Thursday 24 February 2011

Transport for London: Baffling people with huge numbers

Does this man look trustworthy ?
Transport for London recently told the world that 2.5 million cycle journeys have been made on the "Barclays" London bike share bikes.

Instead of straightforwardly presenting data which shows the true extent of the usage of these bikes and puts it in a context which helps the reader to understand what the numbers mean, London has used a style of writing which often is used for projects like this. i.e. Trying to impress the reader by quoting what sound like huge numbers, and doing so without any context so that the meaning of it remains clouded. In this blog post, I explain what the numbers really mean.

What they say
In London's press release, you'll read of:
  • Cycling to the moon and back 13 times in the first six months
  • Covering more than 10,000,000 km
  • Building on the massive 117% growth
  • Making a huge contribution to the cycle revolution
Some context
However, let's look at those figures. Six months, 2.5 million journeys. The numbers sound great, but actually if you look closely at them you quickly see that this is not actually very impressive at all. This equates to 2.28 journeys per bicycle per day or less than a quarter of the estimated ten uses per bicycle per day that promoters used when selling this system.

London has a population of 8 M people. Between them, they make around 20 million journeys per day. If all the six months worth of shared bike journeys had been made on just one day (requiring each bike to be used an impossible 416 times), then even that would make up only 12% of total journeys in the city. However, actually it took half a year, 182 days, for this many journeys to be made. The total usage equates to only around 0.07% of the total journeys in the city. On average, Londoners are using these bikes not once per day, not once per week or once per month, but about once every 18 months. In fact, as each individual trip is being counted here, you could also say that this is the equivalent of the average Londoner going out and returning home again about once every three years by using these bikes.

So far, the scheme has cost 140 million pounds. The cost to Transport for London for each of the 2.5 million journeys made so far is £56 pounds. The hypothetical Londoner making their once per three year trip to the pub and back is subsidized to the tune of £112. Does that sound remotely like a reasonable or sustainable cost ?

Is this really deserving of the amount of hype which it receives ? It is deserving of comments from the mayor of London, Boris Johnson, such as 'The zest in which people have taken to two wheels and joined the cycling revolution we are engendering in the Capital has gladdened my heart.'. I'd love to see a genuine cycling revolution in London, and in Britain as a whole. However, let's please be realistic: this isn't the revolution you were looking for.

Back in March 2009 when I first wrote about the potential of the bike share scheme in London, I calculated that based on the promoters estimates of future use, it still had a maximum potential capacity of only 0.3% of the total journeys made in the city. Some people criticised me for pointing this out. However, actual usage has been under a quarter of what I wrote about.

Is this mass cycling ?
By way of contrast, a few days ago I ran a story about the amount of cycling in the Netherlands. The Netherlands has only twice the population of London, but there are over a million bike journeys an hour for the daytime hours of working days. That's more than 14 million journeys per day by bike, or 2.5 billion per six months. Let's state that in another way: The Dutch population is twice as large as London's population, but they're making a thousand times as many cycle journeys as the "Boris Bikes" are used for.

In fact, no shared bike scheme can ever possibly scale to cope with mass cycling. It would simply take too many bikes. What's more, as shown above, it can never introduce cycling as part of a regular habit to anything like the whole population of a city because the total capacity is so small for any affordable system (even a very expensive system)

Large bike share schemes are what cities install if they want to give off the appearance of being interested in cycling without investing in the infrastructural change which is necessary to make a real difference. There were already over a million bikes in London before the bike share was installed. People need safe and attractive conditions for cycling, not more bikes.

Of course, Londoners do actually ride other bikes as well. It's estimated that half a million cycle journeys are made every day in London on bikes other than bike-share bikes. These are far more significant numbers than the bike share bikes provide, in fact 36 times as large. What's more, this number has a far greater potential for growth than bike share does, as no-one wants to put 36 times as many of what are already the most expensive bikes in the world onto the streets of London.

You may also remember that a while back it was revealed that 84% of users already have their own bike and that almost half of users would otherwise ride their own bike. So, only half of the 2.5 million journeys over the last six months are actually new cycle journeys. To a large extent, the system is being used as a cheap to the user, but expensive to the tax payer, method of keeping existing cyclists bikes safe from vandalism and theft. Perhaps if London had spent some money on improving its inadequate cycle parking then this wouldn't be the case.

