Bike share programs: good feelings, bad economics?

My local mall does not provide short-term bicycle rentals. It also does not sell roast-lamb-and-mint flavour potato chips, or jeans in a size 32 inch waist/36 inch leg.

I would like to be able to purchase all three of these goods and services. For the last two items on the list, the intuition of the average person is the same as the intuition of the average professional economist: quit whining and deal with it. Most Canadian firms do not produce obscure potato chips flavours or clothes in unusual sizes because consumers aren't willing to pay enough for those items to cover the firms' costs. The benefits consumers enjoy from these goods and services are less than the costs of providing them, so it's not efficient to provide the good or service.

Yet when it comes to short-term bicycle rentals, people often reach a completely different conclusion: government should intervene to create a bike sharing program. All the world class cities have them: Washington, London, Paris, Beijing, New York, and Toronto.

Although each bicycle sharing program is unique, almost all receive public subsidies. Some cities provide in-kind support. Paris's scheme, for example, is implemented and operated "free of charge to the city", in exchange for granting an advertising company "rights to 1,600 advertising hoardings around Paris". The company also receives "space to allocate the cycle stations" at no cost. Other cities provide loan guarantees, as Toronto did for its partner's $4.8 million capital investment. In yet other cities, such as London, the bicycle share program is publicly owned. The city has ultimate responsibility for absorbing losses, although the program is privately sponsored and operated.

To justify such large public costs, one needs to point to some kind of public benefit, some positive externality created by bicycle sharing programs. To figure out what these were, I took a look at some feasibility studies for bicycle share programs, such as London's, New York's, and New Orleans's. The benefits of bike sharing identified across the studies are

1. Improved health through increased physical activity
2. Reduced pollution
3. Reduced congestion on roads and public transport
4. Shorter and more reliable journey times
5. Tourism and a strong business climate

These first two arguments for bike shares are easily dispensed with. While increased physical activity is a laudable goal, a study of London's bicycle share scheme found that the typical user is a professional male between 25 and 44 - probably not the demographic one would want to target in a physical-fitness promotion scheme. Although London's bike share scheme has been expanded into lower income neighbourhoods, those bikes - in contrast to the ones in more affluent central areas - are little used.

What about pollution? Bicycle sharing programs are a decidedly inferior solution to this problem. A much better option is to put a price on emissions through gasoline taxes and/or congestion charges. This allows each person to choose the way of reducing their carbon footprint - forgoing a trans-Atlantic flight, carpooling, moving closer to work - that works best for them.

The third and fourth arguments - shorter journey times and reduced congestion - are worth taking more seriously. What's the evidence?

A review of bike share programs (gated here) published earlier this year found that they typically do not take many cars off the road: "mode substitution from cars to bike share is low." People who use bike share are people who would otherwise be walking or using public transit.

Yet switching people from public transit to bike sharing may be a worthwhile. As the New York feasibility study put it:

"Bike-share programs, which typically can be introduced in a matter of months, can be especially valuable as New York faces increasing subway congestion and no clear, quick answers for relief..."

As an argument for bike share, this has some merit, but how generally applicable is it? That is, how many cities are there where (a) a bike share program would make a noticeable difference to public transport ridership and (b) it is cheaper to build a bike share program than to expand the public transport network? I've looked, but I haven't found a study that compares the costs and benefits of bike share with alternative strategies.

Even when increased cycling has the potential to relieve pressure on the public transport system, I would be interested in knowing how bike sharing compares to alternative methods of encouraging cycling, such as subsidizing private cycle rental operations, providing cycling lessons for people who do not know how to ride bicycles, subsidizing bicycles for low income children and families, subsidizing electric bicycles, improving cycling lanes, or creating safe and secure cycle parking facilities.

What about the final argument for bike sharing programs, that they promote tourism and a strong business climate? Based on my own experience, I'm skeptical. Even though I'm an enthusiastic and confident cyclist, in recent trips to London and Toronto, I've not bothered to use those cities' bike shares. It's just too hard to navigate around an unfamiliar city on a bicycle, to say nothing of the challenges of locating docking stations. The pricing structure of bike share programs tends to make them unattractive for tourists who want to spend a day or two exploring a city by bicycle. Another issue for tourists is the lack of locks on many bike share bikes.

At one level, I can really see the appeal of bike share programs. But as an economist, my instinct is to say "if you want to subsidize something, just write a cheque."

Comments

[Darren]

"No carrying and dealing with unwieldy locks."

"Unweildy locks"? I think you're making a big deal out of bike ownership problems that in my daily experience are quite trivial.

