MANAGERIAL USES OF CAUSAL MODELS OF SUBWAY ON-TIME PERFORMANCE

On-time performance (OTP) indicators reflect the performance of employees and the effectiveness of policies and organizational structure. However in New York City's subway system the operating environment includes factors beyond the control of operating employees, such as mechanical reliability of subway cars, route merges, scheduled headways, construction projects, crowding, passenger behavior and so on. These factors differ from route to route and complicate comparisons of routes. Planning decisions affecting ridership levels, route design, and the operating environment can benefit from precise estimates of impacts on OTP, especially to help in evaluating the trade-offs involved. The results of a statistical causal model of OTP are presented. A set of hypothetically important variables was developed from New York City Transit Authority documents and train movement records. Data for over 54,000 morning rush hour trains traveling from terminals to central business district stations during 1988 and 1990 were used. The model quantifies the effect of variables on the probability that a train will be on time. How the results of such a model can be used to make predictions of performance that control for the operating environment are also shown, allowing performance comparisons between routes with different characteristics. How different OTP goals can be set for different routes is suggested. Finally by converting the results to odds ratios, it is shown how small improvements on routes with OTP of more than 90% can provide large benefits from the perspective of riders and how OTP measures obscure that fact.

Language

  • English

Media Info

  • Features: Figures; References; Tables;
  • Pagination: p. 57-64
  • Monograph Title: Public transportation 1994: current research in planning and development; management, marketing, and fare policy; and technology
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 00675412
  • Record Type: Publication
  • ISBN: 0309060591
  • Files: TRIS, TRB
  • Created Date: Mar 29 1995 12:00AM