In Norway, as in many other countries, the public transport authorities' compensation to the operators for running non-profit routes is based on expected total operating costs (gross contracts) and expected revenues (net contracts). The subsidies are determined either through standardized cost-and income norms, negotiations or competitive tendering. As far as the Norwegian ferry sector is concerned, the subsidies given to each ferry operator, have been set through negotiations involving a high degree of trust between operators and transport authorities based on long-term relationships. In November 2002 it was disclosed by the Norwegian National Authority for Investigation and Prosecution of Economic and Environmental Crime (KOKRIM) that one of the biggest ferry companies in Norway, Ofoten and Vesteraalens Dampskibsselskap ASA (OVDS), had violated this trust and intentionally manipulated their accounting figures for over a decade. When the court passed judgement in August 2004, it was revealed that the company had reported too low revenues and too high costs and in this way swindled the Norwegian Ministry of Transport and Communication for about 113 million NOK (about 14 000 000 ) during the period 1993 to 2002. Clearly, the absence of use of standardized cost norms and/or incomplete such norms, do not enable the local transport authorities to check for and react on significant reported cost deviations. As will be demonstrated in this paper, a simple linear multiproduct cost model with two easily available output variables could have indicated the significant cost deviations caused by this fraud. Hence, the aim of this paper is to use the OVDS-case to demonstrate how the local transport authorities using cost norms can disclose significant deviations from expected operating costs so that subsidy fraud can be revealed at an early stage. In order to discuss the problem, accounting and output data for 55 crossings operated by 21 companies were analysed from 1995 to 2004. The data set thus includes observations of accounting and output figures both during and after the period of fraud. Using a multiproduct cost function, both panel data analysis and cross-sectional analysis can be applied. For all the years between 1995 and 2002, the dummy variable indicating crossings operated by this company was significant with a positive sign. The prejudiced hypothesis that this company, because of fraud (overcharging), operated more expensively than other operators, was thus confirmed at a high degree of certainty. This work provides a good example of how public authorities contracting out services can relatively easily reveal attempts to conduct subsidy fraud by putting together output and accounting information. Another significant positive effect of using standardized cost norms is that such norms have a general deterrent effect. When the transport operators perceive that the probability of being caught and punished for such frauds is high, they are less likely to commit them. For the covering abstract see ITRD E137145.
Samenvatting