Study claims bikesharing doesn't do much for congestion or air quality, but has other benefits

A recent study claims that while bike sharing has some real benefits such as improved health; increased transport choice and convenience; reduced travel times and costs; and improved travel experience; there is "no evidence" that it reduces congestion, carbon emissions or pollution. In addition benefits tend to go to younger, more socio-economically advantaged males. This is particularly troubling since some places, DC in particular, have utilized Congestion Mitigation and Air Quality Improvement money to fund bike-sharing; and Montgomery County has tapped into a Jobs Access and Reverse Commute grant which is designed to help welfare recipients and low-income persons seeking to obtain and maintain employment. If Miriam Ricci, a researcher at the University of the West of England’s Centre for Transport and Society, is correct about these claims, that might put future funding from these programs at risk.

I haven't read the study since it's behind the paywall and I'm not made out of money, but a NextCity article on it has more details than the abstract does. It doesn't appear that Ricci gathered any new data, but rather relied on "existing studies and surveys", which I don't say to make it sound invalid. 

Ricci found that bike-share users don’t bike instead of driving so much as they bike instead of taking transit or walking. She writes that, “although Dublin bikes users reported considerable behavioral change, the prevailing trend showed a large modal shift (80.2 percent) from sustainable modes of travel to the bicycle, particularly from walking (45.6 percent) and including transfer from bus (25.8 percent) and rail (8.8 percent).” Still, nearly 20 percent of Dublin bikes users say they now drive less. Other European and American cities saw far lower rates of mode shift. In London only 2 percent of users shifted away from cars. In Lyon, France, and Washington, D.C., it’s 7 percent.

If each bike causes 22-28% as much congestion as a car, depending on how that compares to transit or walking, when a person moves from transit or walking to biking that might increase congestion - of course it's more than offset by each person who moves from driving.  Another source of congestion could be the rebalancing:

bike-share can actually contribute to congestion with the vans and trucks they use to redistribute bikes in the system. Those motorized fleets also harm bike-share’s environmental benefit.

I know that in 2011-2012, CaBi reportedly drove their vehicles about 140,000 miles, a lot of it outside of congestion time periods I suspect. So the contribution to congestion and pollution by rebalancing might not be that much. But these are good questions to ask, still when they say "Ricci found zero evidence that bike-share leads to any significant reduction of carbon emissions" I wonder what is "significant," because if there's significant enough benefit from a health or transportation choice standpoint, then those benefits - though "insignificant" - are still gravy. And they might be enough to continue using CMAQ and JARC money (something Ricci probably doesn't care about).

Ricci argues that bike advocates should focus on the benefits for which more compelling evidence is available. 

According to Ricci’s paper, a recent survey of London bike-share’s active users found that 78 percent started to ride or ride more as a result of the system. Similarly, 68.4 percent of sampled bike-share users in Dublin claimed, “not to have cycled for their current trip prior to the launch of Dublinbikes” and 63.4 say they purchased a private bicycle after using bike-share.

Ricci also found encouraging evidence that shows bike-share is good for user health...

Which is all well and good, if bike-sharing systems are content to find other funding. 

I suspect that a more detailed study - focused on the US, where there is more driving, instead of Dublin - would show that there are modest, but real, congestion, air pollution and transportation equity benefits from bike-sharing. 

Update: A related study adds to the common sense claim that active commuting is good for your health

Despite evidence that more compact settlement patterns enable active commuting, only a small share of workers in these areas choose to walk or bike to work. In general, the activity level of residents in more compact cities and residents in more sprawling areas is very similar. But, there is a robust association between active commuting and lower body mass index that is not explained by unobserved attributes or preferences suggests that policies to promote active commuting may be effective. In particular, active commuting has a greater effect on BMI. Consequently, compact settlement appears to be an effective infrastructure for promoting more active lifestyles. The policy challenge is finding ways to ensure that this infrastructure is more widely utilized.

