Could TriMet become a “free” (or nominally-priced) service, Part 2

Last fall, Portland Transport considered the question of whether or not TriMet could become a free (or nominally-priced) service. Right now, the Farebox Recovery Ratio of the TriMet system in aggregate is about 25%, and last fiscal year fares provided nearly $100 million of TriMet’s overall operating budget.

This discussion occurred right after TriMet first announced that it would be having major budget issues for FY2013. Since then, the announced service changes have been ugly, with fares going up, and service going down; and TriMet suggests that more will come next year, even if the economic recovery continues–a state of affairs TriMet blames on its current labor contract.

However, as noted in the open thread, another Oregon transit provider recently switched to providing free service. That agency is the Corvallis Transit System.
About CTS

CTS, as the name suggests, provides public transportation services for the city of Corvallis, OR. Unlike TriMet, which is an independent transit authority organized under ORS 267, with a board appointed by the governor, CTS is a department within the City of Corvallis. While being a municipal department puts CTS at somewhat greater risk of being a political football (it is subject to the whims of Corvallis’ city council), as a department within a city it has a wider array of taxation options at its disposal. Whereas TriMet is statutorily limited to the payroll tax as a source of tax revenue, and it can only change the levy within parameters approved by the Legislature, the City of Corvallis has plenary taxing power. And recently the council voted to end the collection of fares on CTS, converting it to a free service. Transit service is instead funded with a Transit Operations Fee levied on homes and businesses, with the average household paying about $30 per year. Not per month, per year.

Corvallis itself is a college town in the mid-Willamette Valley, with a full-time population of about 55,000 residents, many of which are students at Oregon State University. Not surprisingly for a mid-sized college town, the demographics are dominated by degreed professionals and college students. While OSU has long been regarded as the more conservative of Oregon’s public universities, the city is quite progressive. According to the US Census Bureau, Corvallis has the highest percentage of bicycle commuters in the United States. The city is quite compact and flat (and is a good place to ride), and also offers numerous recreational cycling opportunities in the surrounding area.

CTS operates 12 bus routes, with eight urban circulators, running at hourly or half-hourly frequency, and 4 peak-hour commuter routes. Service on the non-commuter routes is provided for about 13 hours a day, from approximately 7 in the morning to 8 in the evening, with the “Beaver Bus” running evenings on Thursdays to Saturdays to haul drunken students around. :) Service is provided on Monday through Saturday; no service is provided on Sundays. The 8 regular routes run on either hourly or half-hourly frequencies. CTS contracts out operations to First Student, a private bus operations company. (Previously, CTS contracted with Laidlaw, but the latter company was acquired by and merged with First Student several years ago). CTS drivers, like TriMet drivers, are represented by Amalgamated Transit Union local 757, and are having their own fights over health-care benefits.

A more detailed comparison of TriMet and CTS is appropriate. (For those who want to consult the National Transit Database, CTS’s code is 0047, and TriMet’s is 0008.)

TriMet CTS
Service district area (mi2) 570 13.8
Population 1.67M 55k
Density (p/mi2) 2930 3986
Service hours 2.1 million (bus, MAX, WES) 27k
Service hours per capita 1.26 0.49
Boarding rides 100M (FY2011) 884k (FY10/11)
Boarding rides per capita 60 15
Boarding rides per service hour 47.6 36.8
Operating budget $400M (FY2011) $2.8M
Op. Cost/Svc Hour $190 $103
Op. Cost/Population $240 $51
Routes 79 bus, 4 MAX, WES 12
Vehicles ~600 40′ busses, 127 LRT vehicles, 6 WES cars 11 35′ busses, 1 trolley

One of the first figures that jumps out is that CTS’ pays only about 20% of what TriMet spends on operations per capita. The biggest reason for this discrepancy is that TriMet supplies over 2 1/2 times more service hours per capita than does CTS, with numerous frequent service and rapid transit routes. However, even with that taken into account, TriMet’s operating costs per service hour are nearly twice as high. (TriMet publishes a separate calculation of “bus-equivalent” service hours which weights MAX trains by a factor of nearly 5 to account for the capacity difference, giving it an alternate figure of 3.17M service hours, and a cost/svc hour of $126, nearly the same as CTS. We will ignore this figure, however, as the extra capacity of a larger vehicle vs a separate vehicle is only useful at those times when the capacity is needed, whereas two independent vehicles can be run on two different routes, providing twice the coverage.) Whether the operating costs are truly apples-apples comparisons is unclear–as CTS is a city department, things like administrative overhead may not be included in its cost structure.

