photo by VeloBusDriver

By unfortunately-unchallenged tradition, most transit agencies obsess with a deceptive metric they call “Farebox recovery ratio” — the amount of fares collected over a specific period of time, divided by operational costs over the same period.

King County Metro has a goal of 25% farebox recovery. Sound Transit has varying farebox recovery goals on its different services: 20% for ST Express, 23% for Sounder, and 40% for Link Light Rail.

These numbers are actually gross farebox recovery, and can be cooked simply by redefining the divisor.

A somewhat more useful metric would be net farebox recovery, which is what is left over after the costs of fare collection are subtracted from gross fares received. These costs include costs for printing fare media, cash handling costs, costs for fare collection equipment replacement, costs for staff devoted to fare collection and enforcement, and operational costs caused by extra dwell time to collect fares. Metro estimates its annual cash fare collection costs at $2.5 – 3 million (page 28), not counting operational inefficiencies caused by cash fumbling.

I’d like to suggest a third metric, which would involve no additional data collection beyond what is needed to calculate gross and net farebox recovery: farebox recovery efficiency, which is to say, net farebox recovery divided by gross farebox recovery.

This metric could help bring a reality check to how the current fare systems and fare changes being considered are impacting agencies’ bottom lines.

As covered last month, Metro is in the process of considering a fare change. Metro has three oddball aspects of its fares that don’t mesh with the fare systems of other agencies in the region.

* a peak-hour surcharge
* a zone boundary that isn’t a county line.
* paper transfers

The committee vetting Metro’s fare change proposal will likely get to see how getting rid of the first two features will impact gross farebox recovery. The third feature, which adds confusion to inter-agency transfers, is a political hot potato unlikely to be touched by the committee.

But dealing with that third feature could impact what level of flat fare Metro could afford. Eliminating paper tranfers would immediately wipe out $200,000 of annual printing costs, and likely yield a significant reduction in change fumbling. However, such savings would not be reflected in the gross farebox recovery projections.

Bringing the net farebox recovery and farebox recovery efficiency data to the table, and putting the elimination of paper transfers on the table, could make the difference between a less efficient $2.75 Metro flat fare and a more efficient $3.00 Metro flat fare.

Author’s Note: It took over a year for the author to notice he got “more” and “less” backward in the last paragraph of the original post. He apologizes profusely for the error undermining the whole point of the post, and curses his poor editing skills.

35 Replies to “Farebox Recovery Efficiency”

  1. What would it realistically take for Metro to seriously consider reducing its expected farebox recovery? Metro’s higher-than-usual value for this metric has pushed fares up, and made short bus trips uneconomical. In fact, it drives fares up so much that they had to create a low-income fare, because the regular fare is too high! And, to keep this arbitrary farebox recovery goal, the normal fare has gone up even more. Is Metro a public service or a taxi company?

    1. You’re right that Metro is unlikely to propose a revenue-negative fare restructure. Step one would be to get Metro to look at the right measure for revenue impact, which gross farebox recovery is not.

    2. A change in the political climate at the county (and state) level. Perhaps this change has already happened, and perhaps it hasn’t.

      The goal for Metro was, as I recall, part of the political deal that ended 40-40-20, and more generally was part of a deal to secure additional funding authority [basically, there were a bunch of efficiency related changes that the County committed to making in return for this funding].

    3. The county council sets the farebox recovery ratio based on what looks good politically. If it’s too low, people say riders aren’t paying their share. If it’s too high, people say it’s leaving low-income people out. The fare goes up because expenses go up (gas, medical benefits, materials, labor), but the ratio remains the same. (Although it floats within a window like currencies do. When it reaches the floor, that’s when the county raises the fare.)

      The ending of 40-40-20 wasn’t about raising fares per se although there may have been an increase at that time. It was part of a grand bargain in which the county allowed Metro a 2-year tax surcharge to maintain service after the 2008 crash, plus the ending of the Ride-Free Area, plus the adoption of Metro’s new performance metrics.

