Design and impact of a scheme to spread peak rail demand using pre-peak free fares


Graham Currie

Document Type

Conference Paper

Publication Date


Subject Area

place - australasia, mode - rail, operations - capacity, operations - crowding, planning - surveys, ridership - behaviour


rail overcrowding, Melbourne, Australia, peak demand, demand shifting


Rail overcrowding has become endemic for many major cities as urban populations grow and investment in capacity lags behind increased demand. Expanded rail capacity as a solution is problematic since substantial funding is required and can take many years to implement. Demand shifting by time of day presents an opportunity to reduce peak crowding and better match capacity with ridership. However the concept remains the subject of mainly theoretical research since little actual experience of demand shifting has occurred. This paper reviews the performance of a rail demand shifting program which has been operating in Melbourne, Australia since March 2008. The program uses free fares to rail travellers if trips are completed before 7:00a.m. and aims to shift demand from crowded peak trains to trains operating before the peak. The paper outlines previous research and experience in the field of demand shifting. It then outlines the rationale for the development of the Melbourne program and describes how the scheme was developed and operated. It presents the results of a series of user surveys which describe how passengers have changed travel in relation to the scheme notably those who have shifted demand from the peak and by how much. The short and medium term impacts on peak rail crowding are described and trends and influences on time shift behaviour analysed. The paper closes by reviewing the viability of the program in terms its financial and economic performance in reducing rail overcrowding and considers the transferability of the program to other cities. The program costs $Aust 6M p.a. in lost fare revenues and around 8-9,000 passengers use the ticket each weekday. Some 23% of these have shifted the time of travel (around 2,000 to 2,600 passengers) by an average of 42 mins. This has reduced demand in the peak by between 1.2% and 1.5% from previous levels and is the equivalent of a maximum of 5 average train loads (or 3% of total peak trains). Demand growth during this period has far outweighed this effect so overloading rose after early bird was introduced. Its affect was to reduce the scale of increased overloading. In the medium term take up of the ticket is increasing by 1.7% p.a. which is higher then ridership growth hence net share of free ticket takeup is rising. It is unclear if this acting to further reduce peak crowding since total ridership in the peak is still growing. Since the scheme has been in operation pre 7:00a.m loads have increased by 41% reducing pressures for purchase/operations of new peak trains saving 2.5-5.0 trains (2008) to a forecast of 8.05 trains in 2038. Financial analysis suggests the savings would substantively cover financial costs of free fares and benefits will increase over time. Wider economic benefits increase viability.


Permission to publish the abstract has been given by AET, copyright remains with them.