Advocacy & Safety - Cost/benefit analysis of improved cycling infrastructure

Here are some rough numbers that I just crunched, purely in terms of dollars, to see to what extent improvements in cycling infrastructure would be offset by decreased health care costs. Using the orders-of-magnitude technique pioneered by the famed Enrico Fermi, we can get an answer correct to within an order of magnitude while knowing little information.

First, the US spends an excess of around 100 billion dollars per year in terms of extra health-care costs attributable to obesity (I'm basing this on the CDC figures in 2003 of $75 billion, and rounding up to account for the increase since then). Also, we know that being more physically active has benefits beyond the reduction in obesity, and the roundup is supposed to account for that.

Now, let's say that a complete bike-friendly system were built, and that it would reduce the incidence of obesity by a modest 10%. That would mean that we'd have the equivalent of 10 billion/year to spend for breakeven. In 10 years, we'd have 100 billion to spend, to break even.

Also, let's say that the costs of construction are roughly $330,000/mi ($1 million / 3 miles). Sure, this number is highly variable, but when talking about the country as a whole, the number's probably not too far off. Let's also assume that the costs of maintainence are such that the equivalent of reconstruction needs to happen every 10 years.

$100 billion dollars -->300,000 miles worth of bike lanes. Since we have a total of 4 million miles of roads (let's round down to 3 million), we can have 10% of all road mileage made bike friendly with the money saved from the reduction in obesity costs, not to mention the other benefits derived from reduced pollution, more active lifestyle, etc. etc. If the lanes are built in strategic areas, having 10% of the least friendly, busiest roads made more bike friendly seems like it would help things substantially. (FWIW, famed Davis CA has bike lanes on 95% of its arterials and busiest collectors.) What is a typical percentage for the ratio of arterials+collectors to all roads?

Comments appreciated on this rudimentary analysis. Better numbers and so forth especially welcome.

Now, let's say that a complete bike-friendly system were built, and that it would reduce the incidence of obesity by a modest 10%. That would mean that we'd have the equivalent of 10 billion/year to spend for breakeven. In 10 years, we'd have 100 billion to spend, to break even.

Sure, "we" can "say that." Coming up with credible evidence of $1 one in reduction of obesity costs brought about by a bike-friendly system is a little more difficult.

Your assumption of a 10% decrease in the incidence of obesity due to the existence of improved cycling infrastructure is utter fantasy. 0.01% is closer to the mark.

I know whereof I speak, as I am one of those doctors who frequently treats obesity. If you can even clear 50% of the behavioral hurdles, you're doing great. Even then, there are adverse physiological changes, some irreversible, that come with the disease, making its reversal difficult to do by the mere introduction of more exercise.

Our friends in the Clydesdales subforum are the extraordinary exceptions. Many of them have overcome huge odds to achieve their goals.

We have plenty of bike infrastructure. We just need to do something to get the cars off it.

Just curious, but what happens then in a fat person who begins and continues to exercise? Assuming that the weight stays the same, is it not true that the person nevertheless becomes much healthier? And wouldn't this improved health lead to reduced (both direct and indirect) health-care costs? Thanks.


Your assumption of a 10% decrease in the incidence of obesity due to the existence of improved cycling infrastructure is utter fantasy. 0.01% is closer to the mark.

I know whereof I speak, as I am one of those doctors who frequently treats obesity. If you can even clear 50% of the behavioral hurdles, you're doing great. Even then, there are adverse physiological changes, some irreversible, that come with the disease, making its reversal difficult to do by the mere introduction of more exercise.

Our friends in the Clydesdales subforum are the extraordinary exceptions. Many of them have overcome huge odds to achieve their goals.

Also, let's say that the costs of construction are roughly $330,000/mi ($1 million / 3 miles). Sure, this number is highly variable, but when talking about the country as a whole, the number's probably not too far off. Let's also assume that the costs of maintainence are such that the equivalent of reconstruction needs to happen every 10 years.

$100 billion dollars -->300,000 miles worth of bike lanes. Since we have a total of 4 million miles of roads (let's round down to 3 million), we can have 10% of all road mileage made bike friendly with the money saved from the reduction in obesity costs, not to mention the other benefits derived from reduced pollution, more active lifestyle, etc. etc. If the lanes are built in strategic areas, having 10% of the least friendly, busiest roads made more bike friendly seems like it would help things substantially. (FWIW, famed Davis CA has bike lanes on 95% of its arterials and busiest collectors.) What is a typical percentage for the ratio of arterials+collectors to all roads?

