Quantifying Capital Bikeshare’s Supposed “Success”

by Marc Scribner on September 27, 2011 · 23 comments

in Features, Mobility

Post image for Quantifying Capital Bikeshare’s Supposed “Success”

Last week, D.C.’s Capital Bikeshare program celebrated its millionth trip and one-year anniversary. U.S. Secretary of Transportation Ray LaHood celebrated the milestone with a blog post, calling the government-subsidized bicycle system “remarkable.” Capital Bikeshare is supposedly revolutionizing transportation in the District, reducing nasty polluting and obesity-causing auto use. I will now explain why this is not the case with some back-of-the-envelope arithmetic, and why Capital Bikeshare should be panned rather than praised.

Let’s look at some data. From January 2011 to July 2011 (the most recent complete data available), Capital Bikeshare averaged 3.35 trips per in-service bike per day in Washington, D.C. (This is overestimated given that DDOT/Capital Bikeshare do not break down their active fleet by municipality as they do for ridership, but I’m feeling charitable today.) According to the most recent National Household Travel Survey (2009), Washington, D.C., averages 1.92 million person trips per day. This can be broken down into modes: 44.3 percent of trips were made by personal automobile, 18.4 percent by transit (rail and bus), 1.9 percent by taxi, 1.6 percent by bike, and 27.9 percent by walking. This means that a few years ago, Washingtonians made 11.5 million trips annually by bicycle.

Let me again be charitable and assume that not only did Capital Bikeshare add 1 million trips (bringing annual bicycle trips to 12.5 million), D.C. cyclists using their own vehicles (myself included) added another 1.3 million additional bicycle trips — bringing annual cycling trips to 13.8 million — while holding total trips constant. We’ve now increased cycling’s share of person trips by 20 percent from the 2009 NHTS. Even given these very optimistic assumptions, cycling would only represent a little over 1.9 percent of all trips taken in Washington, D.C. (Capital Bikeshare’s mode-share in this scenario: 0.14 percent of all trips.) It would take a pretty twisted view of reality to herald this figure as somehow revolutionizing urban mobility.

But where did these bike trips come from? Presumably, not every new trip was added by a resident that was previously stationary. No good survey data exist on mode-switching, but a paper presented at the previous annual Transportation Research Board meeting can help shed some light on how these people may have previously got around the city. The authors studied Montreal’s BIXI bicycle-share program, widely touted as a success by cycling advocates. According to their research, of BIXI’s approximately 17,000 daily trips, about 340 replaced personal auto trips and around 1,360 replaced taxi trips. The rest came from people who would have either walked, used existing bus and rail transit, or used their own bikes.

So if 90 percent of the 1 million trips made by Capital Bikeshare came from previous users of “green” transportation modes, why should we be so excited that the District Department of Transportation spent millions of dollars subsidizing bicycle use for mostly white-collar downtown office workers? Much of the initial Capital Bikeshare funding came from the Federal Highway Administration’s Congestion Mitigation and Air Quality Improvement Program (CMAQ) grants. If the money was to be spent on “[auto] congestion mitigation” and “air quality improvement,” and assuming the BIXI mode-switching figures broadly hold in D.C., this was a clear misuse of federal funds. (Here’s FHWA’s CMAQ fact sheet.)

The Capital Bikeshare bikes cost around $1,000 a piece and have a life cycle of six years. Annual operating costs are somewhere closer to $2,000 per bike. In the past two years, I have spent approximately $500 on my personal bike that I commute to work on daily — $250 a year. And I average more trips per day and distance per trip than Capital Bikeshare. The program’s costs given the benefits are simply absurd.

If the District’s transportation elite must maintain their warped set of priorities and subsidize cycling to induce ridership, why not instead offer vouchers to partially cover the initial purchase and annual maintenance of a bike? It would certainly be cheaper, although that would take the look-at-the-shiny-objects-we-wasted-tax-dollars-on fun out of the whole thing.

