US demand response service cost agencies an average of $50.57 per passenger trip in National Transit Database Report Year 2024, against $8.24 for a fixed-route bus trip. Because a passenger trip counts a single boarding, a round trip to a dialysis appointment costs an agency closer to $101.
That’s a 6.1x gap, and nearly every explanation of it you’ll read gets the cause wrong. It isn’t that paratransit hours are expensive. They’re cheap.
Metric (NTD RY2024) | Demand response | Bus |
Operating cost per passenger trip | $50.57 | $8.24 |
Operating cost per vehicle revenue hour | $97.27 | $189.95 |
Passenger trips per vehicle revenue hour | 1.92 | 23.06 |
Farebox recovery | 6.8% | 11.0% |
Average passenger trip length | 8.8 miles | 3.5 miles |
Calculated from NTD Report Year 2024, TS2.1 Service Data and Operating Expenses Time Series by Mode. “Demand response” is NTD mode DR. Cost, trip and hour figures cover all reporting agencies; trip length is Full Reporters only, since passenger miles are collected only from that group.
Why a paratransit trip costs six times more than a bus trip
An hour of demand response service costs roughly half an hour of bus service — $97.27 against $189.95. It costs more per trip because it carries 1.92 people in that hour instead of 23.
The math is the whole story:
$97.27 per revenue hour ÷ 1.92 trips per revenue hour ≈ $50.57 per trip
Smaller vehicles, fewer operators per passenger carried, no station infrastructure. The hourly cost of running demand response is genuinely modest — 49% below a bus hour. What it doesn’t have is throughput.
Door-to-door service, booked one rider at a time, over trips averaging 8.8 miles against a bus trip’s 3.5, produces roughly two boardings an hour.
Which tells you where the money is.
Hold the hourly rate exactly where it is and lift productivity from 1.92 to 2.5 trips per revenue hour, and cost per trip falls to $38.91 — a 23% reduction. No wage change, no fuel saving, no service cut. Just more people moved in the same hour.
That matters because most cost-reduction conversations start with the hourly rate, which is mostly labour, and labour is the hardest line on the budget to move.
The evidence: agencies with cheaper hours often have dearer trips
We calculated cost per trip and trips per revenue hour for every US agency running demand response service with at least a million annual trips. The result makes the point better than any argument.
Agency | Cost per trip | Trips per revenue hour | Cost per revenue hour |
VIA Metropolitan Transit (TX) | $37.73 | 2.06 | $77.80 |
Dallas Area Rapid Transit (TX) | $40.66 | 2.22 | $90.06 |
Miami-Dade (FL) | $41.66 | 1.39 | $58.07 |
Metro Transit / Met Council (MN) | $54.03 | 1.52 | $81.87 |
Access Services (CA) | $59.47 | 1.88 | $111.82 |
Maryland Transit Administration | $64.41 | 1.95 | $125.32 |
Pace (IL) | $68.40 | 1.63 | $111.33 |
Capital Metro (TX) | $73.70 | 2.43 | $178.96 |
King County Metro (WA) | $87.96 | 1.63 | $143.07 |
MTA New York City Transit | $96.01 | 1.22 | $117.51 |
NJ Transit | $114.01 | 1.43 | $162.72 |
WMATA (DC) | $134.99 | 0.87 | $117.56 |
Look at the bottom two rows against Capital Metro. WMATA buys an hour of service for $117.56 — a third cheaper than Capital Metro’s $178.96 — and still pays nearly double per trip, because it moves 0.87 passengers an hour against Capital Metro’s 2.43.
Same story for New York City Transit: its hourly rate is lower than Honolulu’s, Orange County’s and Maryland’s, and its cost per trip is higher than all three.
Across the 272 agencies carrying 100,000 or more demand response trips a year, the median cost per trip is $43.84, the middle 80% runs from $18.71 to $76.24, and the full range stretches from $6.94 to $161.25.
