This is the abstract of Fare Elasticity and Its Application to Forecasting Transit
Demand . Click on Publications
for the APTA publications order form to order the complete report.
Transit managers are under increasing pressure to obtain sufficient fare revenues to
maintain superior service while reducing dependence on government assistance. They need an
accurate formula to estimate the impacts of fare changes on transit ridership and fare
revenues. For years, these managers were given two choices: constructing a fare elasticity
model specific to their transit systems or applying the Simpson-Curtin formula
which postulates a fare elasticity of -0.33; i.e., a 10 percent increase in fare would
result in a 3.3 percent decrease in transit patronage.
The models are usually costly and time-consuming to construct, causing delays in the
implementation of fare changes. On the other hand, the 30-year-old Simpson-Curtin formula
is likely to be inaccurate today. Further, it provides no estimation of the varying fare
impacts between peak and off-peak hours, or between large and small cities.
The objectives of this study are to verify the Simpson-Curtin formula using updated
data and modern technologies, and to provide a set of fare elasticity estimates for bus
service in various cities during peak as well as off-peak hours.
An advanced econometric model, the Autoregressive Integrated Moving Average (ARIMA)
model, was used for the estimations. A special survey was conducted to obtain ridership
data 24 months before and 24 months after each fare change for 52 transit systems. Monthly
information on other factors which may influence ridership, including gasoline price,
vehicle miles of service, labor strikes, etc., were also collected. The purpose was to use
the model to isolate the impacts of the fare changes from those caused by other factors.
Findings
On the average, a ten percent increase in bus fares would result in a four percent
decrease in ridership. This shows that today's transit users react more severely to
fare changes than found by Simpson and Curtin.
Fare Elasticity-Bus Services
| Average (all hours all cities) |
-0.40 |
Transit riders in small cities are more responsive to fare increases than those in large
cities. The fare elasticity for bus service is -0.36 for systems in urbanized areas of
1 million or more population. In urbanized areas with less than 1 million people, the
elasticity is -0.43.
Although the data for peak vs. off-peak services are available for only six transit
systems, the difference between the fare elasticity levels is very clear: The average
peak-hour elasticity is -0.23 while the off-peak hour elasticity is -0.42, indicating
that peak-hour commuters are much less responsive to fare changes than transit passengers
travelling during off-peak hours.
