Docking stations in New York's Citi Bike scheme (in the dead of night - at 3.20am Sunday)

New York’s commercially-financed new bikeshare scheme, Citi Bike, isn’t even a week old but already it’s obvious, notwithstanding some criticism, that it’s taken the big apple by storm (video here).

After just five days, it’s recorded an average 8,105 trips per day and enrolled 25,276 annual members (fn 1). It looks set to join other conspicuously successful bikeshare schemes like those in Paris, Barcelona, London, Dublin, Montreal, Washington DC and Minneapolis.

Each Vélib’ in Paris is used four to six times per day. Hire rates per bike in Barcelona run at seven times a day in summer and in Washington DC each bike is used five times a day in peak season.

The apparent overnight success of Citi Bike highlights the disastrous experience with bikeshare in Australia.  Melbourne Bike Share had just 253 hires in its first week of operation in May 2010. After six months, even with the introduction of a subsidy for helmets, usage rose to 183 hires per day and memberships reached 650. It’s now up to around one daily hire per bike i.e. on average each bike gets used for about 20 minutes a day.

The performance of Brisbane’s CitiCyle scheme, which also opened in 2010, is even worse. Despite having three times as many bikes and stations as Melbourne Bike Share, average usage languishes at around 0.4 daily hires per bike.

Why do schemes in other countries appear to perform so much better than Melbourne’s or Brisbane’s? That’s hard to answer definitively because getting reliable and comparable usage data is very hard. Nor have any researchers undertaken yet the sort of sophisticated multiple regression analysis that would be required to explore the question properly.

The seemingly obvious answer is that Australia has the only bikeshare schemes that mandate and enforce helmets. If prospective riders have to find a clean helmet even before they can hire a bike that’s got to be a serious dampener on demand.

While the helmet law is undoubtedly an important factor in the poor performance of bikeshare in Australia, it’s unlikely it’s the only reason. In fact it might not even be the most important one.

As of 2011, there were over 300 bikeshare schemes throughout the world. Other than Melbourne and Brisbane, none mandate use of a helmet. Yet there is considerable variation between them in terms of usage. For example, although the Dublin scheme has fewer bicycles than the Goteborg scheme, usage is ten times higher. Something other than helmets must account for the difference.

This reflects the pattern for cycling in general too. As noted here, cycling rates differ significantly between countries that don’t have Australia’s strict helmet laws. For example, bicycle use in the Netherlands is nine times higher than it is in France and Italy and more than twice as high as it is in Germany. Those differences have nothing to do with helmet laws.

Nor does helmet policy explain why bicycles capture 34% of trips in Munster, but only 13% in Munich. Or why the corresponding figure for Groningen is 37% compared to 10% in Heerlen; or 20% in Bruges but 5% in Brussels; or 19% in Salzburg but 3% in Wien.

Repeal of the helmet law would undoubtedly lead to increased demand for bikeshare in Australian cities, but usage might still be lacklustre relative to other schemes. It might still be insufficient to generate benefits that exceed the publicly-funded costs.

That’s because there are other factors at play. They include differences in:

  • density of bikes and stations;
  • density of population, employment, education and activities;
  • pricing, access, operating hours and marketing;
  • cycling-specific infrastructure and regulation of vehicles;
  • car ownership level
  • the demography of the population;
  • historical cycling culture;
  • level of public transport supply;
  • climate and topography.

Consider that on launch and with no public financial subsidy, New York’s Citi Bike has 6,000 bikes and 330 stations in Manhattan (south of Central Park) and parts of Brooklyn. In the next iteration that will increase to 10,000 bikes and 600 stations.

Melbourne Bikeshare launched with a mere 100 bikes and now has 600 in 50 stations. CitiCyle has 2,000 bikes in 150 stations. Here’s a map of the Melbourne and Brisbane schemes at the same scale as New York‘s (see exhibit).

The demography of New York’s residents favours bikeshare – Manhattan and Brooklyn are populated by the cognitive elite. Experience with Washington DC’s Capital Bikeshare scheme suggests they’re the primary market for bikeshare (indeed, some observers claim bikeshare in the US is “another kind of luxury item”).

While NYC is still far behind the leading European cities, Mayor Bloomberg also put a lot of effort into improving cycling infrastructure ahead of the launch of Citi Bike. Meanwhile, Melbourne struggles to provide even a short section of bicycle lane on Princes Bridge.

I suspect one of the most important differences is density. Manhattan and Brooklyn are much denser in terms of population, employment and other key destinations than inner city Melbourne and Brisbane, or Washington DC for that matter. In terms of density, Manhattan is comparable with Paris.

Just as importantly, that density extends across a very large area. Whereas most major inner city attractions are in or close to the very compact CBDs of Australian cities, there are many key destinations spread across the territory covered by Citi Bike.

That also impacts on the relative attractiveness of cycling and public transport. New York enjoys a well-deserved reputation for the quality of its public transport system, but travel within Manhattan often involves transfers. That provides a niche for the option of faster direct travel by Citi Bike (and, of course, by taxi).

At $95 p.a. ($60 p.a. concession), a Citi Bike subscription provides a very inexpensive option for those whose frequent travel range lies within the area covered by bike stations. There’s no maintenance cost, no need to carry the bike upstairs, and no need to store it securely in a tiny apartment with no balcony.

Popular expectations of bikeshare schemes in Australia are shaped in large measure by the example of outstandingly successful schemes such as those in Paris and Barcelona. But they’re not typical of all 375 schemes around the world. Importantly, the two Australian schemes lack some of the key attributes that contribute to success elsewhere.

Melbourne Bike Share and Brisbane CitiCycle would undoubtedly perform better if helmets weren’t mandatory, but I don’t think an exemption would necessarily, or even probably, be a silver bullet. Further actions would also be required before either could be considered a “success” by the standards of schemes like, for example, Dublinbikes.

Dublin is often cited as an example of what Melbourne could do if it weren’t for the helmet law (DublinBikes started with a similar number of bikes and stations). However it also has a 30 kmh speed limit in the central city and bans on heavy vehicles. Irrespective of the issue of mandatory helmets, that’s one example of the sort of change the Melbourne and Brisbane schemes require to take patronage to a level that justifies their growing public subsidies.


Update: After ten days operation, Citi Bike is averaging 10,650 trips per day (16,395 on day 10) and memberships have increased to 32,033.

Every docking station in some of the world's 375 bikeshare schemes mapped at the same scale
Every docking station in 23 of the world's 375 bikeshare schemes mapped at the same scale