Subsidies are considered important means to facilitate the provision of public transit, yet the empirical evidence implies that they can have harming effects on costs and possibly also on operators' performance. This paper examines the impacts of deficit-balancing subsidies on the cost inefficiency of local public bus companies in Germany, where a complex system allocates ample financial support. Our empirical analysis relies on a unique dataset of 33 companies observed over a period of up to twelve years for a total of 231 observations. We employ a stochastic frontier cost function for panel data that account for unobserved heterogeneity and provide firm-specific, time-varying inefficiency estimates. Further, we allow variations in the optimal technology by randomizing some cost functions' coefficients in one of our model specifications. Subsidies directly enter the inefficiency function as a heteroscedastic variable. We find a positive effect of subsidies on the standard deviation of inefficiency, which implies that the range of companies' inefficiency increases with the level of subsidies relative to total costs. However, we also find that non-subsidized firms perform better in terms of cost efficiency.
Stochastic cost frontier subsidies heteroscedasticity local public bus transportation cost efficiency panel data