Dave September 27, 2011 at 4:53 pm

I’m waiting for you quantify bikeshare. How much money did it actually cost? unknown? what were the benefits? What is the value of those benefits? How much would the program be worth?

Nada from you. If you’re going to quantify something, quantify it. You said you would explain why it is not “revolutionizing transportation in the District, reducing nasty polluting and obesity-causing auto use [sic']” but you didn’t. Did it change transportation in the District with a new from of transit? Check. Did it reduce pollution? By your own admission, yes. Did it reduce auto use? By your own admission, yes.

So, I’ll give you a do-over.

Marc Scribner September 27, 2011 at 5:14 pm

Confused again by you, Dave. The primary point of this is to argue that Capital Bikeshare is a trivial mode and that people are getting excited over nothing. Secondary points include that it most certainly reduces auto use (and thus negative externalities of auto use) by a trivial amount, that it only exists because of a multi-million dollar federal grant, and that it costs far too much money to achieve whatever the bikeshare people claim it has achieved. Those points are all in there.

Denis September 27, 2011 at 7:32 pm

Why should we be so excited? Because, even though it does not replace so many car trips, it cuts transit times from 30 minutes to less than 10 minutes on a typical trip. This is by far the most conveniant means of transport in Montreal.

Marc Scribner September 27, 2011 at 8:18 pm


I won’t quibble with you there. That’s why I own a bike, but I don’t expect my neighbors — particularly those who can’t bike — to pay for me. Beneficiaries of D.C.’s Capital Bikeshare program are quite satisfied paying 50 percent of the cost. But as is common knowledge (at least in my weird transportation policy world), transportation benefits accrue almost exclusively to the users.

cabi addict September 27, 2011 at 10:56 pm

Autumn is in the air and Marc is at it again. One year and a million bikeshare trips ago in http://www.openmarket.org/2010/10/04/sharing-isnt-caring/ he predicted that Capital Bikeshare would experience large scale theft or damage to equipment and would incur massive losses; that the program would not produce public benefits; and that the system would turn out to be an administrative nightmare.

Now he claims it’s a failure because his partial usage figures (starting during the seasonal winter downturn – remember the ice storms? – and excluding the most recent months of growth in subscribers and in usage) are too low.

Private sector retail figures normally exclude stores that have been open for less than a year; Capital Bikeshare is less than a week out of its first year and is still building out the system; trips/bike/day are certainly higher than they were at launch and than the number cited above.

As transportation infrastructure investments go the Bikeshare system is dirt cheap; a fully stocked bike station costs the same as a single downtown parking spot and can hold more vehicles and produce more trips. Out in the ‘burbs the equivalent is three park-n-ride lot spaces. Add up everything that’s been spent on this project so far and disregard the user fee revenue that’s come in – how much infrastructure would that buy?

Marc doesn’t have to like the system, and I expect him to continue to put out these pieces in this space and anywhere else he can. His political outlook aside, the one place where I’m not sure that he Gets It is the underlying view of transportation. He sees it as vehicles and suggests we would be better off distributing bike vouchers. I see transportation as a service – the whole point of bikeshare is that between trips the subscriber need not be concerned with custody of the vehicle, and the vehicle need not be idle until its last user retrieves it from where it was left. By increasing both utility and utilization bikeshare enables trips to occur that would not otherwise be possible.

So let’s come back to this in another year and see where things stand. Personally I would like to see an expanded system with usage so high that Marc won’t have a way to be disappointed.

Dave September 27, 2011 at 11:35 pm

First of all, your neighbors do pay for you to bike, unless you’re paying some secret bicycle tax that no one is aware of that pays for roads, bike parking, traffic lights, trails etc…

Second, the title is about quantifying success, which you don’t do.

Third, people are getting excited over a new form of transit that pays for more than half it’s cost and that, in the first year, has improved 1 million trips. Ridership is growing, the system is expanding and so that is going to go up. To say that it makes up a small number of trips and is thus unworthy of excitement is to miss the whole point. TDs make up a small number of plays in a football game, but they’re pretty exciting. I’d say CaBi is a TD.

