Bike sharing has a lot of benefits. In addition to providing an alternative mode of transport, saving users money and travel time, increasing access, making roads safer and encouraging bicycle ownership; it reduces GHG emissions, is good for public health and reduces congestion. The last three should not be surprising since bike sharing results in a net modal shift from driving.
The Phase II member survey results show that bikesharing is causing a diverse array of modal shifts within the different cities surveyed.
The survey also found that bikesharing reduced respondents driving by large amounts in all cities. In Montreal and Toronto, 29% and 35% reported driving less. In MinneapolisSaint Paul and Salt Lake City, 53% and 55% reported driving less, and in Mexico City, 53% reported driving less. Very few respondents reported driving more.
And since driving is a significant cause of GHG emissions, bad for public health and a cause of congestion this is exactly the outcome we'd expect.
I'll note that some of these effects are counter-balanced by people who shift from transit and walking. Those shifts involve more complicated trade-offs, with smaller impacts and are not as consistent as shifts from driving. For example, with walking
More respondents in Mexico City, Minneapolis-Saint Paul, and Salt Lake City increased walking than decreased it. In Montreal and Toronto, more reported walking less often than more.
Similar complexities are found with transit and biking. But those shifts don't change the fact that bikesharing reduces GHG emissions and improves health.
So it was surprising when the Rupert Murdoch-owned The Times recently had a disturbing headline about bikesharing "Boris bikes don't improve health or reduce pollution." It was also wrong. But there are layers of wrongness built into it. It's like a game of telephone where each person is not only getting the statement wrong due to normal errors, but each is also acting in bad faith.
The Times was a little wrong
The Times article was basically regurgitating a press release from the Royal Geographic Society about a recent study on bikesharing, so when it was quoting it - which it did often - it was at least getting that right. But even then it got the main things wrong.
They state that the study shows that bike sharing schemes "do not cut carbon dioxide emissions" a claim highlighted, with the most damage, in the title. But that's wrong. The claim from the RGS study is that "Bike share schemes ... are not very effective at ...reducing CO2 emissions" which is not the same thing as doing nothing. I'm not very effective at cleaning dishes, but they aren't as dirty when I finish as when I started (sorry everyone I've ever shared a house with).
The article goes on to claim that bikesharing bikes "aren't having a positive effect on public health" and that "people on the outskirts [of cities], who are likely to be on lower incomes and in greater need, do not benefit." Each of these statements wrongly represent what the RGS press release says. As the press release says there aren't very effective, not that they're doing nothing. The underlying studies make it clear that they DO have benefits for health and CO2 reduction and that they do benefit people with lower incomes.
The RGS press release was a little wrong too
As noted, the press release from the Royal Geographic Society "Bike sharing schemes mostly benefit healthy, wealthy, young white men" makes different claims than those in the Times article. Mostly that it exaggerates the claims from doing "little" to doing nothing.
But even the press release doesn't match the work by Dr. Médard de Chardon. [The press release is about the presentation of a study by Médard de Chardon. It doesn't say which one, but much of the territory is covered in his study on "Bicycle sharing system 'success' determinants" so I'll be using that in this post]
The press release states that
- The total amount of carbon produced in London from the rebalancing vans is not offset by the amount saved from use of the bike share scheme.
- They are not very effective at improving health
- They are not very effective at lessening road congestion
- They are not very effective at promoting transport equity
But here again the claims in the release state things with more certainty than they're stated in the paper and they elevate the negative results of outliers without noting that they are the exception. For example on item 1 above, the paper says.
Multiple studies have shown that publicized estimates of carbon dioxide reductions are often overstated as only a small portion of car trips are replaced using Bike sharing systems (BSS) (Ricci, 2015). In the case of London it is estimated that the vehicles rebalancing bicycles within the system may surpass any emission reductions from modal shift (Fishman et al., 2014a).
So the rebalancing vehicles MAY offset emissions, but they may not. Furthermore, of the four cities studied by Fishman, London is the only city where they found that mode shift didn't offset re-balancing, so it seems deceptive to make that the "example."