To bring about a genuine "cycling revolution in London, the environment must be made more conducive to cycling. People want to do it, but they need their local government to stop wasting vast amounts of money on silly things like bike share and invest properly in those things which have a proven history of success in promoting cycling.


Unknown said...

Toronto is about to implement the "BIXI" bike share system.

My take on it is that it is a "gateway drug." It enables people to try cycling at little cost or risk. It is also one more reason to enable the City government to justify proper cycle infrastructure.

The same mode share criticism can be made of Velib in Paris. Yet, by all accounts, Velib has been successful in changing the culture of Paris. Nice.

David Hembrow said...

Kevin, I've heard this opinion expressed many times. However, is it true, or just wishful thinking ?

I'll be quite straightforward and say that I don't believe it. If the culture had genuinely changed, in Paris or anywhere else, then the modal share would also have changed.

Attempts to measure the effect of bike share schemes have never shown much.

Paris, by all accounts, still looks like this. That's the root of the problem. Just like London, they're also playing the hype game instead of doing what actually needs doing.

amoeba said...

Of Course you are right. I too had noticed the huge disconnect between the overblown hype of neverland [the TfL / Boris Bikes scheme] and the Netherlands. Bungling Boris and numerous other politicians seem willing to go to any length and do anything for cycling, except those things that have actually been shown to work. What a shame that £140 million wasn't spent on high-quality, safe and useful facilities for cyclists. Then Bungling Boris would really have had something to crow about.

Unknown said...

I don't have hard data. But I keep running across comments such as:

"Velib has changed the way we get about in Paris. Riding a bicycle used to be quite marginal and frowned upon, it used to be a problem, but now more and more people ride bicycles and there are fewer cars on the roads.”


I know that in Toronto (and much more so in Quebec) there is a certain culture of deference. So if something is "official" it has the imprimatur of respectability.

It will also be part of the offical tourism promotion. Torontonian's self-image is to some extent shaped by how the government shows us to be. If we are shown cycling, this becomes a self-fulfilling description.

Unknown said...

I see that there are people saying that Bixi changed Montreal. See:

I really like the hilarious rap video produced by three lilly-white Quebecois.

David Hembrow said...

Kevin, if a statement is made that "more people ride bicycles and there are fewer cars on the roads" this should easily be measurable. Where are the measurements ?

Comments such as "people are saying that Bixi changed Montreal" are, I'm afraid, even more dubious. If it's changed, you can measure the change. If you set your standards for change to be unmeasurable then how can you possibly tell whether there has been any change at all ?

I can't deny that bike-share appeals to me as a concept. However, when I see the amounts of money thrown at it with no regard to effect, and see city after city doing this with seemingly no need at all to see others achieve real results first, I can't support it.

It has become a symbolic way of trying to show that you're doing "something for cycling" without having to make any real commitment to changing the status-quo. It also resembles a get rich quick scheme for the promoters, who are being awarded expensive government contracts without having to prove anything.

If your city is proposing this, you should be concerned.

Robert said...

I completely agree with you, David. There have been rumblings here in Prague about a bike-share program and it always makes me roll my eyes. The only thing here that has succeeded in making people "change the way we get about" has been when they built a nice, modern segregated path (sadly, multi-use and full of people walking their unleashed dogs...) through the center of the city. In good weather, that path is full to capacity with cyclists. And thanks to it, I'm now able to commute to work on my bike.

Dave M, Edinburgh said...

Although I agree it's very difficult to quantify, there's definitely more to it than the raw number of journeys made.

Compared with when I started cycling in Edinburgh, I see a sea-change in the interactions between motorised traffic and riders in many places - not to mention the general attitude of colleagues to bike commuting and so on.

I wouldn't be shocked to learn that the proportion of journeys made in Edinburgh by bike has not significantly increased in all this time. Nevertheless, the experience of riding here is much better.

I read that Belfast recently cut 98% of its funding for cycling. Perhaps another advantage of schemes like Velib is that they make it embarrasing and difficult for authorities to do this, as they have become direct stakeholders?

Kevin Love said...


Here are the numbers:

If this increases due to BIXI, then it must be reckoned a success.

The amount of public money that went into BIXI is zero. It is entirely funded by advertising revenues and user fees.

