Transit
Demand-Responsive Transit Costs Most Where It Is Needed
Booked minibus services promise flexibility in areas too thin for fixed routes, but the cost per passenger rises exactly where population is sparse and trips do not combine.
Flexible bus services are proposed wherever a fixed route is uneconomic. The economics that made the route uneconomic apply to the flexible service as well.
The saving depends on shared rides
A demand-responsive service is efficient when several passengers travelling at similar times in similar directions can be combined into one vehicle journey.
Combining requires enough requests within a window to find compatible trips, which depends on the density of demand rather than on the software.
Where requests are sparse, the vehicle carries one or two people and the cost per passenger approaches that of a taxi with a professional driver.
Detours penalise the passengers already aboard
Picking up an additional passenger lengthens the journey for everyone else in the vehicle, and the algorithm must decide how much delay to impose.
Tight limits on detour preserve journey quality and reduce the number of shared rides, which raises cost.
Loose limits improve utilisation and produce journeys that take considerably longer than driving, which reduces demand and returns the service to low occupancy.
Booking is a barrier for part of the market
A service that must be requested in advance excludes spontaneous travel, which is a large share of the trips a fixed route carries.
App-based booking works well for some users and creates a genuine obstacle for others, so telephone booking has to be maintained alongside it.
The staff and systems for that are a fixed cost spread over a small number of passengers, which again pushes up the unit figure.
The comparison is usually made against the wrong baseline
Flexible services are often compared with the fixed route they replaced at the moment of replacement, when the route was running nearly empty.
A fairer comparison considers what the route would have carried with better frequency and reliability, since low ridership on an infrequent service is partly self-inflicted.
Where flexible services have been evaluated over several years, the pattern is that they serve real needs at a cost per trip that few authorities can sustain at scale.
The useful niche is narrower than the pitch
These services work well as feeders to a frequent trunk route, collecting passengers from a dispersed area and delivering them to an interchange.
They also work for defined groups with specific needs, where the alternative is a specialist vehicle and the cost comparison is different.
Presented as a general replacement for conventional buses, they tend to disappoint, because the underlying problem was population density rather than route design.
Questions readers ask
Do bus lanes make congestion worse for other traffic?
They reduce general traffic capacity on the link, and effects vary with how much traffic shifts to the bus. Corridors have generally adapted more than fixed-demand models predicted.
What is conditional signal priority?
Priority granted only when a bus is behind schedule or above a certain occupancy, rather than to every bus. It targets the delay that matters while limiting the impact on other movements.