Comments appreciated on this rudimentary analysis. Better numbers and so forth especially welcome.

Are you saying it takes 1 million dollars to stripe just three miles of pavement? Or are you talking adding width or paths? That is some expensive paint.

As a side note I am not sure bike lanes are the answer when the roads may still resemble freeways... But I know you are not discussing the "what" as much as the cost benefit ratio.

To that cost benefit ratio perhaps one should also consider this... that a 10 foot wide bike path should cost far less than adding 2 typical 14 foot freeway lanes. If highway dollars were also used for bike infrastructure, and the result was a 5-10% decrease in motorists using existing freeways, then the freeway was actually improved for a far lower cost then adding lanes to said freeway.

The $1 million/ 3 miles is for new construction. Say in the case where the roads originally are narrow already and require widening. You're right, in cases where existing roads can simply be reconfigured, the cost is much, much less.

I'm not advocating BLs on actual expressways. But BLs are nice for everything else that's busy, say on 45-55 MPh arterials.

You also bring up an excellent point with regard to bike lanes versus freeway lanes. Since even the best bike lanes are less wide than freeway lanes, if freeway congestion can be substantially reduced by getting people onto bikes, a reasonable case can be made to fund BL construction from freeway funds.

Overall, if it can be objectively demonstrated that well-planned and well-built BLs are cost-effective, as opposed to a cute "toy" beneficial to ~1% of the population, then there is a much better chance of getting the dollars needed to actually build them.


Are you saying it takes 1 million dollars to stripe just three miles of pavement? Or are you talking adding width or paths? That is some expensive paint.

As a side note I am not sure bike lanes are the answer when the roads may still resemble freeways... But I know you are not discussing the "what" as much as the cost benefit ratio.

To that cost benefit ratio perhaps one should also consider this... that a 10 foot wide bike path should cost far less than adding 2 typical 14 foot freeway lanes. If highway dollars were also used for bike infrastructure, and the result was a 5-10% decrease in motorists using existing freeways, then the freeway was actually improved for a far lower cost then adding lanes to said freeway.

Overall, if it can be objectively demonstrated that well-planned and well-built BLs are cost-effective, as opposed to a cute "toy" beneficial to ~1% of the population, then there is a much better chance of getting the dollars needed to actually build them.

I really agree with this... but use the term "infrastructure" not BL. BL are not always the best solution.

Frankly I think we need a bike infrastructure project similar to the 1956 federal highway act (http://en.wikipedia.org/wiki/Federal_Aid_Highway_Act_of_1956) that was the seed for our major freeway system.

I think, to reduce our dependence on foreign oil, we should fund a federal bicycle transportation act that would dramatically enhance the cycling experience in the US. This act should encourage cycling by enhancing existing roadways and adding paths and bike lanes everywhere and by discouraging motoring on pedestrian and cyclist shared roadways... all with the goal of reducing our national dependence on foreign oil.

Couple my anti terrorist (the second largest supplier of foreign oil is Saudi Arabia (http://www.eia.doe.gov/pub/oil_gas/petroleum/data_publications/company_level_imports/current/import.html)) goal with your health improvement goal, and there should be plenty of federal funding to make roadways more cyclist friendly.

Are you saying it takes 1 million dollars to stripe just three miles of pavement? Or are you talking adding width or paths? That is some expensive paint...

Our city is installing a new elevated rail line that will have a 3.5 km (2.2 mile) bike lane below. The lane is unique in that is higher than the street but lower than the side walk. It's cost is $2.7 million CAD.

The lane is also only on one side of the street and is for north bound cycling only.

Think of all the other things you could do for cycling with that money.

If this clearly high-end bike facility is just $200k more expensive than his $1 M/mile bogey, then I think the aggregate cost is significantly less than $1 M/mile.

I have lived in the city where this $2.7 mill bike lane is going in for 37 years and I don't think I have ever seen a cyclist on the road at this stretch.

I've seen a few on the sidewalk and I always see a couple of bikes in the various racks outside stores in the area, but never seen any on the road.

It'll be interesting if I see anyone use this lane when it opens. At $2.7 mill, those will be very valuable cyclists!