Marc Scribner September 28, 2011 at 10:58 am

cabi addict,

1. I never “predicted that Capital Bikeshare would experience large scale theft or damage to equipment.” I did predict that the system wouldn’t leave the red — which it hasn’t. It hasn’t produced significant public benefits, although it has produced significant PRIVATE benefits for a tiny fraction of city residents at PUBLIC expense. I will concede I was wrong about administrative hurdles being significant barriers to roll out. The transportation wizards have decided that this is a priority program for some reason.

2. I used the most recent complete public data provided by Capital Bikeshare and DDOT.

3. The number cited above overestimates the true average of trips/bike/day during that period, as I noted. CaBi/DDOT do not segregate fleet operation data by municipality as they do with use, so I had to use Washington, D.C. only for usage, but system-wide for the operating fleet.

4. Fiscally, cost per trip is not very good. I pointed this out to Dave over on another site. Auto/road transit person trips are actually (very slightly) less costly than Capital Bikeshare trips according to NHTS 2009 (well, they’re within the survey’s margin of error).

5. The TRB paper I cited found that most bicycle-share users are not substituting auto trips or creating new trips. They’re substituting trips made by existing transit, personal bike, and walking. You can claim whatever you want, but there isn’t much evidence supporting your claims.


Really? My cost to the transportation system from riding my bike is close to zero. But take that cost and then add the Capital Bikeshare costs, and that’s what CaBi users impose on the system.

I attempted to quantify the success of the system by comparing mode share estimates. The media and numerically-challenged bicycle proponents were throwing around the 1 million figure as if that made some sort of dent in the transportation system. I pointed out that this was stupid and completely useless without context (i.e., how many trips are made by other modes). What metric would you use? The number of smiles you see on Capital Bikeshare users as they wobble down the bike lanes in Dupont?

JTS September 28, 2011 at 11:57 am

>What metric would you use? The number of smiles you see on Capital Bikeshare users as they wobble down the bike lanes in Dupont?

Absolutely. The as-yet intangible benefits of a money losing system like Cabi have yet to be quantified. Since the city can’t yet realistically compete with the outer areas on a number of economic fronts (COL, taxes, labor, etc.), quality of life is going to be the single most important driver for the city’s competitiveness in the years to come.

Cabi may lose money, or be a relatively small number of total trips, but damn if it hasn’t become instantly synonymous with our fair city. Cities don’t buy bikesharing systems because they think it’s going to take 20 million vehicles off the road; they buy them because they are a very cheap way to improve quality of life for the current and future tax base. Who else is going to pay for all of the improvements the city needs?

I want to live in DC because I can ride a bike share to places where I can spend money. I don’t want to live in Gaithersburg because I have to drive. Are you really trying to say that “The number of smiles you see on Capital Bikeshare users as they wobble down the bike lanes in Dupont” is not something that will eventually be quantified with respect to consumer behavior? To real estate trends? To the tax base? You’ve got to be kidding me.

JTS September 28, 2011 at 12:01 pm

And just as an add-on, the difference between a system like cabi and a direct transfer payment is that the transfer payment doesn’t have the ancillary benefit of giving the city a means to brand itself. Cabi is synonymous with DC’s new identity as an attractive place. Giving everyone 1k to buy a bike may have the same end-result of more people riding bicycles, but it wouldn’t give the city 1,500 quirky looking red bikes that match the circulator system. It’s pretty cool, and people like that.

Dave September 28, 2011 at 1:00 pm

According to DDOT, they made $300,000 on the program this year – that’s without the positive externalities.

What metric would I use? I don’t know, how about cost vs benefit? (again the program made money, so unless the benefit was negative, that’s a winner). How about the recovery ratio? How about comparing it’s use per dollar to use per dollar of other systems? There are plenty of ways to look at it other than saying it made an incremental change in mode share without the context of how little it cost.

