How CitiBike is like Ecuador

by Felix Salmon


© Felix Salmon

© Felix Salmon

What’s the point of having a contract, if there’s no remedy?

It’s not easy to edit all of the verve and humor out of Bike Snob’s prose, but somehow the NYT has managed to do it: today’s op-ed about CitiBike is a dry and evenhanded look at its successes and failures. Specifically, it’s an examination of the fact that the scheme is losing money:

There was one expectation that the world’s most heavily scrutinized bike share system failed to meet: economic self-sufficiency.

As robust as the program’s user numbers were, it turned out that offering cheap annual memberships ($95) and relatively costly 24-hour passes ($9.95) was not the best pricing structure to underwrite the system.

Bike Snob has a simple solution to this problem. CitiBike is a form of public transit, public transit is something which is almost always subsidized by the government, therefore CitiBike should be subsidized by the government.

I don’t disagree. I would love to see NYC subsidize the rollout of CitiBike: right now, it’s in the area of the city which is best served by the subway system, and its expansion could and would do wonders for people living much further distances from a subway line. What’s more, if you want a system which is going to be popular with tourists, you’re going to have to expand it up Museum Mile to the Upper East Side and Upper West Side, even if doing so annoys the bike-rental companies in Central Park.

When CitiBike was launched, the hope and expectation was that it would be profitable for its operator, Alta Bicycle Share, and that therefore Alta would have every incentive to expand the scheme organically. (More bikes = more profit!) It hasn’t worked out that way, however, which means that if CitiBike is ever going to make it uptown, or into Queens, or further into Brooklyn, it’s going to need some kind of public funds.

The problem is that the CitiBike scheme was not designed to be expanded with public funds. As a result, there’s no easy or obvious way for City Hall to strengthen and expand NYC’s bike-share scheme, even if it wanted to.

A bit of background is in order. When NYC decided it was going to have a bike-share scheme, it basically had two choices. The first was to run the project itself. It would order the bikes and the docks, be in charge of maintaining both the hardware and the software, would run the sponsorship auctions, would collect the bike rental income, and so on and so forth. A whole new level of civic infrastructure would have to be put in place, a lot of new civil servants would have to be hired to run the project, and almost certainly the whole thing would cost a vast amount of money.

The alternative was to sign a contract with a private company — in this case, Alta Bicycle Share. Alta would have a contractual obligation to keep the bike network in good shape, and in return it would receive all the rental and sponsorship income, as well as make a profit for itself. The more successful the bike-share scheme, the more money Alta would make — so all interests were aligned.

That, then, is the way that NYC decided to do it. It signed a multi-year contract with Alta, and in turn Alta signed a multi-year contract with Citigroup, worth $41 million, to be the lead sponsor. Between Citi’s money, the rental income, and future sponsorship income from other sponsors, Alta was meant to make enough profit that it would make sense to expand the network.

It didn’t work out that way, in large part because of Alta’s incompetence and in small part because of Hurricane Sandy, which pushed back the launch of the bike-share scheme by about six months.

Mainly, though, the CitiBike scheme was just very badly designed, especially if it was ever going to make money. There were all manner of fiascoes surrounding the software for the project, which were all entirely Alta’s fault, and when CitiBike did finally launch in 2013, entire stations would frequently fail to either dispense or receive any bikes at all.

That problem ended up being fixed, and Alta even started getting a bit better at rebalancing — moving bikes from where they are too abundant (meaning no docks available to return bikes) to where they are too scarce (meaning no bikes available to rent).

But as CitiBike celebrates its first birthday, much deeper, and much more intractable, problems are easily apparent. The biggest is the reason why CitiBike is losing money: there aren’t nearly enough daily and weekly renters.

CitiBike was designed from the start to operate on a cross-subsidy model: the annual memberships would be underpriced, at $95 per year, with all the profits coming from the daily and weekly renters. Certainly CitiBike is losing money on me: I’ve taken over 250 trips in the past year, which works out to roughly 37 cents per trip.

Annual membership also makes it very easy to rent a bike (when the system is working): you just slide your key into the slot, and out comes the bike. I have lots of rides in the five-minute range: journeys I would otherwise have walked without any difficulty, but which are that much quicker and easier if I can just hop on a bike. Once you’re a member, the marginal cost of renting a bike is zero, which encourages one of the most common use cases — taking a CitiBike to a slightly further-away subway station which will get you to your destination more quickly. (“Instead of walking to the local, CitiBike to the express.”)

Daily and weekly rentals, by contrast, have none of that convenience. The docking stations come with tiny screens attached to unreliable credit card readers, and becoming a short-term CitiBike member is really hard work. Putting in your information takes loads of time and involves multiple screens; and even if you’ve managed to get to the end of that process, renting a bike then requires a whole separate set of steps, involving your credit card, dock numbers, and three-digit codes.

None of this is convenient; none of this is easy — even when it works, which, a lot of the time, it doesn’t. (CitiBike has been plagued with reports of failed credit-card transactions.) In most private-sector systems, the most profitable customers get the best service; in this one, they get the worst, with predictable results. I know some people who have actually bought spare annual CitiBike memberships for houseguests, just because they’ve learned that visitors will almost never jump through all the hoops necessary to take out a short-term rental.

