Fixed vs Dynamic Public Transit

Recently, some DART board and city council members have said that dynamic transit systems like Uber and Lyft are the future. They claim that fixed-route transit systems, like buses and trains, are no longer the most effective way to move people around. Below is a quote from Plano City Council Member Shelby Williams, who wrote about this idea:

“What I've advocated for quite a while is a dynamically routed bus system using large buses where appropriate, and smaller buses otherwise, where riders use their phones to indicate where and when they need to be picked up (including here and now), and where they're going. This is the same technological foundation Uber uses. And for those citizens who don't have a smartphone, all of the bus stops around the metroplex are already set up and simply need to be outfitted with a weather-protected display to accomplish the same thing, signal that they need to be picked up, and route them to where they need to go. This can be accomplished much more cost-effectively than the current model which comes with the incentive mount up debt and to "spend every dollar" (quote by DART CEO).”

Public transit using dynamic routing already exists within DART and several other neighboring transit agencies. In DART's case, it's called GoLink, which works like a subsidized rideshare, taking riders to nearby destinations, train stations, or transit centers. If you use the GoPass app and navigate within one of the GoLink zones, it will give you the option to request a ride. A DART vehicle or a third-party Uber driver will pick you up and take you to your destination.

The Benefits of GoLink

For riders, GoLink's curb-to-curb ride can be attractive to people going to and from areas that aren't designed to be walkable. For the transit agency, GoLink is more cost-effective in areas with low density and low ridership, where the price per passenger makes buses less viable. These zones allow people to get around their area or connect to denser areas which have enough people to justify fixed bus or train service. They are broken up into zones to reduce return journey time for drivers and as a form of bucketing to increase the likelihood that a driver can pick up other riders along the way. GoLink also allows more granular data on where riders like to go to and from. If a given corridor gets enough riders, it can signal there is enough demand for a bus route. Similarly, if a bus route is not getting enough ridership due to sparse or scattered ridership patterns, it can be converted to a GoLink zone. This system can allow the routes to adjust to shifts in population or destinations.

The Downsides of Dynamic

One of the major downsides of a dynamic system is right there in the name, the rider's arrival time is now dynamic, and can vary significantly. There are many factors such as: how many drivers are available, how far the driver is from the rider, how many riders will be picked up along the way, and how far out of the way are they and their destinations. These factors mean two downsides: scheduling and rider competition.

A dynamic arrival time means it is unattractive for trips with fixed times like commutes or medical appointments. In a fixed system, a rider can plan around departure and arrival times and be reasonably confident they will arrive when needed. Ridership competition is another problem that separates it from fixed route systems, and is more similar to private automobile usage. For a motorist in a personal vehicle, each additional vehicle on the road has a negative impact on their travel time and experience. Similarly, riders in an on-demand system can have their travel times increased when the driver goes off of the route to pick up or drop off others. In a fixed system an additional rider has little to no impact on arrival time, because traffic and the average number of stops are factored into the schedule.

The Costs of a Purely Dynamic System

As alluded to earlier, there is a tipping point at which the amount of ridership in a given area makes a fixed bus route more cost-effective. GoLink's goals of on-demand, curb-to-curb service mean that in practice, like Uber and Lyft, it's less ride share and more a taxi service. Even though they attempt to group rides together by subdividing zones and searching for riders along a similar route, it's unlikely many people will be at that exact place and time. This means that GoLink costs are relatively linear with demand, with each additional rider requiring an additional vehicle and driver. Buses address this by having specific stops to group people by location and a schedule to group people by time. Each additional rider makes the bus more cost-effective. Eventually, a bus reaches capacity and will need to increase frequency; this will require another driver and another bus, but the rate that these costs increase is much lower than GoLink. The additional frequency also makes the wait time for a transfer lower, further incentivizing ridership.

The relatively linear cost growth of GoLink means the subsidy per passenger is often higher than bus or train service, which can create some concerns around the incentive structure for the transit agency. If buses are more cost-effective when they carry more passengers, the transit agency is incentivized to move more people. This can be done by increasing frequency, adding new routes, or improving service in other ways to entice riders. This leads to better service, more riders, and a better bottom line — exactly the result you want. However, if you have a system where costs are not decreased by increasing ridership, the transit agency has no incentive to improve.

There are several neighboring cities that only have an on-demand system. One of them is Arlington, where the company Via runs a service similar to GoLink. It's notorious for long wait times due to lack of drivers, requiring users to continually request rides or give up and take a different ridesharing service. Via is not incentivized to improve this experience because taking more rides does not offset the cost of hiring more drivers, so the service remains suboptimal. On top of these issues, on-demand systems don't fulfill the other two main goals of transit, reducing traffic and air pollution.

If DART were to transition from the current system to this proposed model, they would need to:

Significantly expand all GoLink zones or make the entire service area one “zone”
Purchase lower-capacity cars and vans to expand the GoLink fleet
Remove fixed-route buses and incorporate them into the GoLink fleet

The Takeaway

GoLink has been a great addition to DART to serve areas that aren't dense enough for bus service. It serves as a bottom rung on the ladder below buses and trains, which have higher upfront cost but better scalability. If we assume the best intentions, the city officials of Plano and other areas saw the cost reduction of replacing some low-performing bus routes with GoLink in 2019, then made the assumption that these cost reductions could be replicated across the board. If we assume the worst intentions, some of the Plano DART board and city council members have known financial ties to Uber and could be looking to benefit themselves by increasing Uber's contract with DART. Regardless of motivation, DATA does not support the replacement of fixed route buses with GoLink at a city or system-wide level, as this would be more expensive and less beneficial to riders.

Fixed vs Dynamic Public Transit

The Benefits of GoLink

The Downsides of Dynamic

The Costs of a Purely Dynamic System

The Takeaway

Sources