Denis September 28, 2011 at 1:22 pm

You don’t understand my point. According to you, Bike Sharing is all bad and you select only the facts that suit your argument. I say that bike sharing has improved a lot the way to get around in cities for the people who use it and that it is a good thing for that reason. It is not true that if you give a bike to everybody who use bike sharing systems, they would used it. For reasons that have been given so many times, owning a bike is not convenient for some people and even for some people who own bikes, bike sharing can be used in many situations where a personal bike cannot.
I can accept that you are against grants for such initiatives even if I don’t agree with you. But this is not a reason to be so biased. For exemple, you say that an increase of 1 million of trips over 12 is small. Apart from being a questionable affirmation, this ignores the fact that Capital Bikeshare is a small system covering only a part of the city, that is it relatively new and that this number could increase with time especially if the system was expanded. In order for this information to be meaningful, you should compare CaBi usage with total bicycle usage in the area covered. You say that the operating cost is $2000 per bike but you say nothing about revenues and you don’t consider that the system is only starting and that the revenues could grow with time. So when you conclude that costs over benifits are absurd, you base yourself on a partial assessment of both the costs and benefits.

Marc Scribner September 28, 2011 at 1:29 pm

Quality of life for which residents? Those who make up the tiny population using the mode that accounts for 1/10th of 1 percent of urban trips? I don’t think most non-CaBi users in D.C. would agree that quality of life has been, or think it will be, increased by the program. And no, I do not believe trivial CaBi will ever result in statistically significant changes to real estate prices or retail sales. If you’re trying to undermine the limited success of TOD across the country, D.C. being an exception to the “fail” rule, redirecting funds that could otherwise be spent on legitimate transit to bicycle sharing would be a good way to accomplish that goal.

And CaBi is “synonymous with D.C.’s new identity as an attractive place”? Really? I thought that was the H Street trolley construction zone… To me, it represents D.C.’s decades-old identity as a municipality run by spendthrift lefty goofballs…

andrew September 28, 2011 at 3:14 pm

Although I appreciate the analysis here, I do think that it’s extremely prudent to point out that Capital Bikeshare is currently covering all of its ongoing operating costs from user fees.

Although it’s true that the up-front cost of the underlying infrastructure is unlikely to be recovered, this is true of virtually all transportation modes (including privately-owned cars). The fact that Capital Bikeshare is managing to cover its ongoing operating expenses at a modest, but healthy profit is nothing short of remarkable. (And, let’s also be honest with ourselves here — as far as transportation spending goes, Capital Bikeshare has thus far been a tiny drop in the bucket. It’s not even a remotely expensive program when you consider what most transportation infrastructure costs)

And, as you mentioned, it’s a good deal for the people who use it. Although I spend less than $250/year to maintain my personal bicycle, I still do spend more than the $75 Bikeshare membership fee.

Bikeshare’s flexibility is also worth noting. Although the overall share of Bikeshare trips is indeed very low, it does a fantastic job of “filling in the gaps” for trips where there is no (or infrequent/indirect) mass transit service, and where parking is not readily available at one or both ends. For instance, I can bike from NoMA to the Logan Circle area in about 10-15 minutes. Previously, I’d need to choose between transferring between two unreliable bus lines; paying $10 for parking; or taking Metro, transferring, and then walking 15 minutes to my final destination.

Bikeshare’s advantage on these routes is magnified during “off peak” hours when bus/Metro service is less frequent (and these days, it’s been getting even more infrequent/erratic). Re-using NoMA as a start point, one can easily bike to Eastern Market, which is an impractical trip via Metro, but a fairly quick bus ride . However, if the next bus isn’t coming for another 10-15 minutes, you can hop on a bike, and be at Eastern Market before the bus even pulls up into NoMA.