Comments

[charlie]

Well, the one I like is that 30% of bikeshare users lost weight! That goes along with the self-reported BMI of white people with a 202 area code having such low BMIs to begin with.

(In other words, white people with a 202 area code are more likely to lie on the phone to a random interview on a status question like BMI and weight. Or income.)

But yes, the use of CMAQ money for bikeshare is a sham. The JARC is even a bigger joke.

I'd suspect that for CABI, a larger percentage is shifting from rail than in other cities. Shifting from bus is also a bump but not as large.

I doubt the rebalancing vehicles affect congestion at all. As long as they aren't diesels they won't do much on air quality.

Right now bikeshare funding is a good way to encourage biking in general. If you could combine bikeshare stations with bike lanes that would be better, but station placement continues to be less than optimal. In the future, continued funding on bikeshare may not be optimal for increasing or maintaining bike share usage.

[Joe D]

It would be more relevant to US cities if, as you mention in your very last paragraph, they study cities that don't already have a significant pedestrian modeshare.

[Jeffb]

I'm not surprised but if you want to reduce driving then you have to provide alternatives and bike share can be part of the solution.

Presently the marginal utility of driving is greater than the marginal cost . So simply offering an alternative is not going to affect mode share much.

If we want to reduce driving we are going to have to increase the disincentives - be it gas taxes, parking costs or a congestion charge.

I do wonder if having bike share, and other alternative transportation, affects car usage over time. In other words do they influence people to remain car free/lite longer?

[ACyclistInThePortCity]

An interesting article, but the key limitation in apply results largely derived from European experience to the US, as Dave has pointed out, is the lower transit share in US cities. I would also suggest that in US cities bike share can add to placemaking (encouraging density which also encourages more walking and transit, and shorter auto trips, and lower auto ownership) which is a much, much, lesser issue in European cities with bikeshare.

[Smedley Burkhart]

Interestingly, it is the up-front, not the marginal costs that seem to be affecting transport patterns around here, with millenials deciding that car ownership isn't worth it.

In other news, bike lanes (with sharrows) in Pyongyang: http://www.reuters.com/article/2015/07/14/us-northkorea-bicycles-idUSKCN0PO06Q20150714

[Crickey7]

I think the effect is a bit more indirect. Bike share fits into other sharing and transit transportation choices that allow a segment of the population to forego buying a car. If you don't have a car, then you're not choosing between driving and biking, you're choosing between non-car modes. But without bike sharing systems, they would be less likely to be car free.

[DE]

Probably a typo, but what's "driging"?

I wouldn't really have expected bikeshare to reduce emissions very much because those heavy pigs aren't made for long hauls; therefore, they're less likely to replace car trips than they are to replace shorter trips like might be done by walking or bus.

Your own personal bike, however, can easily replace a 20-mile trip and can therefore dig into the emissions somewhat more. And perhaps getting people started on bikeshare will then move them on to their own bikes and longer trips.

[washcycle]

Crickey7, I agree and another part of the problem is 2nd-order effects. To what extent does moving young men off of transit and onto bikes make more room for others on transit, thereby increasing the use of transit? I have no idea. I don't even really know how to figure that out. We'll call it a known unknown.

[DE]

All this stuff is incredibly hard to measure. I know that I ride 4,500 miles a year that I would be in a 29-mpg car in stop-and-go DC-area traffic. You can convert that to definite emissions reduction values, CO2 and otherwise, but you can't easily determine that kind of number for the greater cycling population because it's hard to know how they all would have behaved if they didn't ride. Then you have the ripple effects from that, which are even more intangible.

I see washcycle changed that word I couldn't figure out to "driving." I look and feel like an eejit, but that's par for the day.

[Jonathan Krall]

Congestion, barf. In a large modern city, people will always pile on with the cars until the alternatives are less painful. Adding transit/biking/walking shifts the equilibrium, but people will always pile on with the cars to the point where they find themselves complaining about congestion.

