The Global Diffusion of Bikeshares (I)

17 July 2013, 1139 EDT

velibEditor’s Note: This started off as two bullet points, but it’s morphed into a surprisingly lengthy piece about Bikeshares.  Blame the easy availability of both picturesque bike rides and cheap wine in Western Europe.  This is the first installment; a second will follow.

At the age of 19, I moved to the Netherlands for a summer.  Like most tourists, I had read about the famous Dutch “white bikes,” a non-locked bike-sharing system that guidebooks were fond of presenting as proof of Dutch civic-mindedness.  The system was simple: see a white bike, take, use, leave wherever.  Of course, there was an alternative course of action, perhaps unforeseen by white bike planners: take, use, spray paint, keep.*  I never saw a white bike.  There are mixed reports on whether the system was ever particularly vibrant; by the 1990s, however, it was gone.**

In the mid-2000s, the solving of the enforcement issue allowed the birth of modern bikeshare systems in western Europe.  These systems relied on locking stations, credit card-secured rentals, and heavy step-through commuter bikes.  They now exist in major cities on five continents, and we’re inching toward exponential growth in both numbers of systems and riders.  Western Europe is increasingly saturated, and bikeshares are spreading to smaller towns, especially in France, Spain, and Italy. North America has been slow out of the gate, but there are established systems in DC, Boston, and Denver.  New York’s Citibikes are finally online. Los Angeles, Chicago, Miami, San Diego, and others should be up within the next year.  Though information on bikeshares outside the OECD is bikeshares is minimal, they do seem to be a truly global phenomenon. In Asia, Hangzhou currently has the world’s largest bikeshare with 66,500 bicycles, roughly 3 times the number of Paris’s Velib; Beijing’s currently has 14,000 with plans to expand. Several Indian cities have incipient bikeshares, as do cities in several Latin American countries.

Why bikeshares matter.   There are three big social sciences stories here, of which I’ll address only the first today.

  • The global diffusion of bikeshares to a wide range of settings, coupled with a surprisingly quick winnowing-down of technological models to two or three major contenders;
  • Why some cities adopt bikeshares, while others do not.  Is adoption driven by city-level characteristics or by diffusion mechanisms, such as competition?
  • The potentially transformative effect of bikeshares on 21st century public transport and urban development.

QWERTY as the cost of rapid expansion?  Most new bikeshare systems tend to cluster into one of two technology families, one associated with French advertising firm JCDecaux (most associated with Paris’s Velib) and the other with Montreal’s Bixi.  The English-speaking world (London, North America, and Australia) is essentially all Bixi, while much of Europe is split between JCDeaux and a few other providers.  For the few who might remember DC’s Smartbike, similar systems are subsidized by ClearChannel, another advertising firm, in Barcelona, Mexico City, and a few other places. From the pictures of Beijing’s systems, I wonder if it’s not also a ClearChannel design, while Taipei looked Bixi to me.  Unlike Bixi and JCDecaux, ClearChannel uses beach cruiser-style bicycles with a high crossbar, which limits their appeal to the elderly, pencil-skirted women, and nattily-dressed men.  Across all systems, there seems to be convergence toward a single rate structure, which mixes a time-delimited subscription fee with overage charges (i.e., trips of more than 30 minutes involve additional costs).  ClearChannel is again an outlier, as it does not impose time charges.

I spent much of the past week wondering if there’s a Piersonian path dependence problem at work, perhaps exacerbated by the public-private partnership model that characterizes bikeshares. This past week was spent in Paris; the previous two in Valencia, Seville, and elsewhere.  These are all JCDecaux systems. Though JCDecaux’s side-mounted system seems to work well (or is better maintained?) in Seville and Valencia, Paris’s Velib is a bit of a mess. Docking is hit-or-miss, and there is no way for drivers to put in a maintenance request (unlike the Bixi system, which holds all bicycles that require maintenance).  In 1 in 5 trips in Paris, I pulled a bike with a maintenance problem, occasionally severe enough to make riding impossible.  It might be that Paris just does poor maintenance, but I couldn’t shake the idea that the Bixi docking and maintenance system — and, indeed, the bikes — are just better-designed.

Of course, public transport systems are often associated with path dependence problems (e.g., track gauges, overhead tram electrification).  Paris and other early adopters have problems of sunk costs, and cities that wish to expand their bikeshare network must do so with the same technology with which they began.  But even new bikeshare schemes might continue to adopt inferior technology due to the dynamics of sponsorship. For instance, JCDecaux subsidizes the costs of bikeshares, and, presumably, its distinctive docks are now part of its branding.  In Western Europe, where JCDecaux and ClearChannel seem to split most of the contracts, cities may have little choice regarding infrastructure.  The North American / London systems, in which technology and corporate sponsorship are delinked, seem a better model.  But I wonder if there isn’t still cause for worry in the speed with which these three models established dominance over the bikeshare market, which presumably now crowds out innovation.


* See also, Portland’s Yellow Bike scheme.

** In The Hague, one still saw unlocked bikes parked outside shops in the late 1990s, and there are apparently still pockets of the Netherlands – notably national parks – where the white bikes survive.  In 2003, the fairly small-scale Ov Fiets scheme was launched in the Netherlands.  This is funded by Dutch public transport, and the bikes are largely located at rail stations.  It is designed differently from most existing systems.