The real problem, then, is not the pricing structure itself, so much as it is the station design, the user interface, and the general ease of use of the system. All of which was Alta’s problem to solve.

On top of that is another problem, which is the degree to which Citibank has become inextricably associated with CitiBike. That’s great for Citibank, of course, which got fantastic brand advertising at a very low cost. But it has also scared off any potential secondary sponsors. In principle, Alta can and should be selling lots of other sponsorships, perhaps in new geographic areas. But in practice, no one wants to sponsor the Citibikes, because they’re now completely associated with Citibank.

Put all of this together, and you end up with the fact that Alta is losing money on CitiBike. In itself, that’s not obviously a big problem. The beauty of outsourcing the bike-share contract to Alta was that Alta took all the risk: if it worked then Alta would make money, and if it didn’t work then Alta would lose money. If Alta loses money, that’s public-private capitalism at its finest.

Except. When Alta loses money, the quality of CitiBike starts to deteriorate. Obviously there’s not going to be any geographic expansion any time soon: doing that would only increase the total losses. But also, the existing infrastructure has started getting visibly worse. Bikes are often pretty shabby, one year in: they have flat tires, or the screw fixing the seat post will be broken, or (and this is almost universal now) the seats themselves will be cracked, causing them to stay decidedly soggy for hours after any rainstorm. (When new, the seats are wonderful things: they dry off almost immediately.)

The docks, too, are in pretty bad shape: most of them had great difficulty coping with the brutal winter, and far too many of them, at any given point, refuse to accept bikes, or to dispense the bikes that are in them.

None of this is allowed, under the terms of Alta’s contract. Alta’s contractual obligations are spelled out in great detail: 99% of broken docks must be repaired within 48 hours; every bike must receive a mechanical check once a month; 98% of vandalized bikes and docking stations must be cleaned within four days, and so on.

Alta isn’t even coming close to meeting those obligations. In December, for instance, only 34% of bikes were examined; in January, only 50% of broken docks were repaired within 48 hours. It’s reached the point at which Alta now boasts of doing things once a year which it’s contractually obliged to do continuously throughout the year.

Which brings me to Ecuador, and sovereign debt more generally. When a country like Ecuador borrows money, it has a contractual obligation to repay that money in full and on time. But it also has a large degree of de facto impunity: because it’s a sovereign nation, that contract is worth precious little to an Ecuadorean creditor if the country ever stops paying its debts.

Alta, in New York, is a bit like Ecuador: it has a contract, but its counterparties have precious little recourse when that contract is violated. It really doesn’t matter whether the reasons for the violation are weather-related (El Niño, Hurricane Sandy); either way, you end up staring at your contract and wondering what good it serves. To put it another way, a contract, no matter how detailed, has little more weight than a verbal promise, if there is no real recourse or remedy in the event that the terms of the contract are broken.

Alta signed its contract when it hoped and expected to be making good money from CitiBike. With that money, it could keep the whole system in good repair, and even expand it to new neighborhoods. But now that it’s losing money, it’s cutting back on everything, and the whole system is getting worse rather than better, with no real hope of improvement.

Meanwhile, NYC is locked into its decision to grant the CitiBike contract to Alta. Probably there’s some mechanism, under the contract, for the city to remove the contract from Alta and grant it to someone else, or to take it over and run it on a municipal basis. But this is Alta’s proprietary system: no one else can run it, nor would anybody else want to run it, given how badly it is designed.

New York is, realistically, stuck with Alta — which means that it doesn’t have much in the way of attractive options. There is certainly no legal or moral reason why New Yorkers should cover Alta’s losses, which were almost entirely of Alta’s doing. But maybe there’s a practical reason: if we cover their losses, perhaps Alta will step up its game and actually fulfill its contractual responsibilities?

Well, perhaps. But then again, perhaps not. Alta has already discovered that it has impunity: if it fails to keep up its end of the bargain, the worst that happens is that it gets shouted at by a bunch of civil servants. And so long as it’s happy being on the receiving end of the occasional verbal bollocking, every dollar it doesn’t spend on maintaining an excellent bike-share system is an extra dollar in its owners’ pockets. That’s true now, and that would continue to be true even if New York City started writing extra checks to Alta in an attempt to make things better.

Similarly, New York could step in and promise to cover the extra marginal cost of expanding the CitiBike geographical footprint. But that wouldn’t fix the deeper issues with the system — it would only entrench them further. After all, if tourists aren’t signing up for CitiBike in midtown, they’re not going to start signing up if it expands into Long Island City.

What CitiBike really needs is a major operational overhaul. Management should be replaced with people who can keep their promises, and the means of renting bikes for a day or a week should be radically simplified — maybe booths can be set up in hotels and next to the bike stations most heavily trafficked by tourists, which rent out CitiBike keys somehow. A system where you rent a key, and can then return it to any booth, and where the booths are staffed by human beings, might well be more successful than the current model.

It’s not clear that Alta is the best company to give the job: it has, after all, signally failed to meet any of its original goals and promises. At the same time, it’s equally unclear that anybody else would be better than Alta at operating Alta’s own system. But one thing is clear: if we’re going to start throwing public money at bike share, we should definitely demand some changes to the way things are (not) being done right now.