      40-40-20 referred to a formula for adding or reducing service to match changing revenue. Seattle started with more service than the suburbs, so the goal was to gradually equalize the service level throughout the county by giving 80% of the additions to the suburbs and taking 60% of the reductions from Seattle. That coincided with the period when the council often override Metro’s reorganizations to preserve a route for the one person or few people who complained to the council about losing it. The new metrics instead tell Metro to look at the 25% lowest-ridership routes every year and consider shifting hours to the most underserved corridors (where ridership is being held back by no service or infrequent service) or to improve reliability on the most unreliable routes. It doesn’t require the bottom routes to be deleted or restructured but it requires Metro to at least look at them. Now that there have been several years of trimming the bottom, the bottom routes are now stronger than the were and thus less likely to be deleted. There’s also a provision to reserve some of the revenue for coverage service, which these bottom routes might be fulfilling.

      All this allows the council to tell Metro, “Stick to your performance metrics”, rather than micromanaging individual routes and second-guessing the planners and catering to squeaky wheels. The council did a good job of that for several years. The past couple years it has started backsliding a bit, such as imposing the 71 and 78 on the U-District restructure (which took hours from the 67 and 45 which were going to be 10-minute but are now 15-minute). That kind of backsliding is like the bad old days and makes Metro self-censor to avoid being slapped down, which may be why Metro withdrew deleting the 12 in the U-Link restructure, since the squeaky wheels are loudest about the 2, second-loudest about the 12, and third-loudest about the 42. However, the backsliding has been only a little bit, and the council has let large restructures go through, from the 2012 C/D restructure to the 2016 U-Link restructure and others, so it should be commended for that.

      1. “It was part of a grand bargain in which the county allowed Metro a 2-year tax surcharge to maintain service after the 2008 crash”.

        My recollection is that there were several tax payer protections in that grand bargain, and that setting a particular fare box recovery ratio policy was one of them (as was getting rid of the RFA and moving away from 40-40-20). It’s possible that I am misremembering or that my understanding of the deal was faulty at the time.

    1. All things being equal, a higher fare will tend to be more efficient than a lower fare. Higher-than-expected ridership loss might alter that calculation, but so might right-sizing service after ridership loss.

      If Metro is willing to reduce its fare collection costs by eliminating paper transfers, having a cash surcharge, and/or banning cash payment next to ORCA vending machines, and is willing to go with a slightly-lower fare than what it was planning assuming it wouldn’t touch the hot potato of electronic-payment incentivization, that lower fare *might* end up doing better on the efficiency calculation.

      I personally have trouble stomaching raising the base fare from $2.50 to $3.00, but that is likely what Metro has to do to get what it needs out of this exercise: increased *gross* fare recovery .

      Switching to looking at *net* fare recovery, the more accurate way to gauge a fare increase’s impact on the budget, and at the efficiency metric, should show farebox recovery will become more efficient at any fare in the reasonable range when cash payment is disincentivized.

      The reason for tying the incentives to choosing a lower fare increase is mostly about making the incentives politically viable, to be blunt. For purposes of a cash surcharge, it is also about making the cash fare an even dollar amount.

      1. Brent, could you explain the insoluble political problem that sticks us with a really expensive, clumsy money-shredder of a system? Driving around the logging country down here, I’m seeing stumps of thousands of square miles of transfers.

        Look at an ORCA card as a small plastic version of what just about everybody has been carrying their money since the Enlightenment caused people to realize how much gold WALLETS saved, instead of hanging all those scraggly “Cut-Purses”.

        Victims and thieves were able to agree that pickpocketing was easier to deal with. Also lost and found considerations. Doubt that carriage companies liked to have drivers stop the coach in the middle of a dark gloomy moor go to the back of a coach to look for somebody’s bag of gold.

        None of our component agencies has ever lost a dime over people carrying their fare money in a wallet. Which nobody has ever seemed to have any problem finding or buying.

        You might also mention that the transit industry doesn’t have an engine or motor powerful enough to haul a wagon carrying the bag that fare-connected delays cost us every year.