Our city is installing a new elevated rail line that will have a 3.5 km (2.2 mile) bike lane below. The lane is unique in that is higher than the street but lower than the side walk. It's cost is $2.7 million CAD.

The lane is also only on one side of the street and is for north bound cycling only.

Think of all the other things you could do for cycling with that money.

Like what?

Frankly I look at any and all new freeway structure and ask nearly the same thing...

Like what? ...

Gee, off hand I could say better (or more) enforcement of traffic law, more education of drivers and cyclists, how about advance signals at intersections for cyclists, a couple of simpler, longer bike lanes in areas where they are more likely to be used by cyclists, building traffic calming measures, more staff for cycling advocacy at city hall.

How's that?

To clarify, the figure I'm using is $1 M for every 3 miles. Still a lot of money, but not as outrageous as it might otherwise seem. And yes, paved roads are very expensive everywhere.

I'm sure it's been well-documented that roads and cycling facilities both demonstrate the networking effect. Namely, one 2 mile stretch is more than twice as effective as two 1 mile stretches, and as the network reaches places people actually want to go to, the level of usage increases exponentially until saturation. How well connected is that stretch of BL to the rest of the network?


I have lived in the city where this $2.7 mill bike lane is going in for 37 years and I don't think I have ever seen a cyclist on the road at this stretch.

I've seen a few on the sidewalk and I always see a couple of bikes in the various racks outside stores in the area, but never seen any on the road.

It'll be interesting if I see anyone use this lane when it opens. At $2.7 mill, those will be very valuable cyclists!

Like what?

Frankly I look at any and all new freeway structure and ask nearly the same thing...

Lots of bike parking for one.

How about a blinkie giveaway that targets the immigrant -- around here that would primarily be Latinos -- community?

Ohhhh ... getting fancy here and 2700 K might be too little ... how about a subsidized downtown commuter center with showers/lockers/safe parking?

EDIT: How about some surveys and/or live data collection?

... How well connected is that stretch of BL to the rest of the network?

not very. In fact, there was story in the paper this past week pointing out how the new lane cannot be connected to the existing network. If cyclists want to use this lane they have to brave some of the most difficult roads in town.

Just curious, but what happens then in a fat person who begins and continues to exercise? Assuming that the weight stays the same, is it not true that the person nevertheless becomes much healthier? And wouldn't this improved health lead to reduced (both direct and indirect) health-care costs? Thanks.

How many people, so obese as to create serious heath problems for themselves, especially middle aged or older, do you expect to begin or increase their cycling, for any reason? And increase their cycling/exercise regimen (because of improved facilities) enough to effect their overall health costs? Dubious proposition at best, if the facilities are in locations that are not generally favorable for year round cycling - do you really believe that bike paths will cause obese people to ignore hills and weather factors?

How many people, so obese as to create serious heath problems for themselves, especially middle aged or older, do you expect to begin or increase their cycling, for any reason? And increase their cycling/exercise regimen (because of improved facilities) enough to effect their overall health costs? Dubious proposition at best, if the facilities are in locations that are not generally favorable for year round cycling - do you really believe that bike paths will cause obese people to ignore hills and weather factors?

I think that this is the killer question. That is, who takes up cycling is anything but random. Moreover, you need to really consider the root causes of obesity.**

http://www.webmd.com/diet/news/20070820/obesity-virus-more-bigger-fat-cells

Now if you want to change your argument to something about preventing obesity you might have more traction. But it would make analysis much more complicated since you would need to consider the present value of the cost/benefit.

** I am no expert on the subject. But from casual reading, I get the sense the relationship between activity, obesity, and health is more complicated than typically reported.

Why are the costs of obesity more? Where are these short term calculations coming from? The only study that I've read about that tracked health care costs from birth to death showed that obese people and smokers died much sooner, and their health care costs per lifetime were $80,000 and $90,000 less than thin "healthy" people over the course of a lifetime.

Why are the costs of obesity more? Where are these short term calculations coming from? The only study that I've read about that tracked health care costs from birth to death showed that obese people and smokers died much sooner, and their health care costs per lifetime were $80,000 and $90,000 less than thin "healthy" people over the course of a lifetime.

Trouble is, during those shorter lives, obese people tend to be heavier users of the medical system, due to comorbidities.