So there are no "expensive government contracts."

It is, however, becoming an argument in support of the proposed network of separated cycle lanes downtown with protective barriers.

David Hembrow said...

Kevin: There is of course no reference to BIXI at all in that document now. It's not clear how their methodology would be able to indicate how many extra journeys such a system had generated at some point in the future. Even counting the BIXI bikes won't work. In London it was found that half the usage was by people who would otherwise use their own bikes.

Also, sadly it's another example of just the sort of thing I criticised in the blog post. This is another document which lives in its own universe, full of "impressive" numbers but which doesn't compare them with anything else.

Five million people live in the urban area of Toronto. That's a bit more than half the size of London. I think this document does the cyclists of Toronto rather an injustice in counting just 19000 of them who entered the urban core each day. The real level of cycling in the city must surely be higher than that ! In a city of that size, you'd expect the population to make about 12 million journeys a day by all modes combined. Even a 1% cycling rate ought to mean 120000 cycle trips per day.

Unfortunately, these figures are very limited indeed. They don't give any indication of what happens to cyclists elsewhere but in a small area, they don't give any reference as to the total number of journeys made in the city. The danger with using such figures to try to work towards an increase is that future efforts will go towards increasing only those cycle journeys which increase these counts. Most of the city is excluded.

The closest that this document comes to discussing actual modal share is on page three where there is a map showing "Bicycle Mode Share for Work Trips in Toronto, 2006 (from place of residence)". Even this contains no accurate figures, but just shades on a map, and no figure is given for the city overall even just for commuters, let alone for all journeys.

As for funding by advertising... Yes, that's possible. However, it doesn't actually counter the argument. If you're willing to accept more advertising in your city, in exchange for benefits from an advertising company, that doesn't mean that the resulting money has be be wasted on bike share. There are a range of other things which could be paid for, such as cycle infrastructure for all cyclists rather than a tiny number of bikes which can be used by a few people.

J.. said...

I can't comment on Canada, but I've cycled in Paris and I must say, it's better than I expected. But then, I had never heard of Velib, so I was expecting it to be utter crap. What I found was only moderately crap with some nice things here and there. I was amazed that biking across town was even feasible. Of course, I am an "experienced cyclist" smack in the middle of the vehicular cycling demographic, so what do I know.
Although to Dutch standards (the only ones that matter) cycling in Paris is still medium crappy, I'll admit that Velib has had positive effects. Most importantly, it has made cycling normal. And it has also seemed to have raised awareness among drivers. The group of potential users (given the quality of infrastructure) is still too small, but it's bigger than it otherwise would have been.
These positive effects by no means constitute a "cycling revolution" as our dear Boris might have you think, but we have to be honest and appreciate whatever positive trends have been brought about. We have to get the ball rolling somewhere, so it might as well be here.

michael neubert said...

The BIXI system has been implemented in the Washington DC area - the "Capital Bikeshare" program. summarizes the program accurately. provides additional information.

I don't think anyone would claim the CaBi (as it is known) program represents revolutionary change and perhaps most present cyclists here could think of better infrastructure support to spend the initial 6.8 million US dollars on for their benefit as riders. And the numbers I have seen don't represent a significant change in how trips are taken. Still, the public relations benefits, which are difficult to quantify, seem pretty good so far.

Cycling culture here is unlikely to look like the Netherlands any time soon, but CaBi helps push things ever so slightly in that direction by encouraging daytime riding by folks for more modest distances in street clothes. Until CaBi started in September, I rarely saw such riders in my longer distance commuting, wearing cycling clothes. Now I see these folks riding CaBi bikes as I get towards DC , and they are always in street clothes.

A neighbor said his office has purchased several annual memberships and they are shared by staff (I don't know if CaBi intended to support that, but it works) so that they can ride around downtown between offices in different buildings, replacing pricey six block cab rides. Again, while hardly revolutionary, it is pretty unusual to see Americans in business attire riding in downtown in the middle of the day. The convenience of CaBi (admittedly, depending on the distance to the nearest station) trumps having bikes in some office parking lot, it seems.

CaBi is getting good press - practically anything that gets good press for cycling is hard to view as negative around here. And it is becoming part of the transportation landscape (perhaps not the best phrase) - in the Washington Post, a recent question on how Lady Gaga should arrive at her concert here included the suggestion of arrival by CaBi bike (since it is "so green").