        Mark Dublin

  2. A $2.75 fare is not more efficient if you have to take two buses and therefore have to pay $5.50 if a transfer is not available.

    I agree getting people to stop paying cash is the way to go.

    1. The efficiency metric is for the agency, not the rider. I assume people taking two-seat rides more than once in a while will acquire an ORCA card (which will hopefully become free, if the county council gets its way, well before September 2018).

      Yes, that makes more one-time work for infrequent riders, but the ride for everyone else on that bus becomes a little bit faster.

    2. The $5 ORCA fee would be paid off in just two trips under that scenario. Then all other transfers after that would be free. I have sympathy for those who have only three dollars in their pocket at a time, and visitors and occasional riders who find themselves at a bus stop far from an ORCA vendor or they don’t know where the ORCA vendors are. But for those who have $10 or more in their pocket and ride at least semi-often, they can get themselves to an ORCA vendor.

  3. As discussed in a recent post, designing a “fair” fare is really hard. In part, I think this is because Transit is an example of a “club” good ( Basically, it is easy to keep someone off a bus or train, but one person’s use does not diminish the ability of other people to ride (except at peak periods).

    This is case for moving as much as possible to a membership cost structure, where you pay a monthly fee and then have unlimited rides. This better reflects the economics of transit (high fixed costs, very low marginal costs). You could have tiers of membership – ie a low cost membership that only covers non-peak periods, etc.

    Obviously these already exist, but maximizing the use of monthly passes should be the goal for transit agencies.

  4. Is there a way of finding out the fare box recovery of single lines? I’ve always wondered if the 41 makes a profit.

    1. I remember seeing a Metro report showing farebox recovery by route from a few years back. No route made a “profit” – IIRC the highest routes were in the 60-70% range.

      Not sure if it is still published.

      1. Found a 2008 report on the website – it breaks out Peak, Off-Peak, and Night. All of these %s below are Peak. Also, keep in mind these are 2008-era routes:

        Highest farebox recovery: Route 1 (71%)
        Other routes above 60%: 2N, 2S, 3N, 4N, 12, 13, 15, 48S, 68, 72

        Other routes mentioned below:
        7: 35%
        41: 39%
        44: 47%
        49: 49%
        212: 36%
        301: 40%
        358: 50% (proxy for the E)

      2. Route 1 is the highest? That sounds like something is wrong because the 1 is a low-performing route (that’s why it was split from the 36 which needs higher frequency, and joined with the 14 which is also low-performance). Are you sure you’re not interpreting the numbers inversely (that the 71% is the non-farebox share rather than the farebox share)? It’s hard to believe the 7 is down near the bottom when it runs a gazillion times a day and carries lots of bus-dependent people up and down the valley all day.

      3. That was in 2008, and there have been some pretty big changes since then. We don’t have more current numbers, but as I suggested above, we can make some guesses using the reported numbers for trips per platform hour. It’s not perfect, most significantly because you have to allocate transfers somehow, but it’s the best proxy to which we have easy access. [I am not entirely convinced that route by route fare box recovery is a useful metric, but…]

        Based on 2015 figures from the service guideline report

        The best is almost certainly the D that manages 80 rides per platform hour peak, 70.7 off peak and 46.2 at night. [Remember no D in 2008]

        The 1 is 54.8/46.4/29.8
        The 7 is 53.9/61.7/34.9

        That 7’s off peak 61.7 is better than any other route except the aforementioned D and the 82 (61.8)

    2. The routes most commonly cited here for high farebox recovery are the 48 (pre split), 49, 7, and I think the 44. Maybe the E, although it may be just high in ridership. I haven’t heard anything about the 41.

    3. Metro’s performance reporting included that metric pre-Service Guidelines Report, but no longer does.

      My recollection is different from Alex’s. I believe just a couple of peak commuter routes had 100%+ farebox recovery. The ones that are coming to mind are 212 and 301, but memory is hazy.