For these oddball short-distance trips in or near the downtown core, Bikeshare is absolutely the fastest (and cheapest) way to get from your origin to your destination. Better still, it prevents the core from becoming clogged with parked bicycles, and users don’t need to worry about carting their bike around during the course of the day. Unlike riding your own bicycle, you don’t have to worry about making sure that every trip is a round-trip.

CBGB September 28, 2011 at 4:00 pm

Does that 1.92m figure include commuters coming into the city? Because I would expect those trips to be overwhelmingly by auto and to a lesser degree transit. The story is much different for DC residents who are much less likely to drive to the corner store, dry cleaner, bar, dinner, etc.

Also what qualifies as a trip? Someone may commute to work by car or transit but bike or walk to the amenities in their neighborhood. If I go out to lunch is that two trips? What if the restaurant is across the street or three blocks away? Quantifying trips in dense urban environments is inherently a difficult approach.

Marc Scribner September 28, 2011 at 4:45 pm


No. It only included D.C. households who were subject to the NHTS. Trips are defined as going from one address to another, which is why walking is at a 27.9 percent share. I used the annualized Travel Day data rather than the Travel Period data to weed out long-distance trips that wouldn’t be relevant to urban mobility of urban residents. Of course, these should only be treated as estimates. The NHTS data are certainly of lower quality than Capital Bikeshare data.

Bryan September 29, 2011 at 5:55 pm


Like Andrew said, what about the numbers from the GGW article published one day after you. http://greatergreaterwashington.org/post/12176/capital-bikeshares-first-year-results-exceed-expectations/
You said they wouldn’t be in the black, but they did pull in a profit for DDOT after just one year. Sure it’s not like they’re paying back federal funds, but they were never required to, and I’m a taxpayer, who like many is GLAD that those federal funds went to bikes, and to a private/public partnership like Alta/DDOT, and towards generating revenue for the city.

Cabi is self sustaining, and New York is signed on. The battle is over. Now enjoy the ride.


Marc Scribner September 30, 2011 at 5:18 pm

If by “self sustaining,” you mean “reliant on federal grant money,” you are correct. As to the benefits, a proper comparison would be between rates of return on investment to CaBi and alternatives. The GGW people do not seem to realize there is a difference between economic (opportunity) costs and fiscal accounting costs.

Bryan October 2, 2011 at 12:39 pm

I appreciate that you changed your dismissive initial reply with something less personal and more substantial. The whole point is that it is no longer “reliant”. It was, in order to get built (like many federal projects, some of which even most libertarians agree were a good idea). They continue to request expansion funds from the CMAQ (Congestion Mitigation and Air Quality Importment) Fund, because the money is there, so why the heck not request it. But they don’t have to, so they aren’t reliant on anyone. DC made $300,000, and could spend that operating profit to build 6(?) stations if it wanted and execute a slow independent expansion. But CMAQ exists, so they use it.
But you want to compare, so let’s compare. They made 2.27 million in revenue, and after expenses, $300,000 -ish was an operative profit. Comparing just operations, that’s 13% profit.
But you want to include initial investment, so let’s do that as well. DDOT claims that the program is a 6M dollar program here (http://www.dc.gov/DC/DDOT/About+DDOT/News+Room/Capital+Bikeshare,+Largest+Bikeshare+Program+in+the+United+States).
But they also say that DDOT payed for 20%, and that Virginia didn’t use federal sources at all. So the initial capital from the CMAQ fund to D.C. was 4.8 million, roughly, if my assumptions are correct and to be honest I’m not sure, but here goes.
That brings the total first year cost to the District (I’m now ignoring Arlington if they didn’t get federal funds, as libertarians do believe in state’s rights to spend state funds on bikes however they wish) to 2.27M + 4.8M – .3M = 6.77M. Divide the 300K by 6.77M and we get 4.4%. So there’s your basis for an apples to apples comparison. An Alternative measure would be to only calculate the operating profit / initial federal expenses (300K / 4.8M = 6.25%). Or lastly, full first year costs to profit (6M + 2.27M – 300K = 7.97M) 300K/7.97M = 3.76%, but this measure includes DC’s local, liberty filled, contribution, and excellent choice of Alta to produce a profit, so I prefer the 4.4% figure.
Either way you pick, you have a system that in its very first year made 3.76%, 4.4% profit, 6.25% profit, or 13% profit in a public/private partnership arrangement that created steady jobs and provided a give and take benefit to the taxpaying public and businesses.
After a single year of operation, and ignoring all non-cash positive externalities (I ignored all that hippy green crap and potential business benefits and kept only the cold hard money), what other transportation solution can you name that hit a 4% profit after just one year, again, ignoring ALL externalities? At this point, investing in CaBi gets a better ROI than a high yield savings account ever did, if those even exist anymore.
And it’s important to reiterate, this program will very probably never cost the city a single UNBUDGETED dollar again. If DC wants to allocate more money, they can, but if they don’t CaBi will actually start to raise DC money from now on. Since DC’s initial investment was 1.2M, it will actually be entirely profitable for the city in only 3 more years, at the current pace.
I’m not a transportation expert and I’m genuinely curious. Apples to Apples: 4% on full first year cost to first year cost, Oranges to Oranges: 6.25% on initial cost to initial cost, or Grapefruits to Grapefruits: 13% on annual operation cost, after one year? What was it? How much was it? And did it ever pay itself off?