If you have the citations, I would love to see the contrarian studies. I can kinda see the point that they are trying to make, that shorter lives = less resource consumption. I'm just not sure that applies to the medical industry, which has its own topsy-turvy economics.

I've heard it said too that smokers actually cost the system less, due to their shorter lives. As another poster noted, medical economics has some strange curves. Nevertheless it's very true that insurance rates for smokers are substantially higher than that for nonsmokers, and I'm sure there is a reason for that. FWIW, Canadian smokers pay extremely high cigarette taxes to pay for the expected increases in medical costs (Canada has a government-pay system) due to their smoking.

Anyway, maybe those who are extremely short-lived, and those who live extremely long lives end up costing the least. Centenarians typically are healthy more or less until the day they die, so they cost very little :lol:. My great-grandma lived to 99+, and she was healthy til the end.


Why are the costs of obesity more? Where are these short term calculations coming from? The only study that I've read about that tracked health care costs from birth to death showed that obese people and smokers died much sooner, and their health care costs per lifetime were $80,000 and $90,000 less than thin "healthy" people over the course of a lifetime.

such a silly point. People who require health care cost more than people who do not use health care. If someone can't understand that, what can they understand?

Why are the costs of obesity more? Where are these short term calculations coming from? The only study that I've read about that tracked health care costs from birth to death showed that obese people and smokers died much sooner, and their health care costs per lifetime were $80,000 and $90,000 less than thin "healthy" people over the course of a lifetime.

I'm trying to remember some of the authors. I seem to recall Kip Viscusi testifying on the topic arguing that it was an offsetting factor. Whether earlier death completely offset a lifetime of generally greater costs is not what I remember. But this is pretty fuzzy.

On another vein of research I recall that, roughly, most health care costs are incurred during the last years of one's life. So a "healthy" person just delays the extraordinary costs of emergency/intensive care. Since health care expenses have been outpacing general inflation, it might lend more weight to the aforementioned argument that earlier death offsets costs from "sin" behavior such as smoking.

Interesting idea. But what about the added healthcare costs that would be incurred as a result of injuries to cyclists from crashes and accidents?

I have read many times (which doesn't make it true) that the first and last year of our lives is when we use the most healthcare. Obesity would likely be reduced mostly for the next generations if there was biking infrastructure. Injuries to cyclists from crashes and accidents are much cheaper than car accidents.

Interesting idea. But what about the added healthcare costs that would be incurred as a result of injuries to cyclists from crashes and accidents?

who says the health cars costs to cyclists are any worse than motorists suffer?

who says the health cars costs to cyclists are any worse than motorists suffer?

Are you kidding? The injury and fatality rate per mile is much much higher for cyclists than for motorists. More injuries and more fatalities equate to higher healthcare costs.

Are you kidding? The injury and fatality rate per mile is much much higher for cyclists than for motorists. More injuries and more fatalities equate to higher healthcare costs.

OK, but why are these injury and fatality rates higher? Could it be due to cycling among autos? What if "improved infrastructure" provided a way for cyclists and motorist to be separated?

Trouble is, during those shorter lives, obese people tend to be heavier users of the medical system, due to comorbidities.

If you have the citations, I would love to see the contrarian studies. I can kinda see the point that they are trying to make, that shorter lives = less resource consumption. I'm just not sure that applies to the medical industry, which has its own topsy-turvy economics.

I believe it was in an article in 2008. I didn't actually read the study. It was a Dutch study. The higher costs were incurred during the end years of the thin, healthy people that lived longer. The cost over the number of years of longer life is what drove their total up. Folks in their elder years end up having some of the same health problems that smokers and overweight people get at a younger age, so the costs for those instances end up being equal.

Are you kidding? The injury and fatality rate per mile is much much higher for cyclists than for motorists. More injuries and more fatalities equate to higher healthcare costs.

You're mistaken. There are numerous studies showing commuter cyclists live longer, healthier lives than motorists and as far as collisions go, there are certain areas where cycling can be more dangerous than others, but overall, there is little evidence this is the case.

In my own province, our mandatory provincial insurer post each years traffic collision statistics and you can compare the death and injury rates between motorists and cyclists. There's little difference. Pedestrians fair far worse than cyclists. They also show that proportionately, cyclists are involved in half the collisions that motor vehicles are.

