Quick summary:
When budgets tighten, raising fares looks like the obvious fix. New peer-reviewed research on a real on-demand microtransit system shows the math often doesn’t work the way agencies expect: a simulated 40 percent fare increase raised revenue by only about 7 percent, because the riders most likely to leave are exactly the riders an agency can least afford to lose.
What you’ll take away:
- Aggregate ridership numbers can look stable after a fare increase while hiding very different reactions underneath
- Occasional, price-sensitive riders cut their usage sharply; frequent riders barely change their behavior at all
- A 40 percent fare increase modeled on real rider data produced only a 7 percent revenue gain
- Across-the-board fare increases tend to shrink the rider base an agency serves, even when total revenue holds roughly steady
- Agencies weighing a fare increase should look at rider-level data, not just system-wide totals, before deciding
Table of Contents
A tight budget year narrows an agency’s options fast. Federal pandemic-era subsidies are winding down, operating costs keep climbing, and a fare increase can look like the cleanest lever available: raise prices, collect more revenue, close the gap. It’s straightforward math on paper.
It’s not straightforward in practice. A new peer-reviewed study gives agencies a real, data-backed reason to slow down before pulling that lever.
What the Research Actually Found
OnDemand Microtransit is transit service that dynamically routes vehicles based on real-time rider requests, without a fixed path or schedule. A growing number of small cities and outer suburbs run this kind of service to cover low-density areas that a fixed bus network can’t reach efficiently.
Dr. Jia Li’s study, published in Transportation Research Part A: Policy and Practice and based at Wake Forest University School of Business, examined a citywide on-demand microtransit system in Wilson, North Carolina, that replaced the city’s Fixed Route bus network. The study used detailed rider-level trip data around an actual fare increase to see how riders really responded, not how a system-wide average suggested they responded.
The headline result: a simulated 40 percent fare increase raised total revenue by only about 7 percent. The reason is exactly why aggregate numbers are dangerous to rely on. The study’s data shows fare sensitivity split sharply by how often someone rides. Low-frequency, occasional riders cut their usage substantially when fares rose. High-frequency riders, the ones who depend on the service most, showed only modest reductions. The losses from the large group of price-sensitive occasional riders ate up most of the gains from the smaller group of riders who kept paying regardless.
Why the Surface Numbers Can Mislead an Agency
On paper, Wilson’s overall ridership looked stable after the fare change. That stability was the problem, not the reassurance it appeared to be. Underneath a flat topline number, existing riders were reducing their trips while new riders replaced them. An agency looking only at total boardings could easily miss that the composition of its ridership had shifted underneath it, and that the riders leaving were disproportionately the ones for whom the service was a genuine lifeline rather than an occasional convenience.
This matters beyond one city. The same dynamic shows up anywhere a service serves a mix of riders with very different levels of reliance on it. A fare policy that looks revenue-neutral or even revenue-positive at the system level can still be quietly pricing out the riders the service was built to reach, particularly those connecting to jobs, healthcare, or school who have no comparable alternative.
What This Means Before an Agency Raises Fares
None of this means fares should never move. It means the decision needs a finer instrument than a single revenue projection based on current ridership times a new price. Before adjusting fares, an agency benefits from looking at:
Rider-level usage patterns, not just system totals. Segmenting riders by frequency shows who actually absorbs a fare increase and who walks away from it, information a system-wide average can’t provide.
Which riders depend on the service versus use it occasionally. A fare increase that a low-frequency, price-sensitive rider abandons entirely is a very different outcome than one a daily commuter barely notices, even if both show up identically in a simple ridership count.
Alternatives to an across-the-board increase. Tiered pricing, targeted subsidies for high-need riders, or off-peak discounts can raise revenue from riders with more flexibility to pay while protecting the riders who have the least alternative options.
What the increase is actually meant to solve. If the goal is closing a specific budget gap, it’s worth modeling whether that gap is better closed through fare policy at all, or through route efficiency, vehicle utilization, or grant funding that doesn’t put the burden on the riders who can least absorb it.
The Underlying Lesson
A 40 percent price increase producing a 7 percent revenue gain isn’t a rounding error. It’s a sign that the demand curve for a service like this doesn’t behave the way a simple revenue model assumes. Agencies that model fare changes using rider-level data, rather than system-wide averages, are in a much stronger position to know what a fare increase will actually buy them, and what it will actually cost the riders who need the service most.
40% → 7%
Modeled fare increase versus the resulting revenue gain in a real on-demand microtransit system
Wake Forest University School of Business, 2026
Stable on the surface, shifting underneath
Aggregate ridership held steady after the fare change, while existing riders cut trips and new riders replaced them
Wake Forest University School of Business, 2026
2 rider groups, 2 different reactions
Low-frequency riders cut usage substantially after a fare increase; high-frequency riders showed only modest reductions
Internal framework, based on the cited research
Frequently Asked Questions (FAQ): Transit Fare Increases and Revenue
Does raising transit fares reliably increase revenue?
Not necessarily. Peer-reviewed research on a real on-demand microtransit system found a simulated 40 percent fare increase raised total revenue by only about 7 percent, because revenue gains from less price-sensitive riders were offset by ridership losses among more price-sensitive riders.
Why did ridership look stable if the fare increase was pushing riders away?
Because the topline number masked rider-level change. Existing riders reduced their trips while new riders took their place, so total ridership held roughly steady even though the underlying rider base shifted.
Which riders are most affected by a fare increase?
Low-frequency, occasional riders are the most price-sensitive and cut their usage substantially after a fare increase. High-frequency riders, who tend to depend on the service most, show only modest changes in behavior.
What should an agency look at before deciding to raise fares?
Rider-level usage data segmented by frequency, not just system-wide ridership and revenue totals. This shows who is likely to leave, who is likely to stay, and whether an across-the-board increase will actually net meaningful revenue.
Are there alternatives to an across-the-board fare increase?
Yes. Tiered pricing, off-peak discounts, and targeted subsidies for high-need riders can raise revenue from riders with more flexibility to pay while protecting riders who have the least alternative transportation options.
Does this research apply only to microtransit, or to fixed-route service too?
The study specifically examined an on-demand microtransit system, but the underlying dynamic, that aggregate ridership stability can mask very different rider-level responses to price, is a general risk any transit agency should account for when modeling a fare change.
Thinking Through a Fare Change for Your Network?
Fare policy decisions carry real tradeoffs between revenue and access, and the right answer depends on the makeup of your specific ridership. If your agency is weighing a fare adjustment, modeling the rider-level impact first, rather than relying on a system-wide estimate, can help you see the real tradeoff before it’s made.