We can add lots of transit capacity, but cars, when not gridlocked, will always be very convenient. Thus, absent congestion pricing, people people will always pile on with the cars until they find themselves complaining about congestion. Any politician with a clue knows this.

All this is by way of saying that the choice to measure transportation using a congestion metric is a highly illogical thing to do (why measure something that you know isn't going to change?). The exception is where transportation capacity (public transportation) is so weak that congestion is pathological, a situation that occurs nowhere in the USA.

Politicians have a reason to wade into these illogical waters. They have to respond to the illogical demands of their constituents. Hence they babble about congestion while using CMAQ money to add capacity. This is a smart thing to do and I appreciate it.

Researchers, on the other hand, should measure something that matters. My suggestion s that they measure the number of trips. As others point out, bikeshare makes transit more attractive and helps with the decision to go car free or car lite. A newly car-free person can keep moving while opening up a slot for a some other clown to drive[1].

Presumably the number of trips relates to economic activity, regardless of the length of the trip. I'd prefer that researchers focus on this metric.

[1]http://www.mrmoneymustache.com/2013/04/22/curing-your-clown-like-car-habit/

[ACyclistInThePortCity]

JK

I respectfully disagree. Even absent congestion pricing, and holding constant for car ownership, marginal trips are costly. At some point driving to a more distant but cheaper store, say, is not worth my time and the gas and wear and tear on my car. So I will not due it even if the road is free flowing. That is why not all roads become congested. And why, AFAICT, the studies of induced demand show that the amount of induced demand from an addition in capacity is under 100% of capacity.

Ergo, looking at congestion is not illogical (I agree that reducing congestion is an illogical use for CMAQ fund though)

[washcycle]

How is congestion in Amsterdam and Copenhagen?

[Jonathan Krall]

"marginal trips are costly"

Not costly enough to keep congestion from being a perceived problem. Otherwise we wouldn't hear about it so much.

"studies of induced demand show that the amount of induced demand from an addition in capacity is under 100% of capacity."

I'd be curious to see a study of induced demand that differentiated a city like Washington DC, that has a large pool of potential drivers riding the Metro, from one without that pool. My hypothesis is that congestion will be relatively constant in DC because of that large pool of potential drivers.

Over time, it is possible that increased attractiveness of Metro or increased driving pain (see below) will reduce congestion, but I doubt that driving in DC will become any less of a PITA than it is now. And, if they don't do something about the Rosslyn tunnel bottleneck, then Metro could become less attractive, pushing people to cars. In any case, the gains come from focusing on adding capacity (transit), shortening trip distances (planning), and improving the user experience for most users (in DC, that leans towards people not in cars).

The main reason I am speaking up is that planners/engineers/decision-makers tend to optimize what is measured. Since the future is in transit, we should measure something other than cars.

"How is congestion in Amsterdam and Copenhagen?"

I don't know. I suspect that it is less than in DC because they've introduced a congestion cost that is largely absent here in the form of having to wait for all those bicycles to pass before proceeding in many circumstances. Always, people will shift their habits to avoid pain. But, really, I don't know.

[ACyclistInThePortCity]

"Not costly enough to keep congestion from being a perceived problem. Otherwise we wouldn't hear about it so much."

No, nor did I say it was. Congestion is a problem on select roads, at select hours of the day. But it is sometimes possible to relieve congestion (ceteris paribas - over time of course growth will offset) by adding highway capacity, and often possible to relieve it with transit, demand management, and active transportation. Fully pricing auto usage would be a better policy, but that does not mean that congestion should never be a metric. If I am comparing say, bikeshare expansion, to added PBLs, it is quite reasonable to use congestion relief as a metric - since not all of that will be eaten up by induced demand.

"I'd be curious to see a study of induced demand that differentiated a city like Washington DC, that has a large pool of potential drivers riding the Metro, from one without that pool. My hypothesis is that congestion will be relatively constant in DC because of that large pool of potential drivers."