Steve O October 3, 2011 at 5:00 pm


Marc Scribner October 2, 2011 at 4:22 pm

Your attempt at this quasi-CBA ignores key components of a proper calculation of ROI. First, you omit a life-cycle cost analysis. A single data point cannot tell you much, especially when it is deeply misleading. Second, what is the impact on congestion? Assuming the BIXI figures hold, this would be negligible, meaning that travel time improvements are trivial. Third, what sort of economic growth did this program foster? Since you chose to dismiss the network effects (positive externalities) that result from road capacity enhancements (in addition to non-work travel and commuting, roads move trillions of dollars worth of freight annually in the United States, as well as bicycle-share bikes), these are ignored. All of these components are standard in transport project CBAs.

I’m also still confused why you choose to ignore costs simply because they don’t show up on DDOT or CaBi’s balance sheet. The “$300k profit” is false since it fails to consider total costs (simply because another entity is paying for capital costs does not mean they don’t exist). Someone is bearing these costs, right? This is the old “Amtrak is turning a profit because we only consider operating costs” trick. They don’t disappear just because fiscal accounting is balkanized. It’s fairly easy to secure accounting profits if you’re heavily subsidized.

Dave October 7, 2011 at 1:10 am

But CaBi isn’t a business. If CaBi gets ad revenue and continued growth, it could probably capture enough money to cover the start up costs – but they won’t try to, because it’s not a loan they have to pay back. CaBi won’t try to make a lot of money, because that’s not it’s goal. They could make more money by raising rates, just as the government could make more money on roads by raising the gas tax. But they’d get fewer trips. And this is a DDOT enterprise, not a business. DDOT is interested in moving people, not making money.

So you’re calling it a failure because it doesn’t achieve goals that aren’t it’s goals. That’s hardly fair.

“a proper comparison would be between rates of return on investment to CaBi and alternatives”

Excellent point. Why don’t you – the transportation economist – do that. Tell me how to mode shift 1 million people a year for a ~$5M investment. $5M is a trivial amount of money.

By the way, the Bixi data is almost surely not applicable. They’re stations are much closer together and cover a smaller area – so naturally they replace more walking trips. CaBi’s average trip is 1.7 miles. No one walks that far regularly.

Mike October 14, 2011 at 5:30 pm

I ride everyday on the capital bike share bikes to work and I really like it. But I would much rather just ride my own bike than have the government shell out tax dollars to support it. If it can’t support itself then it shouldn’t exist.

Comments on this entry are closed.

{ 1 trackback }

Previous post:

Next post: