Do Bikeshare Systems Hurt Bicycling?
A woman using the Bicing bikeshare system in Barcelona, at one of many large diamond-shaped intersections. Did this system hurt bicycling in Barcelona? (Photo, Jason Meggs)
Much debate and many growing pains have accompanied the fanfare around the rise and rapid adoption of bikeshare systems such as Vélib in France and BIXI in Montréal. A recent article appeared critiquing the Bicing system of Barcelona, and by coincidence it came to my inbox (and was published) on the day of my first ever arrival in Barcelona. It took me more than a month to respond, but am glad to have done so. My commentary is published below, with my qualified conclusion in this case: no, it has not hurt bicycling!
The original article was here:
My commentary (soon to add photos, perhaps video too!) is here:
Thank you for looking at the big picture.
It was very interesting to receive this article because on that day, the day it was published, I had actually just arrived in Barcelona for the first time, and was in the process of analyzing the relationship between infrastructure and cycling in the BICY project (http://bicy.it). So it gave me a much appreciated puzzle to work on while in the area.
I spent several days pondering this during my visit, and even did some traffic counts and tours of the bikeways as well as cycling to the outer limits of the city to get a better understanding of the bicycle situation there.
First I must say that bicycle data is notoriously unreliable and collected in inconsistent ways, making it further unreliable or incompatible to use for analysis and comparison. It can also be quite political, and I’ve heard from people from other cities who do not believe the figures for cycling in their cities, because funding is sometimes allocated based usage.
Given this and my first impression, I was initially skeptical about the low cycling rates reported, particularly as in certain areas of the center one sees many bicycles. In fact the relatively low rates for everything but walking (45.5%!) presented a challenge.
However, having taken a further look, it’s quite likely these numbers are reasonably accurate. Barcelona, like Paris, is one of the great walking cities, and high cycling in one area does not translate into high overall cycling.
The low rates of cycling are also predicted by our model developed for cities in Central Europe, based on the amount of infrastructure present, although we would have estimated 4% not 2%, which given the many factors and the uniqueness of the city is still quite close.
The unique diamond intersections where cyclists must made several rapid sharp turns were fun in a race course sort of way but struck me as dangerous and inconvenient, due to poor sight lines, tight turning, and unexpected conflict zones, although adaptation may reduce that risk considerably. However it probably further discourages cycling.
Interviewing residents I heard repeatedly that the frequent stopping for red lights creates even more frustration, and leads to dangerous and aggressive behavior. We saw this repeatedly as drivers peeled out with screeching tires after waiting at red lights, even if cross traffic of pedestrians had not yet cleared.
I had to wonder why not turn the diamond intersections into circles, probably removing some or all parking in the process), with many net benefits in noise and air pollution reduction, traffic calming, reduced wear and tear on the entire system (infrastructure and vehicles), and collision reduction, while presumably making it a more free-flow and bicycle-friendly environment.
Aerial view of the unique diamond-shaped intersections of Barcelona. (Source, Google Maps)
Observing the use of Bicing bikes, I noted they made up more than 50% of bicycle traffic in some central areas, at some times of the day, but in other areas made up much smaller percentages (and in large areas of the city that are not served, zero). Traveling with a group I saw how local Barcelona residents used the bicycles, in these cases enabling groups to travel together when some didn’t have bicycles.
In the end I certainly didn’t have time nor resources to conduct a full assessment, and respect the analysis in this article and the city’s publications. However I would like to put in some words of consideration in defense of the Bicing system:
1. Cost justification
While the system may cost a tiny bit more than the Metro or bus per trip, this is not a reason in itself to discontinue the service unless the Metro and bus systems should be discontinued as well (which, by the way, would be expected to increase bicycle use a lot, but likewise would increase motor vehicle use – the good news is bicycling has much more room to grow than car use, the bad news is motor scooters are almost as unlimited in their potential to increase, and carry worse emissions and noise pollution than cars for the most part). A bicycle trip has many benefits that a public transit trip does not, it deserves an even higher subsidy if need be. If we develop an ideal for traffic in Barcelona (choose your favorite), surely the total investment in cycling is much lower than it should be.
2. Political avoidance of building bikeways
Given that investment in cycling facilities is too low, the concern that the Bicing system is being used as an excuse for preventing development of a true and quality bikeway network is serious, and something raised other places (I’ve even heard officials in Montreal voice this regarding the BIXI system, for example).
However, the fact that many times more are invested in Bicing does not have to mean nothing for infrastructure. There should be more for both.
The kind of analysis presented here is a first step to action for an increase of funding for and implementation of infrastructure including secure parking.
But the situation is not so bad: it is very encouraging that based on the official report, infrastructure increased 8.3% 2009-2010, and has steadily increased since the system opened after a long plateau last decade.
Yes this needs to be much faster, but is better than many cities and increasing, and going in the right direction. Also good news, parking has increased a great deal since 2007, suggesting Bicing was related to increased bicycle parking (p. 57). Again too little but in the right direction.
3. Suppression of cycling
First I have to question that Bicing reduced cycling. If you look at the graph of bike trips per day, they nearly doubled when Bicing was introduced in 2007, and have only fallen slightly since then (see p. 55 of the 2010 report).