The claim that they aren't very effective at improving health is also not backed up by the study. Instead it says:
The shift from sedentary travel modes to cycling has clear health benefits but net quantities are overstated due to the reduction
in walking, which has greater health benefits for a fixed distance traveled.
Overstated is not a synonym for "ineffective."
On congestion, again the PR gets it wrong. The study calls the claims of congestion reduction "unproven", it does not say that they are proven in effective.
Finally, on equity, the study does make the claim that bike sharing "members [are] more likely be wealthier, younger, white, male and own a car, compared to the local population" and that it is "one of the most inequitable forms of sustainable transportation infrastructures.” So one out of four isn't bad. Perhaps this is the part he should show as an example. [More on the equity issue below]
The release also asks whether a bike-share system in which each bike is used less than twice a day (as many are) is the best use of public funding for cycling. That's a great question, but he's not an economist and he, as near as I can tell, he doesn't even try to answer it.
Then he really goes off the rails.
“In reality, bike sharing schemes are a false solution. They look sophisticated and are technologically cool, but they don’t create much useful or progressive change.
It’s worrying that we are getting bike share schemes instead of concrete improvements to transport infrastructure.”
A solution to what? His own studies show that they have many small, but real benefits. That seems to indicate that it's at least PART of the solution to some of the problems he brings up. And if it doesn't create much useful progress, the question is compared to what? He's not arguing that bike sharing is a bad thing, only that other things are better. That's always a complicated question and he makes a poor case that something else is.
I do agree with him when he says that it's important to more effectively redistribute public space for better cycling infrastructure, and if we have to make a choice between street changes and bike sharing (as they did in London) it's wise to ask which is the better investment. But this press release, and the 2017 study, don't do that.
The study is wrong too, and isn't really a study of 'success determinants'
Again, I'm going to use the 2017 study here and I recognize that he may have a new study out which would make this wrong.
Médard de Chardon's study is really a study of how many Trips each Bike takes per Day (TBD) in various systems (Of the 75 systems studied CaBi is 22nd with 3.0 TBD, 3rd best in the US). In that sense it's useful and interesting.
Where it gets off track is where it tries to push an agenda - namely that because many systems have a low TBD, bike-sharing is a poor use of resources.
He first claims that "success" for bikesharing is not defined. And for that he cites a paper that states that
Whilst predominantly enabling commuting, bike sharing allows users to undertake other key economic, social and leisure activities. Benefits include improved health, increased transport choice and convenience, reduced travel times and costs, and improved travel experience. These benefits are unequally distributed, since users are typically male, younger and in more advantaged socio-economic positions than average. There is no evidence that bike sharing significantly reduces traffic congestion, carbon emissions and pollution.
But, at least for Capital Bikeshare, that's not true. When the program was starting, MWCOG submitted an application to the US DOT for a TIGER grant. In that they clearly stated what they saw as the benefits/goals.
The benefits they foresaw were user cost savings, user travel time savings, increased access, congestion reduction, emissions reduction, healthcare cost savings and accident reduction. They also saw benefits in getting users to purchase and use their own bikes - something that just doesn't show up in tpbpd.
Then they calculated the value of each trip at $1.20. So as long as the subsidy is less than that, it would seem the system is a success. I did the calculations after one year and showed that it was. Now if Médard de Chardon wants to reanalyze the MWCOG calculations, update them and/or make them more robust that would be great. But he doesn't. And I'll note that of the 9 benefits listed here, his paper misses 5 of them.