A couple of things are important to remember. First, I suspect what worries us most are collisions with motor vehicles. Thankfully, these collisions make up only a small portion of collisions while cycling. Second, it's a person who drives a car or rides a bike and it's their control (or lack of control) that determines crashes. If one is prone to crashing on a bike, it's a good bet they're prone to crash in a car.

You're mistaken. There are numerous studies showing commuter cyclists live longer, healthier lives than motorists and as far as collisions go, there are certain areas where cycling can be more dangerous than others, but overall, there is little evidence this is the case.

In my own province, our mandatory provincial insurer post each years traffic collision statistics and you can compare the death and injury rates between motorists and cyclists. There's little difference. Pedestrians fair far worse than cyclists. They also show that proportionately, cyclists are involved in half the collisions that motor vehicles are.

A couple of things are important to remember. First, I suspect what worries us most are collisions with motor vehicles. Thankfully, these collisions make up only a small portion of collisions while cycling. Second, it's a person who drives a car or rides a bike and it's their control (or lack of control) that determines crashes. If one is prone to crashing on a bike, it's a good bet they're prone to crash in a car.

Hang on... he used the phrase "per mile." When one compares miles traveled per vehicle and injuries per mile... bike travel doesn't look all that great.

Scan down on this web page http://www.kenkifer.com/bikepages/health/risks.htm to "Calculations Based on Miles Instead of Hours," and you'll see the same conclusions.

Hang on... he used the phrase "per mile." When one compares miles traveled per vehicle and injuries per mile... bike travel doesn't look all that great.

Scan down on this web page http://www.kenkifer.com/bikepages/health/risks.htm to "Calculations Based on Miles Instead of Hours," and you'll see the same conclusions.

I was thinking of time of exposure, but mostly think the problem is one of behavior over all other factors

OK, but why are these injury and fatality rates higher? Could it be due to cycling among autos? What if "improved infrastructure" provided a way for cyclists and motorist to be separated?

Great point. If we could use the new infrastructure to increase safety (of not only existing riders, but also new ones), that would seem to provide yet another reduction in healthcare costs.


Hang on... he used the phrase "per mile." When one compares miles traveled per vehicle and injuries per mile... bike travel doesn't look all that great.

Scan down on this web page http://www.kenkifer.com/bikepages/health/risks.htm to "Calculations Based on Miles Instead of Hours," and you'll see the same conclusions.

Great article/website, thanks for sharing.

There are numerous studies showing commuter cyclists live longer, healthier lives than motorists...
Can you provide references to any studies that actually measured and compared the health and/or life span of commuting cyclists with similar populations that don't cycle commute?

One example comes from Mayer Hillmans work.

He has explained the concept of "life years" which he had used. This allowed him to make a comparison of the life years lost in road accidents against the life years gained through improved fitness. He took each cycle death in 1989 and looked at the acturial evidence as to how long a person of a certain age could expect to live and multiplied that by the number of deaths in each age group in order to arrive at a figure of approximately 11,000 life years lost in cycle fatalities. To calculate life years gained he looked at some American evidence of the increased longevity of those who cycled regularly and this showed that people could expect to live two years longer. When this is multiplied by the numbers of those who cycle regularly, as revealed by the National Travel Survey, you arrive at the extraordinary ratio of 20:1. In other words, he says, for every life year lost through accidents, 20 years are gained through improved health and fitness.

Another example was a study done at Copenhagen University Hospital, Copenhagen, Denmark.

The title of the study was "All-Cause Mortality Associated With Physical Activity During Leisure Time, Work, Sports, and Cycling to Work" and can be found at
http://archinte.ama-assn.org/cgi/content/abstract/160/11/1621

The study concluded, concluded:

"Even after adjustment for other risk factors, including leisure time physical activity, those who did not cycle to work experienced a 39% higher mortality rate than those who did."

A policy report on the health benefits of increasing levels of cycling in Oxfordshire - http://www.modalshift.org/reports/tandh/print_version.htm - showed one rough calculation suggesting that new cyclists covering short distances can reduce their risk of death (mainly due to the reduction of heart disease) by as much as 22 per cent.

But really, doesn't common sense tell someone regular exercise makes you healthier and sitting on your butt isn't going to improve your health?

As much as we fear trauma, isn't it common knowledge that far more people die from poor health than from trauma?