For those not already familiar with the concept, bike share consists of short-term rental bikes located at kiosks spaced densely throughout a city. If you’re a subscriber, you walk up to a kiosk, swipe or tap a card, grab a bike, and go on your way. You can then return the bike to any of the other kiosks located around town.
The pricing scheme is typically designed to encourage short one-way trips. For subscribers, the first 30 minutes of travel is free, with per-hour charges increasing steeply after that. For example, in the upcoming CitiBike program in NYC, a five-hour bike ride will set you back $80! Typically both annual (for locals) and one- or several-day (tourist) subscriptions are available.
In this post I investigate the feasibility of bringing bike share to Santa Cruz. Santa Cruz is fairly well-situated for a bike share program. For a medium-sized city, it is relatively dense: with 53,000 residents over ~9000 acres in the core of the city. It has a nice pedestrian-oriented downtown. It attracts tons of mostly Californian tourists to its beaches, bluffs, and boardwalk, especially during the summer. It’s a friendly, sociable town. And it has a lot of college students, grad students, and other young people.
And I’m calling it Cruzer Bikes. The pun is there, waiting for you to take it, Santa Cruz!
I’ll map out three plausible bike share systems in detail: a feasible pilot program, a more aggressive pilot, and a scenario with a successful, mature bike share system. Then I try to estimate the costs and benefits of each system. All three networks require some subsidy beyond solely subscription revenue: this could be provided by public funds, advertising, sponsorship by businesses or business districts, or some combination of the above. However, the plausible (though optimistic) “mature” system provides significant benefits to the city (including uncaptured external benefits such as cheaper and quicker transportation for citizens and congestion relief): annual per-capita subscription revenues of $15 plus $15-$20 in external benefits, as compared to an annualized cost of $16 to operate the system.
Plan A: Pilot
The focus of this pilot system is a dense network of bike kiosks serving downtown and the beach. There are 16 stations: 9 large stations with 13 bikes and 22 docks each, and 7 small stations with 4 bikes and 8 docks. This adds up to 145 bicycles. (Note: all of the systems proposed here are composed of large and small kiosks, as defined above. These kiosk sizes were basically chosen for expediency — it’s easier to work with 2 types of stations than with 4, and I could find construction cost estimates for these.) The kiosks are spaced an average of 350 meters apart, fairly close to the common rule of thumb of 300 meter spacing. On the embedded Google maps below, markers with and without a dot correspond to large and small kiosks, respectively.
This pilot provides access to a significant fraction of the shops and restaurants in the city — for instance, 6 grocery stores are very near the kiosks, including 2 downtown and 4 on the east side. The system is convenient for many of the residents of Santa Cruz: 18,300 people live within a 5-minute walk of a kiosk (about a 1/4 mile, indicated by the circles on the map), and 28,500 live within a 10-minute walk.
The system would also be attractive to Santa Cruz’s many tourists. Three kiosks are located directly adjacent to the beach and boardwalk park, and would enable easy access between many hotels, the beach, and the very attractive downtown. (A downside is that many bike share operations require the users to be 18 or older — the bikes are too large and heavy for children, and the operators don’t want the liability risk.)
For reference, the population of Santa Cruz is 60,000, of which 7,000 live up the hill on campus, mostly students. You’ll notice that in all of these plans I totally ignore the UCSC campus, which is well up a hill from the rest of Santa Cruz. This is not because I think bike share would be a bad idea at UCSC — in fact, I think it would be great and very convenient for students getting across campus. Furthermore, getting students in the program would presumably also get them to come downtown more frequently, and make them feel more like a part of the city.
However, bike kiosks on the campus would probably more appropriately be funded by the university itself, as they would largely be used for getting around campus. To the extent that these kiosks would connect to the larger network, they would also pose a significant “rebalancing” problem: many bikers would only use the Cruzer Bikes to go down the hill to town, and then take a bus back up to campus. This would require frequent uphill hauls of Cruzer Bikes by the bike share operators. Thus, I’ve chosen to ignore the logistical hurdle of the UCSC campus entirely, and all per-capita figures below are calculated on a denominator of 53,000 citizens.
Plan B: Aggressive Pilot
Here’s an alternative scenario, if the city were interested in quickly rolling out bike share to much of the city, as opposed to a system that is focused on downtown and the beach. This plan adds several small bike kiosks to the commercial strip along the west side of town, as well as a few more stations along the southern bluffs and in the east side. There are now 12 large stations, 17 small stations, and 224 bicycles. 32,500 residents would live within a 5-minute walk of a bike kiosk, and 45,500 would live within a 10-minute walk, suggesting that the system is broadly accessible.
