frequent-rider discounts to decongest the peak?

Just in from the Aspen Ideas Festival, via Alexis Madrigal at the Atlantic.  Stanford Electrical Engineering and Computer Science Professor Balaji Prabhakar thinks he has the solution to peak overcrowding.

The frequent commuter program has two goals. One is to increase people's loyalty to the public transport system. We want people to be disloyal to their cars, to cheat on their cars. And the second major goal is to decongest the peak time trains and buses. The problem is that it is unpleasant to take a trip during the peak time. If we could achieve both goals with the frequent commuter project, it would be great.

The nice thing about this project is that it is going to do exactly what the airline miles do. You take a 10 kilometer trip, you get 10 credits. And Singapore can measure the kilometers. But if you make that same trip in the off-peak time, you'll get 30 credits. This creates new bonding between you and the system. People don't think of the indignity of taking a three-stop trip on their preferred airline versus a direct cheaper flight sometimes. In fact, they see the angle as, "I'm earning more miles."

Does anyone with regular experience in transit think this is a good idea?  If so, I'd like to hear.  My first reaction:

This sounds like a very very complicated way to do discounts for off-peak riders, and to reward very regular riders.  The fact is, the transit industry already has a system for rewarding frequent riders; it's called a monthly or annual pass.

A simpler solution to the peak overcrowding problem is to provide discounts for off-peak trips, as Seattle, for example, has done for decades.  This costs very little to administer and has the desired effect much more directly. 

When the need for sheer service is so urgent, why would a transit agency take on the massive administrative cost of a frequent-rider program, when the same money could go into driving buses and trains instead? 

I'm sure transit professionals will appreciate the interest from the "big ideas" people.  But from Madrigal's summary, this idea sounds like a fun metaphor inappropriately applied, suggesting the lack of any technical understanding of the transit capacity problem.

But I'm curious what others think …

26 Responses to frequent-rider discounts to decongest the peak?

  1. Alan Robinson June 30, 2011 at 3:58 pm #

    With airlines, complicated incentive programs are designed to feel like a game. Either you feel like you’re getting ahead by playing it or you feel disgusted by its attempt to disguise the value of the incentive by increasing the amount of time you have to spend to understand it.
    One of the few appeals of driving is the game one plays in ‘beating the traffic’ or finding the best route. Perhaps a transit points program might appeal to that mindset, but it would do so by rejecting ease of use and a feeling of accessibility that seem to be fundamental to good transit systems and the cares of most transit users.

  2. PBY June 30, 2011 at 4:36 pm #

    Loyalty programmes attract people and a certain type of person seems to thrive on them. I dont think passengers travelling off peak and paying a discounted rate see the cost of travel as a reward, they still see it as a cost.
    Perhaps an exisitng loyalty scheme could be combined with PT travel, combining smart cards so that you get points at the grocery store for your food shopping and then also get points when you use the card on the bus, train or ferry.
    That way the PT provider dosnt administer the loyalty scheme, it is run by a third party who already knows the loyalty scheme business.
    After all you can get loyalty points when you fill up with gas in your car!

  3. ant6n June 30, 2011 at 5:12 pm #

    Reduced monthly passes valid past 9am or 10am seem much nicer – I’ve used them myself. Seen in many German cities.

  4. BBnet3000 June 30, 2011 at 5:56 pm #

    Amtrak has a points system, though they don’t use it to encourage anyone to ride at different times. It could however encourage people to ride more, if it wasnt extremely inconsistent and glitchy as far as actually giving the points members are supposed to have earned.

  5. Brent June 30, 2011 at 6:01 pm #

    Ottawa used to have a fare system where you paid three tickets to board during peak periods, and only two to board in off-peak periods.
    They still have a system where you pay with multiple tickets, but now it’s only based on the type of route and passenger:
    1 ticket for children
    2 tickets for regular routes
    3 tickets for express routes
    4 tickets for rural express routes

  6. Rob June 30, 2011 at 7:17 pm #

    Back when Seattle instituted a peak/off-peak fare differential the purpose was not to decongest the peak, but to raise more revenue because peak trips are inelastic, meaning price can be raised with less of an impact on ridership. It’s hard for most people to switch to an off-peak trip in part due to workplace expectations, but also because off-peak transit service is usually inferior to peak service.
    Personally I think the bigger challenge for transit is managing the peak one hour, or the peak six minutes, but nobody wants a fare system complicated enough to tackle that. I wonder sometimes whether overcrowding on local buses just after the peak surcharge ends (and after peak capacity is available) has anything to do with the peak period surcharge. I’ve never understood that phenomenon, but it clearly happens here in Seattle on some routes.

  7. John June 30, 2011 at 8:57 pm #

    On Loyalty Programs: Metro Transit (Minneapolis/St. Paul) has a points-based loyalty program called Ride to Rewards ( A cool idea and fun, but I’m not sure it would be in any position to leverage a large shift in daily riding patterns even if more points were awarded for off-peak travel. It couuld, however, work well to get people to try something new (new route, new destination, weekend trips, etc.)
    On peak/off peak fares: the same Metro Transit has had peak/off peak fares for 20+ years. (peak local is $2.25 vs. $1.75 offpeak, peak express is $3.00 vs. $2.25 offpeak) I agree with the comment that this tends to be more about revenue generation and less about shifting demand.
    One piece of evidence: the ridership numbers at the fringe of the peak (just before and after) are lower not higher than the ridership just inside the peak period. If the peak fare caused a shift in ridership, there would be surges in ridership just before and after the peaks. There is anecdotal evidence of a few riders holding off a trip a few minutes to avoid the AM peak or rushing to get a trip in before the PM peak. But overall, commuters travel when they need to travel. And many purchase monthly passes that cover the peak fare and just don’t worry about it.
    The problem with peak/offpeak fares is that they are confusing, for customers and operators. The complication isn’t just the base cash fare, but it also complicates how daily, weekly and monthly passes are priced and marketed.

  8. Beta Magellan June 30, 2011 at 9:39 pm #

    I’m with Allen–this is meant to be somewhat complicated in order to get the commuter emotionally hooked to the program, which Prabhakar thinks will drive up ridership.
    Although it might be worth a try, I’m skeptical of such an incentive program’s efficacy because, in general, you want to keep people’s daily commutes as easy and quick-to-understand as possible (and almost every serious suggestion I’ve seen for increasing ridership focused on keeping things simple or simplifying things further). Although such an incentive program might work well for occasional travel, it seems like a hassle for everyday trips. Still, it might find a niche.

  9. Aidan Stanger June 30, 2011 at 10:11 pm #

    I had the same first reaction, but then I thought more about the purpose of loyalty cards. AIUI most of them are really designed to gather personal information which can then be commercially exploited, or to give them an excuse to email out details of special offers without being regarded as spammers! Any actual customer loyalty is regarded as a bonus.
    I don’t think transit companies would benefit that much from the personal information of the people who already use it regularly, and special offers for people who are already regular users aren’t going to increase its market share much. So this idea is a non starter.
    The situation’s slightly different with airlines – air miles can genuinely attract more customers because there are instances where the flights are booked by companies but the air miles go to individuals, and because there are some customers where quality of service and ambience are much more important than money. But I don’t think any of these are relevant to transit either.

  10. Andy Nash June 30, 2011 at 11:54 pm #

    One of the main reasons American Airlines started the first loyalty program was because air trips are on different airlines are essentially the same. The idea was to create a relationship where users would always select American because they were committed to earning miles, status, upgrades.
    But is the competition to public transport really offer the same service as public transport? Not really.
    On the other hand, I do think that public transport operators need to create better relationships with their customers, not because these customers will ride more (as Aidan points out), but rather to increase the political effectiveness of public transport agencies in the public arena.
    Public transport needs committed advocates to support complex and controversial government policy. We need to figure out how to create and support these advocates. I am not sure that ‘Frequent Rider’ programs are exactly right, but the idea of developing some sort of “loyalty” or “recognition” program could help in this regard.
    I’m working on a project called GreenCityStreets that is developing ideas for helping create and support these advocates, read more here: or