      1. My recollection is that even the 212 doesn’t break even [and that it isn’t particularly close]. The problem is that that the platform hours per trip are actually fairly high because of the slog through downtown to drop passengers off; the lightly loaded, congested, counter-peak return trip; and the small number of journeys against which the cost of the lengthy deadhead before and after service can be amortized.

        Arguably, fare box recovery understates how expensive routes like the 212 are because it ignores capital expenses for providing their lightly (in the sense of few trips per day) used buses.

      2. fare box recovery understates how expensive routes like the 212 are because it ignores capital expenses for providing their lightly (in the sense of few trips per day) used buses.

        That’s a fair criticism, maybe..But if those buses are “by the book” only charged to those runs then they should last longer which evens it out. Now, light rail capital costs…

      3. It’s not that hard to get an upper bound on the number though, since rides per platform hour is available and maximum fare per ride are known. Yes this ignores reduced fares [probably the right way to handle things, although it’s healthy to be transparent at a system wide level about the nominal lost income] and transfers. It also ignores any per platform hour cost differences between routes (my understanding is that this is pretty negligible).

        Looking at the 2015 service report. The 212 manages 43.2 rides per platform hour. That’s pretty good, especially since they are almost certainly mostly peak 2 zone rides, but it doesn’t quite get you to $150. The 218 might just make it at 50.1 rides per platform hour.

        I stand corrected on the 212. It almost certainly isn’t profitable, but it’s probably not as bad as I implied above.

  5. most transit agencies obsess with a deceptive metric they call “Farebox recovery ratio” — the amount of fares collected over a specific period of time, divided by operational costs over the same period.

    These numbers are actually gross farebox recovery, and can be cooked simply by redefining the divisor.

    In our system both numerator and denominator can be “cooked”. There is a formula where someone paying their fare for a Metro bus gets appropriated some amount of $$$ when they x-fer to an ST bus (or vice versa). But the cost of fare collection should already be accounted for in operational costs so I don’t see why you’d adjust for Net Recovery; seems like double dipping. Although, I’m not an economist and don’t play one on TV :-)

    Bottom line is as long as your consistent it is a “fair/fare” metric. What I think matters is that it be a way to compare performance with other transit agencies. Metro was woefully lacking in comparison prior to the 2008 reset. Using this metric they upped their game. I understand Metro works at the whim of the country council but putting a spotlight on this metric, when used as a comparison with peer agencies I think is a good thing.

    1. I think the suggestion is that the numerator should be (fares collected – cost of fare collection) and that the denominator should be (operating cost – cost of fare collection). The point is that calculating things this way rewards being pragmatic about any costs that may be associated with maximizing fare collection.

  6. Also to be considered is the cost of the fare collection equipment. A single stand alone registering farebox can easily be $20,000 and I would imagine implementing a separate smart card system per vehicle could be just as much if not more. Metro does need new fare boxes and for your $20k they do come with smart card features built in, so it may make sense to combine systems where practical. Operationally, you also need to consider the costs of cash handing as well, whereas a smart card/electronic system is mostly back end. The problem with a smart card based system is that you need additional retail locations or machines to handle on the spot or cash purchases to make it equitable for everyone. Collecting fares any way you look at it is not a cheap proposition.

    1. I suspect a QR code-based system would be really, really cheap. I’m not a technology expert, but given how this is the way that airline tickets are working these days, there is little reason that bus tickets couldn’t work the same way. People with smart phones could seemingly just flash their QR code on their phones. buses could have a very small printer that uses adding machine rolls to print QR transfer strips when people pay cash. Machines could even dispense QR receipts at Link stations so people could pre-pay fares. The major headache would be to make sure that the machines are connected, and that any printing is done reliably.

      Metro and many other major bus transit operators are implementing various versions of smart phone fares. I think that it’s only a matter of time that this is expended to be the way that fares will be collected. Once this becomes more popular, bus boardings will be faster and fare policies will get more imaginative.

      On the other hand, I think we will probably look at having separate smart cards as old technology in a few years as we continue to reduce the number of cards in our wallets and purses.

