilometres of fuel consumption increases from 0.37 to 0.39 litre between the pre‐ and post‐APDPTS period.

7. All the frontiers are derived using observations from that year plus the previous year. In other words, we use the frontier constructed with 1992–1993 data as our 1993 frontier, the frontier constructed using the 1993–1994 data as our 1994 frontier, and so on, until our 2002 frontier is constructed. Each frontier contains 34 observations.

8. Since the two‐year‐window DEA method is used in this article, the performance indices start from 1993.

9. It should be noted that the most progressive efficiency change does not mean the greatest efficiency in terms of bus company’s operation, since a larger increase in efficiency change in a given bus company during one period might be due to the bus company performing inefficiently at the beginning of the period.

10. Bly and Oldfield (1986) found clear evidence that increases in operating loss subsidies raise unit costs, wages and employment while decreasing vehicle‐kilometres per employee and per vehicle, and may also have encouraged a loss of efficiency.

11. The subsidy responsibility was suggested to shift from the central government to more decentralized county and local sources since this would increase the pressure for cost control and potentially produce efficiency gains in transit systems (Shughart and Kimenyi, 1991; Pucher, 1995).