The fiscal situation in Washington DC at present is dismal. The sums needed to fix most problems dwarf the capital investment to create CaBi. At the same time, the strong perception is that the program works and it is growing much more quickly than anticipated. It is a success - a bicycling success! Six million dollars of cycle paths would not, could not, have the same effect, I think, in raising awareness of cycling, of associating cycling with positive outcomes.

And even if they weigh a ton, they are kinda cool looking.

Olaf Storbeck said...

that's really an interesting post but I think your numbers are misleading because you're partly comparing apples to oranges. See my own blog for more details:

David Hembrow said...

Olaf: You make a mixture of good and not so good points.

It's reasonable to say that the scheme currently only covers (roughly) "zone 1" of London at present. However, what has happened is that they've covered just the easy bit. This is the most dense part of the city, where the scheme has the greatest chance of working. I doubt it will ever be expanded to cover the entire city, because what is already the most expensive single cycling project in the history of Britain would be enormously more expensive if it were expanded to cover the whole of the capital.

I don't think it reasonable to argue that usage is higher than Transport for London currently claims, based on another "huge number" throw away comment from earlier last year. If they had reached 2.5 million journeys earlier, I'm sure their over-active marketing department would have told us about it. They're not shy.

Your calculation where you try to show that the rate of usage is as high as I estimated it could be as a maximum is flawed. That calculation is based on taking the entire city. The maximum capacity for part of the city is much greater. It's still falling behind by the same proportion as I said it is.

I'm not against the idea of these shared bikes as such. However, this is actually an astonishingly ineffective way of spending a lot of money. It is easily the most expensive cycling project in British history, but the results are far from impressive given the expenditure. You do realise, don't you, that dividing the 140 million pound expense so far by the 2.5 million journeys, you'll find that each journey so far has cost TfL 56 pounds. A trip to the pub and back costs TfL 112 pounds.

Olaf Storbeck said...

David, thanks for your answer. My thoughts are too long and are hense not accepted by your website. You find them here:

John! said...


Thanks for exploding the myth.

Some other things not in press releases about 'Boris Bikes - the nature of the scheme seem to have encouraged much inconsiderate cycling (on the pavement/sidewalk; against the flow on one-way streets; stright through red lights; and aren't these little 'lights' illegal in the UK?

I live less than five miles from central London - you never see a Boris bike there. As it's all just a hug promotion for Barclays Nasty Bank,I can't say I'm upset.

Neil said...

@John! - which little lights are you referring to?

Note that AFAIK it's not illegal to have non-legal lights on a bike if you also have legal lights. And obviously that only applies after dark.

Anonymous said...

the london cycle hire scheme exists for one reason - increasing capacity on the underground is incredibly expensive - and capacity is all about morning and evening peaks. even small reductions in usage at these points (which is when the cycle hire is most used) can help reduce the need for very expensive investment in more trains, better signaling etc.

of course, better cycle infrastructure would do a better job. but this would involve taking road space away from cars - which tfl is never prepared to do.

Colibri said...

Hi David,

In the same vein as the Barclays hire scheme and superhighways, what do you think of the "Cycle To Work" scheme in Britain?
They put forward nice numbers of CO2 saving (so what?) but they don't mention how much it cost and if it really put new cyclists on the road.
Speaking of "road", when I see the main image on the frontpage of their website (i.e. young, fit, dressed-up, helmeted people with a huge SUV behind), I'm kind of worried...

I'm asking this because over here (France), officials are taking this scheme as an example, promoting it as the great idea that will change it all...


David Hembrow said...

Colibri: It sounds like yet another case of "looking in the wrong direction".

While the cycling rate of France is low, that of Britain is even lower.

No-one sensible would look to the UK to see how to promote cycling. Britain is not an example of shining success in cycling.

However, this is often how things seem to be. Mostly, Britain likes to look the America (I'm writing an article about this will will be live on Monday). However, the hyped bike hire in London is the result of London copying the hyped bike hire in Paris. Now the French appear to be about to return the favour by copying the UK.

What none of these nations ever seem to do is to look to the successes of the most successful country for inspiration.

Reaperexpress said...