There are significant cons here, particularly in the western network. I noted above that the west side of Santa Cruz is not totally ideal for bike share. Most of the commercial real estate of west Santa Cruz is along Mission St, aka, the famous and scenic California Route 1. As famous and scenic as it may be, it’s a pretty terrible street for biking and pedestrians in Santa Cruz, dominated by fast-moving traffic. The businesses almost universally have their own parking lots, and so the street doesn’t have on-street parking, further increasing the speed of cars. Some of the signalized intersections don’t even have all four crosswalks, requiring pedestrians who’ve just gotten off the bus on the north side of the street to take long, circuitous routes just to reach businesses on the other side!
Nevertheless, there are a lot of interesting restaurants and shops along Mission, as well as a lot of residential lots in the blocks off of Mission, so I thought it would be worth it to consider. Frankly, though, I don’t think it makes a lot of sense with the current road design.
Ideally, the Cruzer Bikes would come along with a redesign of Mission St, to make it more bike and pedestrian friendly — for instance, replacing a traffic lane with two bike lanes, or a buffered, two-way cycle path. (The fact that it’s a state highway will make this complicated — state DOTs typically resist any changes to their roads which reduce auto speed.) Just as importantly, the land use along the corridor could gradually be allowed to densify, with reduced parking minimums for new construction.
Plan C: Mature
This one is less projection and more a thought experiment: what would a successful, mature Cruzer Bikes look like? That is to say, a system which is thought of as a normal component of the overall transportation system, where the bikes are a convenient way to get around the city, as well as a point of civic pride.
In this plan, the system has been extended so that bike kiosks blanket both the east and west sides. There are 15 large stations and 28 small stations, for a total of 307 bikes. 38,800 residents live within a 5-minute walk to a kiosk, and 49,600 live within a 10-minute walk (at this point, coverage is good enough that some of these 49,600 are in Live Oak, an unincorporated place just to the east of Santa Cruz). Almost all commercial areas in Santa Cruz are now within reach via Cruzer Bikes, except for the car-oriented big box at the very north of town next to Highway 1. One could imagine that the system would allow a non-zero number of families to forgo owning a second car, or even a first car, because almost the entire city is now conveniently accessible by bike.
Here I dealt with the problem of Mission St. by simply moving the bike kiosks a block or two to the north and south, along major bike paths. These paths are mostly residential streets which have either been restriped with bike lanes, or have relatively slow and light traffic, so that bikes can be comfortable sharing the road. Ideally, you’d want to the bike kiosks to be along commercial thoroughfares — this provides many benefits lots of pedestrian and bike activity outside of businesses, and a far more interesting route for bikers. However, Mission St. is just too scary for bikes at the moment, so this plan is probably appropriate for the current state of Santa Cruz’s roads.
Costs and Benefits of Santa Cruz Bike Share
The cost of a small station, including 4 bikes and 8 docks, is around $27,500. A large station with 13 bikes and 22 docks is about $56,000. (Source: pdf.) According to that same source, bikes need to be replaced about every six years, while the stations need to be replaced every ten. We’ll simplify and assume that every station (bikes and all) needs to be replaced every eight years, and annualize these capital costs at a discount rate of 3%.
The main costs of operating the system are maintenance of bikes and kiosks, and rebalancing stations when they’re empty or full. Capital Bikeshare estimated its annual operating costs at $1,860 per bike. (Source: the same pdf.) Santa Cruz will be a smaller program, so we’ll ballpark this at $2,000 annually per bike, owing to reduced economies of scale.
Ridership and Revenue
According to a blog post at Greater Greater Washington, Cabi made about $1,500,000 in subscription fees and $770,000 in overage fees, on a total of 1,000,000 rides. They had 18,000 annual subscribers and almost 70,000 short-term users. Thus, as a very hand-wavy estimate, we can say that each bike ride in that system generates about $2.30 directly in revenue. An annual subscription to Cabi costs $80. Because Santa Cruz will be a smaller system, annual subscriptions will cost less, probably around $60 (compare this to $60 for Boulder, CO, $65 for Charlotte, NC, and $65 for Madison, WI). So instead of $2.30, we’ll assume that each ride generates $1.70 in system revenue.
