Wednesday, August 14, 2013

Bikeshare budgets, how do they work?

I can now tell you that I'm a proud Citibike annual member, and I've been enjoying it. The biggest advantage is being able to use all that great Manhattan bike infrastructure without having to go through the inadequate bike infrastructure to get to it. The Ninth Avenue bike lane was installed in 2007, and I still haven't used it, because I haven't wanted to ride across from the Queensboro Bridge on 55th Street. With Citibike, I can take the subway to the West Side, or to Greenwich Village where the streets are calmer. I still haven't ridden the Ninth Avenue lane, but I have ridden the Eighth Avenue one.
It would be nice to have Citibike here in Queens, and in nearby parts of Brooklyn, and I'm glad that my City Councilmember Jimmy Van Bramer and State Senator Michael Gianaris are lobbying for it. But there's one thing I don't understand: where the money comes from. I've read a lot about potential Citibike expansion, and everyone just repeats the claim that the expansion will have to be paid for by the government. There is no explanation; it's just treated as though it's obvious, and then people move on to the question of where the government will get the money.

It's not obvious to me, though. Imagine a private bike rental business with five locations. The business can expand without money from the government. All it has to do is earn a surplus that the owner reinvests in new locations. The owner can even make a bet on future success by taking out a loan to pay for expansion. If Citibike can earn a surplus, it can do it too.

Is Citibike earning a surplus? I haven't seen anything one way or another. There are three main possibilities. It could be running a deficit and burning through the initial Citibank outlay of $41 million plus the Mastercard $6.5 million. It could be earning a surplus, but not enough to expand at any significant rate. Maybe it's not at a surplus yet. Or maybe the surplus is going to something other than expansion.

It turns out that we can actually estimate quite a bit. We know, from the Citibike website, that as of Sunday there were 69,830 annual memberships. The rate at which new people are joining is constantly dropping, as is standard, but the system may get up to a hundred thousand members a year, bringing in $9.5 million. There are about 1500 24-hour passes and 150 seven-day passes sold per day on average, earning $18,750 per day, which will come out to about another six million dollars a year, allowing for weather conditions. So the total membership income for the year will probably be around $15 million, which dwarfs the $10 million per year that the city gets from its sponsors.

So what are the expenses? According to this article, Bixi costs $400,000 Canadian a year to run 1800 bikes in Toronto, for an average of $222 per bike. Everything's more expensive here, so let's say $1.5 million a year for our 6,000 bikes. That means that we could pay for the the system out of 24-hour passes, or that it broke even with annual memberships before it even launched. Or about $23 million in profit, which is split between the city and Alta, leaving $11.5 million a year for expansion.

(Interestingly, this means that we don't actually need sponsorship; even without it, the city would still be on track to earn $6 million a year from the deal.)

The next question is how much expansion we can get for $11.5 million. Alta got $47.5 million from the sponsors, a $42 million loan from the vampire squid, and $5 million from its insurance company, for a total of $92.5 million. But a lot of the equipment was damaged by Hurricane Sandy, and it's not clear how much the bikes currently in use cost.

That said, before Sandy hit Alta told its insurance company that it had $20 million in equipment on the ground, so let's assume that that was for 7,000 bikes. That means that for $11.5 million we could expand the system by more than half its planned launch size - 3,500 bikes - every year.

To me that suggests that by this time next year I could be riding Citibikes from the Upper West Side to Long Island City to Bed-Stuy to Red Hook. In 2015 I could ride from my house to Tremont to Inwood, and south to Ridgewood and Brownsville. In 2018, who knows?

Feel free to go over my "back of the envelope" and point out anything that doesn't look right. But if I'm right, we don't need government money to expand Citibike. We only need it if we want to speed up the process. And you know, if we're spending $800 million to widen a bridge that carries hardly any transit, I have to wonder if we couldn't find a hundred million for Citibike expansion. Imagine what that would get us.


justforcomments just said...

The numbers I've seen on yearly costs per bike are generally closer to $1800/yr/bike.

Tal F said...

In any large venture such as citi-bike, it is rare that expansions are funded purely by reinvesting profits. Rather, if an enterprise can demonstrate that a project's expected payoff is more than sufficient to pay off a loan at the terms currently available in the market, then the simple solution is not to wait for profits to come in and take out a loan to undertake the investment immediately. Otherwise, you're leaving money on the table by waiting around until enough profits come in to fund expansion to the optimal level. I, too, have done the sort of back-of-the-envelope calculation you've undertaken here, and I really wonder why the city doesn't just borrow the money for expansion, say from the muni bond market, backed only by their share of future revenues. Given how much more popular citi-bike has been (per bike) than other bike share systems, despite charging more than any other city, it must be hugely profitable even after adjusting for the higher costs of operating in NYC.

James Sinclair said...

A few things to throw in:

A standard station, ie 15 docks (small for Manhattan, average elsewhere) goes for around $40,000 each. Bikes are about 1k each.

You have to factor large startup costs - the warehouse, the vans, the tools, the training and hiring etc. Cost of insurance as well.

MRB said...