Toronto is peculiar in that its (commuting) cycling rate varies wildly in different parts of the city. In New York, the rate generally varies from 2-6% in different areas, whereas Toronto varies from 0-17% (not including the car-free Toronto Island, which has a rate over 20%).

The overall cycling rate is actually fairly similar to the commuting cycling rates, based on vehicle counts on individual streets.

Here is a map of the cycle commuting rate in 2006. Since then there has been a significant increase in cycling, mainly in the downtown area.

It's certainly not spectacular, but it's pretty impressive how high the rate gets given how little infrastructure there is.

Anonymous said...

I hardly think it fair to criticize a bike scheme with only 6000 bikes simply on the basis that even at full possible usage rates It can only represent a small proportion of the overall mode share in city with over 7 million people.
Everything has to start from somewhere demand will grow as they make it more user friendly and make improvements to how they operate it. It's great that people visiting london and casual users now have a access to a convenient alternative means of transport.
The criticism also seems a little premature since it has only been running a relatively short time, no one can expect instant success without first gaining some experience in managing such system.

David Hembrow said...

Anonymous: You've swallowed the propaganda good and proper.

How would gained experience make a scheme which has inadequate capacity suddenly make an appreciable difference ?

How much growing of demand and improved user friendliness will make the bikes capable of transporting more people ?

Bike share systems simply cannot scale to the point where they have an appreciable effect on modal share, no matter how much hype their promoters produce.

They have failed to live up to the promises everywhere that they've been installed, including in London.

These schemes cost way too much, and don't address the issues which prevent people from cycling. They might be nice to have in a city which already provides a good environment for cycling, but they're not the right way to invest funds in a city like London which still has such appallingly badly designed streets.

Multiparty Democracy Today said...

To even come close to what London needs of a bikeshare system, it would already need an effective system of bicycle infrastructure. Main streets with cycle tracks and protected intersections/simultaneous green (good chance the latter could be used given how many cyclists would make right turns I imagine in a city this dense), gateway treatments at minor side roads, quiet, traffic calmed 20 mph zones on every other street, and lots of shortcuts for cyclists and concentrating car traffic and diverting them, and adding efficiency, like using single lane roundabouts with separate cycle paths around it, left turns that can be made regardless of what traffic lights are showing, etc. The details need to be paid attention, like making sure cycle tracks (one way) are at least 2 metres wide, 2.5 metres if possible, (two way), 3.5 metres minimum (3 metres on secondary routes), 4 metres if possible. Bus stop bypasses which do not constrict or slow you down, and angled curbs forgiving errors and optically widening the cycle path. No plastic armadillos.

And the bicycles themselves used in London for the Boris Bike scheme could be better. A wheel lock and chain like the OV-Fiets bikes do allows you to pop into a shop and grab a chocolate bar or coffee for example would be good to have. An enclosed chainguard makes them more reliable. Much more reliable. This is why most Dutch bikes have one. A rear rack would be very helpful. You could even carry another passenger if it was as strong as most Dutch racks are. It might be useful to have 7 gears (on a hub gear of course) rather than 3, though this is probably not required. The scheme could have more of a focus on the fact that once most people own a bicycle themselves in accommodation of modern Dutch buildings or at least neighbourhood bicycle parking facilities as are popular in Amsterdam, then most people will have a easy to use bicycle that they can use for practical purposes. But when you need to go further, a bikeshare hire station could be located at bus stops, train and transit rail stations, and maybe even the gondola terminals. Bike parking for the privately owned bikes are of course important, the main rail stations of London should have parking for at least 1000 each, with some of the very important ones could hold 3000-20000, which by the way is a range that is common in the Netherlands. Thousands and thousands of double stacked racks are common in underground parking stations. Ordinary racks, or maybe bike lockers, could be used too, at bus stops and less important transit terminals. The Dutch use their OV-chipkart (hope I am spelling correctly) to subscribe to both the bus, train and OV-Fiets system which is the closest equivalent, especially if you modify the Boris bikes to my design principles. You can subscribe to a combination of or all of them. You could change the way to rent a boris bike so that you tap a card against the rental terminal, take out the key which is locked until you rent, unlock the wheel-lock, adjust the saddle to the right height and put your belongings in the rack (actually do these things before you rent so you get best use of your 30 minutes of free time), and then ride on the very convenient and accessible bicycle infrastructure, which is the most important part of good bicycle cultures.