Assessing total factor productivity growth of the passenger-bus transit systems in Indian metropolitan cities using the sequential Malmquist-Luenberger productivity index approach

ABSTRACT

This article examines the patterns of total factor productivity (TFP) growth in eight state-owned passenger bus companies operating in major metropolitan cities of India. A state-of-the-art and innovative sequential Malmquist-Luenberger productivity index (SMLPI) approach is applied to compute unbiased estimates of TFP growth and its decomposed components. The empirical results based on the panel data set for the period 2011–2016 suggest that technical inefficiency is a common characteristic of bus companies providing passenger services in metropolitan cities. The analysis of TFP growth at the aggregate level reveals that the public bus transit system placed in Indian metropolitan cities experiences mild productivity regress, which is entirely attributable to efficiency loss. The disaggregated analysis provides that only four bus companies experience productivity gains throughout the sample period. Pune Mahamandal appears to be the most innovative bus company.

KEYWORDS:

• Accidents
• total factor productivity
• sequential Malmquist-Luenberger productivity index
• India
• metropolitan bus operators

JEL CLASSIFICATION:

• C61
• L91
• R41

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ Chung, Färe, and Grosskopf (1997) argue that, ‘in order to obtain the true productivity scores under MLPI it is necessary to follow constant return-to-scale (CRS) assumption’. Zelenyuk and Zheka (2006) point that DEA efficiency has greater discriminative power under CRS relative to other forms of returns-to-scale. In light of above, we assume the CRS while computing SMLPI and its components for sample bus companies.

² We prefer undesirable output to be weakly disposable in the SDDF model because in certain situations the bus company has to undertake cost in terms of desirable output (i.e. passenger-kilometres) to contract the level of undesirable output (i.e., number of accidents).