He then goes on to state that the benefits for road and public transit congestion, carbon emissions, cycling modal share, health and equity have been shown to be hard to measure, trivial or non-existent. But the studies he cites contradict that. One is quoted above listing all the benefits. Another merely says that "the majority of scheme users are substituting from sustainable modes of transport rather than the car" which is not the same as saying there is no mode shift from cars or even from modes that do less for health. The third, written before CaBi started doesn't even make that claim. But even the most damning of these, by Ricci, which says that "there is no evidence that bike sharing significantly reduces traffic congestion, carbon emissions and pollution." (emphasis mine) also notes that nearly 20% of all bikeshare trips were shifted from car. There is no explanation for how hundreds of thousands of trips can be shifted from car to bikeshare without reducing congestion, emissions and pollution. It seems the word "siginficantly" is doing all the work. Médard de Chardon does cite a study that claims that, in the case of London, the low car-to-bike shift rate combined with much higher than average rebalancing miles, leads to an increase in carbon emissions. It makes the case that in a city where few drive, bike-sharing will do little to reduce driving.
But one thing Médard de Chardon does is cherry-picks his studies. If one claims that it does little to reduce emissions, but does improve health. He'll cite it when he talks about emissions and use another when he wants to talk about health. For example, he cites one study to make the claim that
Additionally, women using London’s BSS have reduced health benefits, compared to men, owing to increased rates of injury
Which sounds negative, but that studies conclusion is that
London’s bicycle sharing system has positive health impacts overall, but these benefits are clearer for men than for women and for older users than for younger users.
Which sounds less negative.
When he notes that the systems have benefits
They provide an alternative mode of transport, increase accessibility, trip resilience and flexibility, lower the barrier to exploring urban cycling, increase the visibility of bicycles, bicycle awareness by drivers and normalizing the image of cyclists in casual clothing
He then attacks those benefits on the grounds that they "do not spread evenly among classes and race," which is a way to minimize them. For that he relies on Melody L. Hoffman's "Bike lanes are white lanes."
There may be lessons to be gleaned from that book, but we should be clear that it's not a "study" of equity, it's not based on such a study and when she writes about bike sharing she's talking entirely about one system - Minneapolis' Nice Ride - and almost exclusively relying on interviews. So her claim that bike sharing is "one of the most inequitable forms of sustainable transportation infrastructure" in the United States; that's her opinion. It's not based on some metric of inequality and a comparison of various "forms of sustainable transportation infrastructure."
It is true that there's an equity issue in bike sharing, but that is true within biking as a whole, and so if the argument is that we should not invest in bikesharing to instead invest in protected bike lanes, you're not really solving the problem. It's also true that Médard de Chardon's study does a bad job of making the case about equity. And it makes no attempt to show why a system which unevenly spreads benefits among classes and races does not benefit those with lower incomes. It's possible the benefits are uneven and that low income people still benefit.
Finally, he makes a dubious claim that 1 TBD is a significant threshold since.
This value is psychologically important as systems below this have some bicycles being unused each day. More worrisome are the 10 systems with ratings below 0.5 TDB, as this means most bicycles are not used on a daily basis
It may be that 1 TBD is so low that the costs exceed the benefits. Or that 0.5 is. But if so, it won't be for psychologically symbolic reasons. And it likely won't be universal. In an un-system like New York's, where the taxpayers pay nothing, who cares how low the TBD is? And as he notes, any attempt to maximize TBD run counter to efforts to increase equity.
Just to pile on a bit, he also ignores the number of people who are encouraged by bikeshare to buy and ride their own bikes.
This is the heart of his failure to define bike sharing as a false solution. He has no idea what level of TBD/subsidy is cost-effective. He has defined no set of metrics for defining benefits. He makes to attempt to set a cost on systems. It may be that London's system increases emissions but is still cost-effective.
It would be fair of him to say that systems need to define goals and metrics for success. That they need to analyze those regularly to see if those goals are being met and to what extent. That they need to try to define those metrics as benefits and compare them to costs. But he's simply not in a position to call bike-sharing a false solution. Certainly not across the board, and not in a city like DC with 3.0 TBD and high car-to-bike mode shifting.
So, I would love to see a comprehensive cost-benefit analysis of bike-sharing across a broad spectrum of cities. But this isn't it. And without one I don't see how anyone can claim that they are a "false solution."