Discussing with people in Barcelona about the concern that the system led to lower levels of cyclists and is contracting, local residents told me that the surge and contraction of membership had to do more with the promotions and newness at the beginning (many people encouraged to try and interested to try) coupled with the glitches in the system at the beginning (people encountered major system unreliability being unable to obtain a bicycle, and then unable to leave it, so gave up). These aberrations are normal in a new system, the thing to look for is the long-term trend which seems stable with higher ridership since its introduction. In fact the report supports this, particularly looking at the first two years.
Whether there are more negative effects of a bikeshare system on ridership for some people, or on long-term growth potential, is of great interest, but needs more clear evidence.
One bike rental operator told me he thought the system had increased private bicycle ownership and boosted bike shops, by allowing new cyclists to try bicycling and then realize having their own bicycle was better. (At the same time, he was deeply concerned about opening the system to tourists, because many bicycle rental businesses would collapse.)
Regarding the slight decline of bicycle use in the one-year span 2009-2010, this seems troubling but is actually small, and may be a statistical anomaly. There is error in modal split surveys, which appears more exaggerated for small numbers (the same error in walking would not be noticed, like a regressive tax the minority is hit harder). It’s even possible the true number increased. Certainly it is possible the Bicing system encouraged some people to give up private bikes, because it was easier to use the bikeshare occasionally for them, but on balance many more people are cycling since its introduction. The problem is this is still far too small for the potential.
It is also heartening that the total trips in Bicing, which should be a very reliable number, have increased by 377,744 2009-2010, although the rounded number for the higher 2010 figure suggests it’s approximate with unknown error.
4. Bicycle behavior
I did not see much evidence of bad cycling behavior during my stay, but it was a short stay. I don’t assume all non-bike lane trips are on the sidewalk, however; certainly I used the streets often, as in any city, and saw others doing so. (I’m also skeptical of any figure that attempts to know where all cyclists are, it’s not an easy task as recent GPS-based studies of just a fraction of cyclists attest.) However, there is a learning curve in any culture adopting cycling, and public support for education of cycling skills as well as sensitivity and awareness from non-cyclists should be added to the imperative goal of increased quantity and quality of infrastructure.
My conclusion: cycling in Barcelona is on the rise. Bicing is not causing harm, it has overall helped. However, it must not become a barrier to major actions to increase cycling.
The Copenhagenize Index recognized this by allowing it to score highly despite having very low cycling overall and a relatively poor network.
Next steps are to consider the relationship of Bicing to bicycling (because scaling cycling begs for scaling Bicing, or changing its use), and to re-envision the relationship of all bicycling to the public transport system, as bike trips can be more direct, reliable and cheaper and will be preferred if people feel safe and accommodated to do so.
A truly bicycle-friendly Barcelona would mean a shift from all other modes: motor vehicles as well as from public transport, and even a decline in walking. For this case, Barcelona must choose to be a bicycle city.
Certainly from a public interest perspective the argument is strong: The money saved on more expensive modes more than justifies the investment in cycling; the cost is less to begin with, even before the tremendous benefits (health benefits, boosts to the local economy, and more, including tourism).
Meanwhile, consider conversion of traffic signals where possible, particularly in the diamond-intersection areas, and installing roundabouts and shared space/slow zones in large areas of the city, hand in hand with new bicycle infrastructure and an array of new restrictions on driving both private cars and motor bikes (motorcycles and motor scooters).
Picture this; if you like it, make it your goal; it can happen very quickly.
A final note: this data is now almost two years old. It would be good to know the latest developments now that the system, and the public, have had more time to adjust. (And speaking of delays, my comment has been delayed first because I wanted to do more research, apologies for the time lag.)
The Meggs Report 
My question would be whether bike sharing increases bicyclists. How many customers are people who ride anyway? All fine, of course, but not going to make any major converts. I still have issues with helmetlessness. That turned me off the DC bike share. That, plus crazy DC streets, traffic circles and traffic.
Thank you Phyllis,
I think it did increase bicycling. First, it just makes sense that a large new system in support of bicycling would increase cycling — even if it reduced some private bike use, it surely creates new use as well, and puts the possibility of bicycling at everyone’s fingertips, a giant leap forward.
The numbers support this: when I looked at the official data, it seems quite clear that the total trips by bicycle roughly doubled in short order when the Bicing system was introduced. This can partially be explained by other measures, but it seems certain (assuming the data is correct), that the bikeshare system contributed a lot of new trips.
I fully agree we need to understand the changes of behavior better, this can be achieved through surveys and by analyzing the data collected by the system, amongst other methods. The support and attention to bicycling brought on by the system surely must have had effects, some of which might not be immediately measurable, but many of which could be. I’m optimistic that bikeshare systems are an important step toward widespread social adoption of bicycling and again, at this point it doesn’t appear there’s any argument that the investment is not worth it in a full cost-benefit analysis. This doesn’t mean it’s the only or even the best investment. Providing lots of infrastructure and no bikeshare system might bring on a more sustainable system even faster. But what is the loss without the benefit of bikes on demand?
Regarding cycling in D.C., having been a bicycle messenger there before any of the improvements that define cycling there today, my perception is D.C. has become a leading bicycling city in the USA, and the numbers support that. Yet there’s still much more to do.
The helmet concern is a large and longstanding debate, certainly there’s a lot of justification for not requiring helmets — the benefits of the bicycling in the balance far outweigh the safety benefits, which in fact are controverted. I’d be interested to know more about why you choose not to ride a bicycle because of another person’s personal practice regarding protective equipment. Does this extend to other things as well? Driving, for example? If not, why is bicycling special that way?