  7. Does anybody really think that somebody who can pay a $2.75 fare will stop riding transit over a quarter fare increase? Transit does make arrangements for people whose incomes are too low to be passengers.

    Found out today that coffee is due for a price increase. I don’t see a customer drop-off. Could be because it’s addictive. But also same principle as for trillion dollar a shot cognac. Large strata of people gain prestige by paying too much for things.

    Good example is an Indonesian coffee called “Kopi Luwak”. Google it. Also the price, for perfect illustration. Dare you, Howard. Dare you!

    And c’mon, South Lake Union’s JP Morgans! We know why you’re not riding your own streetcar. How much do we have to raise the fare before it’ll impress enough friends and investors to order their own “Palace Cars”, like History’s real capitalists.

    But here’s a “Not If But When” prediction. When a badly-needed Revolution frees all the real accountants from their present “Black Site” prisons in Poland, every universal public need, including whole range from schools to buses, will simply be paid for with taxes.

    Precisely because for the average person, fair taxes are the simplest way to collect the money. Toll roads will go the way of per-flush coin slots on every individual toilet. Probably never existed.

    Though bet somebody’s got the original patent on one that looked like an elegant miniature antique cash register. And that Revolution Hardware has its attorneys working 24-7-365 to get the papers.

    Google “Old-fashioned mechanical transit fareboxes.” My favorite ones in Chicago and Denver had a little hand-cranked wheel that spun around and counted the change as it dumped the coins into the farebox.

    Could save us a fortune in fare evasion as generations of children demand that their mommy and daddy pay their fares. Give these kids free rides for monitoring the box all day. Though their fare inspectors’ uniforms should have brass buttons and starched white collars. Transit police too!


  8. The only time I get paper transfers is if I pay a friend’s bus fare. Since I am more transit savvy than many of my friends and family, this happens every few months. It seems like a strange system. The advantage is if we split after sharing the start of the journey, we can each use our transfer on our separate second leg. The advantages end there. It takes a long time for the driver to set the reader for two. It takes time for the driver to rip the two transfers. The biggest disadvantage is the paper transfer only works for metro.

    I’m always dismayed when a computer system through poor design has to fall back on manual systems. Here is a first thought about how to design multiple fare from one card. Every fare reader can have +1 and -1 buttons for the rider to add or subtract how many fares the card will pay. The card will hold that many transfers till the next time it is used. If I only use the card to admit one person at the 2nd leg, I’ll only have one transfer on the card for a third leg. If at the start of the 2nd leg, I use both transfers to get two in, then I still have both transfers for a third leg.

    1. If you have a group of 12 riders, and pay for 9 of them, who gets the warnings when the FOE comes along?

  9. As I understand it, Metro includes advertising revenue in their numerator, which is why their farebox recovery is around 30%, another “cooking” method. Other agencies merely include the fares.

    I agree, net farebox recovery would be more useful, though some of the costs you’re interested in getting to are a cost to obtain, especially for our behemoth agency in the region, King County Metro.

    Another cost you didn’t mention is off-board fare collection/vending machines, for those require regular maintenance, troubleshooting, and a firm to pick up the money. There are also costs that are specific to a particular type of service, as in the case of off-board fare collection machines and BRT shelters.

    The peak-hour surcharge is oddball, but I witnessed first-hand how this freed up seats on the commuter routes to downtown Seattle when Metro had a surcharge on seniors’ fares in I think it was the early 1990s. My speculation is that many seniors changed their appointments to off-peak as a result of that fare change, for the difference was noticeable on the buses I rode (as an aside, not all seniors are low-income, which a senior fare assumes). I have less of an issue with this charge other than the confusion as to when it begins in which direction and at what point. The other two items you mentioned – a zone boundary that isn’t a county line and especially paper transfers make no sense. Eliminating paper transfers in conjunction with lowering the cost of obtaining an ORCA card – maybe even offer them free for the first 90 days after transfers are eliminated and greatly expanding where one can get them – makes the most sense.

Comments are closed.