Nonetheless, TriMet is more successful from a ridership perspective. Due to the differences in city size, it is difficult to meaningfully compare the two systems. Corvallis is a very walkable city, with minimal traffic and only a few miles from edge to opposite edge. There is significant residential density around the university, with lots of mid-rise (6 stories or so) dormitories and apartments, fraternity and sorority houses, and other compact student housing, but the rest of the city is dominated by single-family homes. Demand for transit is intrinsically less in a smaller city. CTS performs well compared to other systems of similar size.

Do the numbers work for TriMet?

In considering whether such a scheme would work here, we’ll consider several cases. The amount of revenue generated by the residential portion of the CTS transit fee is approximately $670k, or about 1/4 of CTS’s budget. We assume a similar ratio would apply to the Portland metro area. Assuming an average household size of 2.5, the same fee structure in Portland ($30/household/year) would produce revenue of about $20 million, only 5% of TriMet’s current operational budget. If businesses contributed $60 million, that would provide $80 million, only a fifth of what would be needed. For this sort of fee to replace both the payroll tax and fares, a per-household fee of $150/year would be necessary, with a correspondingly higher levy on businesses.

Another option would be to only levy the fee as needed to replace fares; in which case a slightly higher fee on households (averaging about $37.5/household/year) and businesses would suffice to replace fare revenue.

This analysis ignores one significant cost savings from free service–the cost of fare collection. According to the recent white paper on electronic fare collection, about 10% of the fares TriMet currently collects goes to pay for collection and enforcement, so the net revenue from fare collection is only about $90M, not $100M.

A few issues with this:

  • First and foremost, TriMet lacks statutory authority to levy such a fee. For a scheme like CTS’s transit service fee to work, the law would need changing.
  • It’s not clear that there would be political support in the wider metro area. Corvallis is a very liberal city, and the proposal had no problem passing. However, one could expect widespread opposition from the suburbs.
  • One other issue is the CTS fee is regressive–it is essentially a head tax on households. At a price of $30/year or thereabouts, this is tolerable; after all, motorists pay more to license and register their cars. However, were the fee to get into the hundreds, it would become increasingly a burden on the poor.

A few other options: A city within the metro area, most likely Portland, could elect to “buy” free fares for boardings within its jurisdiction, much as downtown Portland merchants at one point subsidized Fareless Square. It then could use its plenary taxing powers to enact a scheme similar to CTS. (Such a proposal is more likely to be politically tractable in Portland, as well). I don’t have data about what percentage of service hours and boardings occur within Portland– a fair price for such a thing would depend on these. Also, having a partial free-fare zone would still require TriMet to provide fare collection infrastructure for those locations and persons outside, reducing the benefit of completely fareless service.


So far, this article has focused on the mechanics and logistics of free-fare service; however broader policy implications should be considered as well.

  • A common objection is that by making transit service free, one will encourage overconsumption. The first article in the series also discussed the possibility of nominal fares (say, fifty cents per ride) to discourage activities like camping on the bus. Even ignoring that issue, many will complain that a free service is economically inefficient.
  • A related issue is that if a service has a farebox recovery ratio of zero, this removes both the incentive to provide better quality service, and reduces the ability of TriMet to scale service to meet future demand.
  • On the other hand, the primary competition of transit (the automobile) is also quite heavily subsidized, and has a cost structure that encourages people who own cars to prefer them to transit. Given the economic externalities associated with transit and driving, encouraging people to use transit is arguably wise public policy even if offends the economic sensibilities of some.
  • There are plenty of other public services that we, as a society, choose to provide for free (meaning financed by general taxes, as opposed to by consumption or usage charges). And while transit (like most services) can be overconsumed, it cannot be hoarded. Were TriMet to provide taxpayer-funded free beer, I’d bring a tanker truck to the tap (and so would everybody else), but I can’t ride more than one bus at a time, nor take the seats home and put them in my garage and prevent others from using them..


13 Responses to Could TriMet become a “free” (or nominally-priced) service, Part 2