IIRC FTA did such a study almost 20 years ago, and it found that on roads directly parallel to transit lines, the speed of the road remained in a fixed relation to the speed of the transit line - IE the substitution effect you hypothesize is correct (and suggests that you can speed up traffic on the parallel road by improving transit speed and in other ways making the transit line more attractive) However even in greater DC, most roads are not directly parallel to a metrorail line.

"Since the future is in transit, we should measure something other than cars."

A. I am not sure what the future is in transit means. I doubt we will have 50% plus transit share nationally in a hundred years. We may have it inside the beltway, but it will be some time, and is by no means certain, even for commute trips.
B. By all means we need other metrics. We need a complete streets LOS, in which motor vehicle LOS is only one component. And we have other things we need to measure as well.

[ACyclistInThePortCity]

Note, that effect, IIRC, was only when the adjacent road was highly congested, such that the transit line was competitive to time sensitive travelers. I cannot find the study online unfortunately.

[Jonathan Krall]

"I am not sure what the future is in transit means"

The number of people keeps going up while the number of vehicle miles traveled has been flat for about 10 years[1]. That indicates a shift towards transit. In cities, there is not much more room for roads, so new transportation capacity means adding transit. In DC, "drive alone" is still the biggest commute group but they are a plurality, not a majority. In Alexandria, "drive alone" is below 60 percent.

Often, leadership means seeing where things are going and getting out in front. Given that new transportation capacity will be mostly transit, I'd like to see more focus on making transit work well. I'd like to see metrics that focus on people instead of cars. To me, focusing on congestion seems like pandering instead of leadership.

[1] http://usa.streetsblog.org/2013/12/17/study-transpo-agencies-are-terrible-at-predicting-traffic-levels/

[Miriam Ricci]

Thanks for writing about my paper, which is free to download until 19th July from:
http://authors.elsevier.com/a/1R6t47sdbMZRLC
If you have problems with access please email me at [email protected] and I'll happily send you the paper electronically :-)
Follow me on Twitter on @RicciMiriam

[ACyclistInThePortCity]

"The number of people keeps going up while the number of vehicle miles traveled has been flat for about 10 years[1]. That indicates a shift towards transit."

The growth of transit ridership is far too small to account for more than a tiny percentage of that. It is fewer trips - either because of reasons associated with family formation issues, or greater use of the internet to substitute for trips.

" In cities, there is not much more room for roads, so new transportation capacity means adding transit. In DC, "drive alone" is still the biggest commute group but they are a plurality, not a majority. In Alexandria, "drive alone" is below 60 percent."

I agree there little room for more road in cities. I did not realize that you meant 'the future in cities is transit" Thanks for clarifying. Even so the number of drivers in a place like Alexandria is large, and measuring a transit project, by its impact on those drivers, does not seem to me to be illogical.

"Often, leadership means seeing where things are going and getting out in front. Given that new transportation capacity will be mostly transit, I'd like to see more focus on making transit work well. I'd like to see metrics that focus on people instead of cars. To me, focusing on congestion seems like pandering instead of leadership."

I do not see how including impact on LOS is pandering. People are in those cars. I agree that it should not be the ONLY criteria.

But let's say we were comparing a transitway on Duke Street, to adding additional segments of transit only ROW to the West End Transitway. Assume, for the sake of argument, that time savings to transit riders was identical, and other transit related benefits were identical (and that cost was identical) But the Duke steet transitway reduced improved motor vehicle LOS on Duke (because of a a lot of lane capacity, say, combined with high diversion of drivers to transit) while adding more transit ROW to the West End transitway resulted in worse LOS on that corridor (because new riders were mostly diverted from walking, say, and there was a reduction in general lane capacity) In those circumstances I would prioritize the Duke Street project. It sounds to me like you would not. We would both be seeing transit as the way to increase capacity, and the priority for investment - but I would explicitly include improving motor vehicle LOS ALONG with other metrics as investment criteria.