Simultaneously consideration subsidy allocation and target setting in low-volume offshore ferry routes: an empirical study of Taiwan companies

Abstract

To satisfy the need for resident transportation between offshore islands and a mainland, and among offshore islands, some ferry firms are asked to run low-volume routes. Since low-volume routes often bring losses to operators, the government may provide compensation to those offshore ferry routes. However, the current subsidy methods may result in operating inefficiency. Also, the previous studies lack a mechanism for auditing the performance of these ferry routes while considering subsidies. In order to solve these problems and consider the operational characteristics of ferry routes and the fairness of input and output multipliers, this study proposes a different performance-based mechanism built by a cross-efficiency two-stage network data envelopment analysis to allocate subsidies and set targets for individual offshore ferry routes. By setting targets, the government can supervise subsidy usage. An empirical example of seven offshore ferry routes in Taiwan is illustrated. The different results of allocation between stages can be found for all routes and the individual intermediate product and final output targets can be set for all routes. The results provide constructive suggestions for the policymaker on how to allocate subsidies and set targets among offshore ferry routes.

Keywords:

• Subsidy allocation
• target setting
• ferry
• route
• two-stage network data envelopment analysis

Notes

1 Although Bi et al.’s (2011) work in a network context, it only considers the activities in parallel and omits the activities in series.

2 The number of variables in this study does not fully comply with the "rule of thumb" reported by Dyson et al. (2001) but the rule of thumb is for the black-box DEA model. In addition, Banker et al. (1989) indicated that the main concern of the rule is the discrimination power. This study applies the cross-efficiency to improve the discrimination power (Sexton, et al. 1986). Hence, we can capably and sufficiently discriminate the whole observations and obtain a full rank among all observations.

3 Since the information of input and output parameter for each route is obtained from the MOTC, it is confidential and must be approved by the MOTC before it can be published in the study. However, it can be obtained from the MOTC upon request.

4 The other operating costs include port charges, ship and equipment depreciation, insurance cost, administrative cost, ship repair cost, terminal rent, capital investment, outsourcing cost, and other costs.

5 The inputs (fuel cost, personnel cost, and other operating costs) are used to generate the capacity (seat-nautical miles), and then the capacity are applied to provide the services (passenger-nautical miles).