@Francis, perhaps we should triple the Bixi rental rate if you take the last bicycle from the rack :)

[Declan]

This post brought something forward that had been lurking in my subconscious, that the affinity of economists for blogging likely lies in the commonality whereby blogging and economics both seem to suffer from an arrogance that leads one to come to conclusions based on analyses that are far too simplistic to begin to even cover the topic at hand. I'm not meaning to pick on Frances here, all bloggers and economists seem guilty to one extent or another.

For example, in this post consider the 'ease' of dispensing with the notion that bike shares improve health by increasing physical activity because one study in one city noted that a majority of users are males between 25 and 44. Are we quite sure that men never benefit from increased physical activity? And what about the minority of users who aren't men (women, presumably), even if not a majority, there are still lots of them. And are we sure that the results of the study are accurate, and if accurate will remain that way over time, and are we sure it is the same on different places. etc.

And all those questions, any one of which might sway the 'economicity' of the program depending on our answer, all stem from just one point which was 'easily dispensed with'.

Once we try to enumerate all the layers of uncertainty surrounding all the possible benefits and costs, positive and negative externalities, network effects, infant industry questions, natural monopolies, 'irrational' behaviours, signalling effects, political feasibility, etc. etc., surely it quickly becomes clear that the question of the economicity of the program is unknowable. Of course, it doesn't hurt to discuss it, as long as those doing the discussing keep in mind that the reality is that we'll all decide to support or oppose the bike share program based on our gut feel or instinct, and the arguments about economicity are just rationalizations that we employ to support our pre-existing positions.

[Bob Smith]

No offense Declan, but you're guilty of the same crime of which you accuse Frances (and economists/bloggers) more generally - "arrogrance that leads one to come to conclusions based on analysis that are far too simplistic to being to even cover the topic at hand".

Of course the analysis is simplistic - it's a blog - not the Journal of Health Economics or the CJM. I suspect Frances' objective is to stimulate discussion for potential avenues of more substantive research. It's fair point to question France's reasons for dismissing the likely health benefits (I didn't follow her logic there). But I note that the subsequent discussion provided a better reason for suggesting that the health benefits are modest - namely that bike share programs don't significantly increase the use of bikes in daily travel - and some evidence to back it up. Moreover, one might also think that bike share programs aren't likely to have significant health effects if the people who use the bikes are likely to be the people who are already fit (i.e., people who ride bikes). Granted, those are only theories which can be proven or disproven (not by me), albeit reasoned and not facially ridiculous theories, but that's the starting point for any evidence-based policy discussion.

To suggest that the "economicity" of a program is unknowable because the facts are complicated is to give the kiss of death to evidence-based public policy. Ironically, it's a call to precisely to make policy decisions based on gut-feel or instinct.

[Shangwen]

Declan: "Are we quite sure that men never benefit from increased physical activity?"

This was not the point of Frances' comment. I agree with her that the presumed health benefits can be easily dismissed. According to the research that was presented by other commenters, the typical renter is someone who is already physically active, young (under 50), and very unlikely to be either a smoker or significantly overweight. Since the primary benefits of cycling are cardiovascular, the marginal reduction in risk of cardiovascular disease is negligible for that population, if not zero. If there is a serious argument being made about using scarce funds to make the greatest reduction in (for simplicity) cardiovascular disease risk, you would put it towards smoking cessation. (I'm not actually not sure how much lower we can go on that in the short term, but that's a separate question.)

Sure, there are plenty of variables in the story. But the evidence on effective allocation of money towards population health is out there and pretty clear.

[Frances Woolley]

What Shangwen said.

If the question is - are there health benefits? - the answer is probably yes. But there are health benefits to all sorts of things, for example, giving everyone a free pony. Or bicycle.

The question is ****how do the health benefits of this particular intervention stack up against the health benefits of other possible interventions*** for example, giving free bus tickets to seniors without driving licences to encourage them to get out of the house more often. Given that the users of Bixi bikes are already a fairly health demographic, I don't really see that there is a convincing case. At least, I'd like to see something more than handwaving.

[Shangwen]

Follow-up thought: if you want to discourage unhealthy behavior and maybe encourage more cycling, just up the rate of drunk driving convictions for low-income males. NPV>0. You can always spot those unlikely cyclists....

[Declan]

"No offense Declan, but you're guilty of the same crime of which you accuse Frances (and economists/bloggers) more generally - "arrogrance that leads one to come to conclusions based on analysis that are far too simplistic to being to even cover the topic at hand". "

No offense taken, I wasn't trying to claim moral superiority, just expressing dismay at the futility of this mode of discourse.

"To suggest that the "economicity" of a program is unknowable because the facts are complicated is to give the kiss of death to evidence-based public policy."

Not entirely, there are other cases (such as banning lead) where the evidence is clear, but I think sometimes believing we have evidence just gives us a false sense of security (arrogance) which leads to worse decisions than we would make if weren't sure we had the 'right' evidence based answer.