A Sampling Method to Estimate the Revenue of Private Bus Service (A Case Study for a Selected Urban Cross Town Route in Colombo District)

Abstract

Public transportation is one of the important service sectors in Sri Lanka. Buses are the principle mode of public transportation. But the quality of the bus service is very low due to many reasons. One major problem is that there are no proper and quality fare policies which give similar benefits for both passengers and the bus owners. Different fare structures have been introduced by the government time to time. The policy introduced in 2002 by National Transport Commission (NTC) is used to calculate the fares at present. It takes into account 12 factors which affect the expenditure of buses (operational costs) and 10 route types. But the revenue earned from a bus is not considered when preparing fare policies due to the difficulties of collecting revenue data as the owners are reluctant to give actual data. NTC has done some surveys to collect the data to estimate the revenue. But their method has many weaknesses. The main objective of this study is to propose a new sampling procedure to collect the data from passengers to estimate the revenue of a private bus. In the new sampling procedure, data are collected from the passengers that arrive to the halts. In the survey done by NTC, the data were collected from the passengers by traveling in buses. New sampling procedure provides estimates not only for the revenue but also for the total number of passengers in the route, number of passengers travel in each route number, etc. In the sampling procedure, a questionnaire is used to collect data from passengers about the number of sections they travel, and a schedule is used to record the number of busses that arrive to the halt and the number of passengers that get in to those buses. The revenue of a bus per turn is estimated by considering the starting time of the bus, the number of passengers that loads in to the buses and the number of buses reaches the halts. A comparison was done by using the estimated revenue of the simulation and the calculated revenue of the collected data. The revenues give only 10% of relative error. The validity of the proposed procedure was checked by comparing our estimate with the data collected from the bus owners. The private buses of the route 120 (Kesbewa, Pettah) in Colombo district were taken as the target population of the study due to limited time period and the resources. A Matlab program was developed to estimate the revenue. Inputs of the program are the starting time of the bus, the matrix of the number of sections that passengers travel and the average number of passengers get in to the buses from each halt. For the validation of the program, the inputs were estimated from the data collected from the ticket machines.