Median productivity is 2.08 trips per revenue hour.
That spread is scheduling, rather than geography.
What agencies can’t change, and why that leaves one lever
Under the ADA, an agency’s demand, service area, hours, fare and trip mix are all fixed by federal regulation. Cost per trip is close to the only variable it genuinely controls.
49 CFR § 37.131 sets the service criteria for complementary paratransit:
Rule | What it requires | The lever it removes |
§ 37.131(a) | Service within 3/4 mile either side of every fixed route | Can’t shrink the service area |
§ 37.131(b) | Next-day service, with a one-hour negotiation window | Can’t require advance booking |
§ 37.131(c) | Fare no more than twice the fixed-route fare | Can’t price-ration demand |
§ 37.131(d) | No priority by trip purpose | Can’t deprioritise discretionary trips |
§ 37.131(e) | Same days and hours as the fixed route | Can’t trim the service span |
§ 37.131(f) | No capacity constraints, waiting lists or patterns of untimely service | Can’t ration by letting service slip |
Subsection (f) is the one that catches agencies out. Letting service quality drift while demand grows isn’t a way of managing a shortfall — it’s a compliance failure, and the kind of thing that surfaces in an FTA review.
Funding offers little relief. Agencies can treat ADA paratransit operating cost as capital under 49 U.S.C. § 5302, but only up to 10% of their Section 5307 or 5311 apportionment. Everything beyond that lands on local and state money.
So two numbers are genuinely in an agency’s hands: cost per revenue hour, and trips per revenue hour. The first is mostly wages and fuel. The second is scheduling.
How to lower cost per paratransit trip
Every lever that works does one of two things — it raises trips per revenue hour, or it converts a non-revenue hour into a revenue one.
- Group compatible trips. TCRP Synthesis 183 puts it plainly: more shared rides means higher productivity and lower cost per passenger trip. Grouping is the largest structural gain available, and the table above shows what it’s worth.
- Attack deadhead. Hours spent travelling between trips are paid for and carry nobody. Every one recovered becomes a revenue hour.
- Fill the gaps between booked trips. Idle capacity mid-shift is a scheduling artefact, not a demand problem. Trip Xchange exists to put trips into those windows.
- Treat no-shows as lost revenue hours. A no-show consumes the hour and produces zero trips, hitting the denominator twice.
- Know your own two numbers first. Until you’ve set your cost per revenue hour and trips per revenue hour against $97.27 and 1.92, you can’t tell whether you have a cost problem or a throughput problem — and the fixes are entirely different.
That last point is the one to act on this week. Our fleet utilization benchmarks piece covers how to work out where you sit, and what route optimization really means covers what actually moves the second number once you have it.
FAQs
Does paratransit cost money for riders?
Yes, but very little of the true cost. Federal rules cap the fare at twice the equivalent fixed-route fare, and fare revenue covered just 6.8% of demand response operating cost nationally in RY2024. Public funds cover the other 93.2%.
Why is paratransit so expensive?
Because of productivity, not price. An hour of demand response service costs about half an hour of bus service to run — it simply carries around two passengers in that hour instead of 23.
Door-to-door service, individual bookings and longer average trips all suppress throughput.
What is a good cost per trip for paratransit?
The median US agency running 100,000+ demand response trips a year came in at $43.84 in RY2024, with the middle 80% between $18.71 and $76.24. But cost per trip on its own tells you little — compare your trips per revenue hour against the 2.08 median first, because that’s usually where the difference actually sits.
Pull your two numbers, set them next to $97.27 and 1.92, and you’ll know within the hour which problem you actually have.
Most agencies find it isn’t the one they’ve been budgeting against.
Figures are national aggregates for NTD mode DR (demand response), which includes ADA complementary paratransit alongside rural dial-a-ride, senior transport and microtransit. NTD does not publish a separate national cost per trip for ADA paratransit alone.