How can we estimate ridership? The Cabi system had about 1,000 bikes over the first year; 18,000 rides then corresponds to about 2.7 rides per bike per day. However, in their TIGER grant application, they estimated about half that, or 1.35 rides per bike per day. We’ll adopt that figure as our baseline. However, for the wildly optimistic Mature System Scenario below, we’ll see what happens if we assume four rides per bike per day. For comparison, Paris’s Velib system averages about 4.3 rides per bike per day, while Barcelona’s Bicing averages 13.3. Four rides per bike per day is certainly a pretty optimistic scenario, but I think it’s a realistic description of what a successful mature Cruzer Bikes would look like. (Update: I put together a collection of ride/bike/day figures reported by other U.S. bikeshare systems here.)
And how about annual subscribers? D.C. estimated that they would have around 10,000 annual subscribers in the first year (as noted above, it turned out to be 18,000) on a system of about 1,000 bikes. We’ll again be overly simplistic and project 10 annual subscribers per bike for our two pilot programs. The Cabi Dashboard informs us that as of the most recent available period (September 2012) they had about 28,400 annual subscribers for a system with about 1,500 bikes: about 19 subscribers per bike — and this figure is still increasing rapidly (as of May 2012, it was 16.4 subscribers per bike). Let’s be quite optimistic and assume that a fully mature system would have 30 subscribers per bike.
But wait, there’s more! The true benefits of the system include more than just the direct income of the bike share operator. Biking also reduces congestion, reduces emissions, and is cheaper for the user than most other transportation options. Higher rates of bicycling are associated with reduced bike accident rates (“safety in numbers“). Furthermore, bicycling is associated with stronger downtown businesses.
Specifically, we can look at the cost-benefit analysis in the Cabi TIGER grant application (linked above). These figures include benefits due to riders due to time and cost savings, as well as pollution and emissions, but do not include local economic benefits. I’ll vastly oversimplify and note that from the 18,000 daily riders Cabi expected in year 2, they estimated $56,000 in external benefits, or about $3 per ride.
Most of these benefits in fact go directly to the users of the system: using bike share is cheaper than driving or taking transit. Since parking and transit are relatively cheaper in Santa Cruz, let’s make a ballpark estimate (read: wild guess) that each bike ride generates $2.00 in external benefits.
Summary of Costs and Benefits
Here’s a table combining all of the costs, benefits, and assumptions from above for each of the three systems:
|rides per bike per day||1.35||1.35||4|
|annual subscribers per bike||10||10||30|
|net cost, per capita||$5.24||$8.32||$2.04|
|total external benefits||$140,000||$220,000||$900,000|
|external benefits, per capita||$2.70||$4.17||$16.91|
|avg. nearest neighbor (m)||352||448||385|
We see that both of the pilot systems require some subsidy: about $280,000 annually for the smaller pilot program, and $440,000 for the larger. They don’t generate external benefits large enough to justify the subsidy, if it’s funded purely municipally.
The mature system, because it has more subscriptions and riders, comes much closer to funding itself: it only requires $110,000 in annual subsidies. On the other hand, all those bike rides are generating large external benefits for Santa Cruz of around $900,000. This system would make perfect sense for municipal subsidy.
It’s worth noting that Santa Cruz METRO, the local bus system, has a farebox recovery ratio — that is, revenue divided by operating costs — of between 21% and 26% depending on the year (pdf source). (This is pretty typical of bus systems, and of course, the figure doesn’t include capital costs at all! Rail systems can be somewhat more efficient by this metric due to decreased labor costs per rider, but also typically have much higher capital costs. See this handy chart on Wikipedia.) The corresponding metric for our systems are 41% and 42% for plans A and B, respectively, and 127% for the mature system. So if we think of bike share as transportation spending (and we should) the systems are relatively cheap.
Other funding sources
As I mentioned above, there are other funding sources available to match the required subsidy. Advertising is a major one: Cabi intends to sell advertising on its kiosks and bikes, though it hasn’t as of yet, while CitiBank is completely sponsoring capital and operating for the upcoming New York City system in exchange for advertising rights. (Any profits will be split between CitiBank and the city.) Velib in Paris is paid for by allowing JC Decaux billboards throughout the city. Similarly, downtown businesses or the business improvement district may be willing to put up a substantial fraction of the cost, due to the presumably increased pedestrian and foot traffic in these areas. Finally, many U.S. systems were primarily financed through federal grant programs like CMAQ or TIGER.