  11. Rob W July 1, 2011 at 12:24 am #

    As a rider I would prefer simplified fares with awards offered at different times. It’s too confusing for many riders here in Seattle to figure out different fares, times, when to pay, which doors are allowed to be used for boarding, etc. If everything were one fare, the ORCA participating agencies could offer a standard rate of x points per trip or a rebate that is credited back to a users e-purse for every ride — with bonuses at different times of day (peak vs. off-peak) or to promote specific routes. On the other end, the agencies could benefit their bottom lines by selling points, credits, etc. to area retailers to be credited back to riders. Awards could be as simple as credits to be applied toward future travel or credit to be used with partners.
    Systems along these lines could be one way of offsetting some of the budget problems currently facing transit agencies. Agencies could increase fares, but also sell bulk credits/awards to partners to be incrementally handed out to customers.
    One thing is for sure — I would rather figure out a rewards program than wait behind riders who protest fares because they don’t understand why fares are different than they were earlier in the day.

  12. John July 1, 2011 at 5:32 am #

    It seems that it would be hard for transit agencies to track the miles traveled by bus riders, and by train riders in systems without exit gates.
    He alludes to Singapore being able to to this though. How do they do it?

  13. Al Dimond July 1, 2011 at 9:49 am #

    Another Seattleite here. If there’s crowding just after the peak ends, I suspect it’s largely because service levels drop off faster than ridership levels post-peak. I don’t have much evidence, even from observation, because I mostly use reverse-commute buses that don’t have extra peak runs, so it’s just a guess.
    I think that, in Seattle, the distribution of “rewards” would quickly become a controversial political issue. I’d probably steer clear of that if I were in charge.
    Overall, in any city, I think it would take pretty drastic pricing changes to decongest the peak, and that a lot of people would be pushed off the system entirely before they changed their work hours. Like some others have said, travel demand during peak commute hours is not too sensitive to price.

  14. Joseph E July 1, 2011 at 9:58 am #

    In Singapore, transit riders slide or tap a card when entering and when exiting a bus or train, and fares are based on distance.

  15. Art Busman July 1, 2011 at 12:40 pm #

    I would never take 3 airline trips to get points versus a direct, cheaper flight. That’s insane. I like the idea of frequent flier programs for transit, but not as way to promote off peak travel. Rather, you should get credits for every pass you buy online, simple as that. In the future, I’m sure you’ll be able to track how many miles you travel by transit, and it would be cool to say I’ve traveled 100K transit miles, and you get a plaque or something. Also, in the future I’m sure there’ll be discounted off peak passes that are only good off peak. Then again, perhaps in the future, work hours won’t be so rigid and there’ll be more people working from home, and further out, 3D holographic representations of you at the workplace.

  16. Rob July 1, 2011 at 3:25 pm #

    Thing is, rewards will always be less effective than price if the objective is to get people to choose trips at less congested times. If price doesn’t have a big impact, then rewards will have a lesser impact still.
    Again here in Seattle, there was a rewards program a decade or more ago. Unlike airlines, who need to pay for rewards offered (so, increases price), the Metro rewards system was to increase the value of owning a pass by leveraging merchant discounts that cost the agency nothing. My impression is that the program just faded away because the cost to administer it was too high considering minimal use, but I’m not actually sure what the reason it ended was.

  17. Danny July 1, 2011 at 5:36 pm #

    The key to the loyalty enhancing aspects of frequent flier miles lies in the high value, low marginal cost benefits that they give.
    After an extended month long stay in a Hilton hotel, I was mailed a card saying that I was a silver something-or-other. It sounded cool because on my frequent flier system you start to get pretty good benefits at the silver level. But then I read about what the benefits were of being a silver level Hilton patron, and I threw the card away and haven’t touched it since. I had basically earned enough for almost a free night at the Hampton Inn.
    So while I like the concept between incentives for the marginal mid-day trips, I have to ask…what do these miles/points/whatever actually get you?

  18. Alon Levy July 1, 2011 at 5:38 pm #

    Jarrett, I think if anything you’re too charitable to this program. There already are programs for customer loyalty: unlimited season passes, as you note, but also prepaid multi-ride cards, in which you pay in advance for many rides in exchange for a lower per-ride fare.

  19. david vartanoff July 2, 2011 at 4:06 pm #

    Long before I would ever support ANY peak/off peak fare differential let alone some airline type point system, we will need to force flex time scheduling for lower wage workers. They have the least elasticity in showing up for work and transit is a higher % of their wages. Unlimited ride passes are the most equitable fare instruments for encouraging off peak usage.

  20. Alex B. July 4, 2011 at 8:01 am #

    Andy Nash hits on the key difference between airline programs (mainly fighting amongst other air travel operators) and hotel programs (amongst those who are staying in hotels) in that they represent a different universe of customers.
    Unlimited ride passes are far simpler.

  21. Chris Yuen July 4, 2011 at 9:45 am #

    I think this is a difference in framing. If you “reward” a non-peak hour user more than a peak hour user with the peak hour user still getting a little bit of “reward”, the peak hour user doesn’t feel penalized with higher prices. In effect, its the same as the cheaper discount system you mentioned, but I think it could be possible that the positive marketing effects of this out-weights the costs.

  22. AbNtransit July 4, 2011 at 3:53 pm #

    It is nice to incentivize riders, whether for loyalty or for usage during off peaks but. That’s just it, nice. PT systems that have to stretch every dollar or Euro should spend their efforts elsewhere, like grant writing for equipment or operating funds, lobbying for support, or sales for ad space. The ridership will come. Loyalty is inherent in their lifestyle. Those who need incentives and could drive, should be ignored. Let privates or market based tools gain their loyalty.

  23. G-Man (Type E) July 5, 2011 at 4:23 pm #

    I agree w/ Andy. Frequent customer programs are meant to override competition. I’ve purchased on a particular airline when the price was only slightly higher in order to get my miles. Could the concept be morphed into an incentive to ride at certain times? sure. But I think it would more likely produce an incentive for discretionary off-peak trips rather than create a reduction in demand at inelastic peak periods. I have asked people responsible for fares at my agency if we could offer a discount Nights and Weekends to encourage more travel when buses have lots of available capacity but the added complexity is a stumbling block.

  24. Henry July 5, 2011 at 9:33 pm #

    Hong Kong has some rewards worked into their smart card system, the Octopus Card. However, the purpose is not to encourage riding during off-peak periods. The rewards offered by some transit operators in Hong Kong is to get you back onto their system again. Hong Kong is covered by 1 rail transit operator, many minibus operators, and at least 3 major double-decker bus companies. So everyone wants a piece of the transit money pie in Hong Kong. MTR company operates the rail transit in the territory and has had offers previously for every 10 rides on their system. Some of the public transit airport lines (bus and rail) offer a discount on your return trip to/from the airport.
    So these “rewards” are not motivated by the need for decongestion, but it is interesting to know that rewards programs exist for some transit systems.

  25. Mike De Blasi July 5, 2011 at 10:11 pm #

    I think this might be a good system. One thing I would add is a way to reward bike commuters. When I lived in the SF Bay area, I biked everyday across the Golden Gate bridge to work. On the rare occasion, I would need to drive across the bridge or use the Marin bus. A fair system would have rewarded me, other than the fact that I was in the best shape of my life, by giving me a credit for the toll or bus ticket costs.

  26. Chris Bradshaw July 12, 2011 at 6:36 pm #

    It sounds like the idea is an attempt to avoid complicating fares by charging less for off-peak travel. As has been pointed out, loyalty programs serve to increase the # of uses with something more tangible as a reward. Both peak and off-peak use have their disadvantages I(shortage of seats at peak, shortage of buses at off-peak).
    Unlike airlines, transit operators charge a flat fare, regardless of distance (although some like Singapore and Wash DC’s subway, don’t). Second, no airline has the equivalent of unlimited-use passes, which would bankrupt them.
    The time has come for fare structures to grow up and charge by distance, with some differentiation by time-of-day, since long trips and the peak periods cost more to service, especially since the commuter trips represent both: long distances and huge volumes of riders demanding these long trips at the same time.
    Such a system can incorporate volume discounts simply provide a sliding scale so that each km is slightly cheaper than the previous one. There would be no need for an unlimited-use rate. Unlimited-use rates, by the way, involve a personal gamble: you have to decides at the beginning of the month/year how much use you _will_ need in the future. My proposal does not.
    Offering all card holders a reduced rate for using shared cars is another win-win promotion, helping to get across the idea that relying on transit makes personal-car ownership ridiculously expensive on a per-trip basis. No need to own two unlimited-use entitlements.
    We need simplification of fares, not more complications and gimmicks.