I think you are overestimating the number of one-day and seven-day passes will be sold over the winter. Citibike has been in since late May, so the past 3 months have been the best biking weather of the year. Your estimate assumes that demand will be at summertime levels for 320 days/year. That is too many.

My guesstimation is that pass sales in Dec/Jan/Feb will be roughly 10% of summer levels; shoulder periods in Nov and March at 40% and Apr/Oct at 75% (and August at 90%).

All this adds up to about $4.2m in short-term pass sales, about 1/3rd less than you $6m projection. This isn't accounting for the novelty factor of year 1 (however, because bike availability has been an issue, perhaps summertime demand is saturated enough that it washes out)

From a business standpoint, I would probably only subsidise expansion via pass sales - at $1800/year in maintenance costs per bike, and each bike supporting (at current levels) 10 users, each yearly pass runs a deficit of $80.

Justin Pollock said...

Your cocktail napkin math is a bit off as mentioned in other comments. Besides the aggressive estimates for yearly fees from daily users and annual members you way underestimate the annual costs of operation per bike.

Also, where did the $20 million in equipment come from? I assume it was paid for by the citibank/MC deal which is not accounted for in your math.

I'm all for expansion and hate the veil of secrecy over citibike's operations but this kind of math doesn't get us anywhere.

Unknown said...

You are not including overage charges. Citi Bike doesn't publish them, but I think they are considerable. I've seen people on Citi Bikes more than 5 miles from the nearest station; I suspect they can't make it back in time for a free ride, even if they are annual members with a 45-min limit. And if you have a day/week pass, you can easily go over 30 min even for a reasonable-looking one-way ride between midtown and downtown (unless you go to the trouble of stopping at a station in-between).

On the other hand, I think you may be underestimating expenses. Again, there is no public information, but can you really pay all those mechanics, drivers, software developers, technical support, management, etc. for 1.5 million? And that doesn't even include material expenses such as gas, parts, depreciation, debt service, rent, etc.

Capn Transit said...

MRB, I did knock the intake from daily and weekly memberships down from $6.8m to $6m. I don't see the daily memberships going all the way down to 10% of summer levels in the winter. Think about global warming. Think about the fact that there are always a number of mild days in the winter, which are perfect for daily memberships.

All you wise guys repeating the $1800/yr per bike maintenance cost figure, do you really think so? Why would it cost eight times more than Toronto?

Justin, you do get the distinction between capital and operating costs, right?

Capn Transit said...

Forgot to mention that even though there are lots of Citibike annual members like myself who don't live within the current coverage area, every expansion will pick up a bunch of new annual members who see value in the system that wasn't there before. I think that will outweigh any attrition due to the novelty wearing off.

SayWhat said...

This is the most slapdash analysis I can imagine- did you do any research on this other than reading a few magazine articles (that you may have mis-construed in the translation from Canadian- no link to that to review...)? The capital costs for the original 7000 bike system would have been about $42M (~$6K/bike to deploy); per the article you link to, this was covered by the loan & is being paid off over time by sponsor $$, so no surplus there. They lost about 1000 bikes in Sandy (the reason that there are <6000 bikes on the street at deployment), which is close to the $5M the insurance company gave them (BTW- the operator is suing the insu co for a lot more, likely delaying expansion further). Operating costs typically run closer to $300+/month/bike (see Cabi contract info avail online), not per year, so operations will be in the $18M+/year range w/the current fleet & higher if they do expand. If you think casual usage (& overall usage) will remain the same in the winter, please reference other systems in weather dependent climes and you'll see a steep dropoff when it gets cold; aside from the even newer Chi program (same system, same operator, same pending issue), NYC is the farthest north of any year-round bike share currently operating in the US, so seasonal attrition will likely be even greater (again, Cabi info from much warmer DC is available online). So, no extra $ from startup (& probably a deficit w/the delays from software probs & Sandy), annual membership revenue of ~$10M, short-term revenues of <$6M, plus some decent overage charges, and you might be breaking even or slightly better on operations (even $1M in excess cash would be decent income for running the system); but there will almost certainly be no substantial money left over for further capital expansion (which is why the city is talking about funding the next phase). I've never seen a bikeshare program that has been able to fund initial &/or replacement equipment from revenues (but hey, roads & transit don't cover capital costs from user fees either, but that's a separate discussion on the best use of public funds)- the only other large scale, privately funded system (Deco Bike, in ideal weather Miami Beach) claims to be doing so, but little public info on the finances is available. Please try to understand what you are talking about before just throwing out some ill conceived and wildly inaccurate postulations- it does no one any good & makes me waste my time rebutting it in frustration.

Capn Transit said...

Oh, please get a life, SayWhat! Yes, I said it was a back of the envelope calculation, based on what I could find with some searching. That's because nobody's sharing this information. Even you don't provide any links to back up the numbers you throw around, not even a link to the Cabi contract that you say is online.

If you're saying that Citibike may be able to break even on fees alone, then the $10m per year sponsorship ought to pay for some expansion.

The fact is that the budget is not available online, and we're being asked to transfer money from other transportation improvements (bus lanes, sidewalks, bike lanes) to pay for the expansion, without being given any details.

Maybe there's a select few in the know who have access to this information, but then they should just tell us and stop using it to scold us like we're children.