The findings from a new study on the costs and benefits of cycling have gotten plenty of attention in mainstream and social media because they tell a story that intuitively seems right and which many of us want to believe. The exhibit shows that both the private and social costs of driving in Copenhagen are much higher than they are for cycling.
The calculations are from a paper by Stefan Gossling and Andy S Choi published in the journal Ecological Economics, Transport transitions in Copenhagen: Comparing the cost of cars and bicycles. It’s a seductive story but as usual it’s not as simple as that.
Social costs and benefits
The authors conclude that whereas each kilometre travelled by car in Copenhagen costs society money, each kilometre travelled by bicycle saves money.
Considering only social costs, each bicycle km is a gain to society (Euro 0.16), while each car km represents a cost (Euro 0.15).
That’s a saving of €0.31 per kilometre. The explanation, the researchers say, is that the exercise benefit from cycling significantly reduces the costs associated with ill health e.g. medical treatment, sick leave. These savings are more than three times greater than the health costs from bicycle crashes.
Although it doesn’t undermine the ‘direction’ of the key finding, the difference in the social cost of driving and cycling isn’t as large as the authors claim. They haven’t included the hefty taxes Danes pay for the privilege of driving (€-0.159/km) in the social costs column, even though they’ve put them in the total column. That revenue goes into the public purse and should be accounted for in the social costs.
When taxes are accounted for, the social cost of the two modes is €-0.01/km for driving and €-0.16/km for cycling. Not as dramatic as the author’s claim but it’s still a big difference (€0.15/km) in cycling’s favour.
The study provides some interesting insights about social costs.
- After taking account of taxation, Danes who drive pay the social cost of their mode choice.
- The social costs of pollution, emissions and road deterioration associated with driving are relatively small compared to the cost of congestion (the noise number is probably wrong – see footnote 1).
- The net health benefits associated with cycling dwarf all other social costs and benefits.
- The principal social health saving is in lower rates of ill-health, not prolonged life.
Private costs and benefits
The findings of the study are more arguable when it comes to the authors’ estimation of private costs. Taken at face value, they’re consistent with the high mode share enjoyed by cycling in Copenhagen; it seems Danes choose to cycle because it’s about half the (economic) cost of driving i.e. €0.24/km for cycling vs €0.51/km for driving.
As the first two columns show, cycling’s advantage compared to driving is mostly due to the very large private longevity (€-0.36/km) and health (€-0.15/km) benefits it provides. They’re worth a whopping €0.51/km combined. At the 2,592 km per annum per rider assumed by the authors, that’s €1,314 annually in increased life span and avoided personal health problems.
I’m sceptical about the size of these benefits. Unfortunately, the authors don’t provide enough detail on how they handled this aspect, but it’s not uncommon for motivated active transport researchers to over-state private health benefits.
One routine error is to assume everyone gets the same (large) exercise benefit from cycling irrespective of their age or level of fitness. Another is to assume riders wouldn’t exercise in some other way if they didn’t cycle. If the claimed extended life span made for every health reform (e.g. smoking, obesity, cycling) were added, I expect it’d show we could all be immortal if only we’d mend our ways (reminds me of the claims of job multipliers for major projects like coal mines).
Whatever their value is, I doubt the extended life span and health benefits offer the key explanation for why so many Danes cycle. One reason is all populations tend to under-value uncertain benefits that come well into the future. Another reason is it’s a truism that Danes are utility cyclists; unlike Australians, they see bicycles as a just a practical means of transport, like a train or a bus, rather than as a fitness aid or an enthusiasm.
Yet they choose to cycle in large numbers even though, according to the researchers, the penalty in terms of time spent travelling is more than three times higher than driving. Indeed, the cost of the time penalty is almost double the savings they make by not operating a car.
Part of the explanation might be that the time differential for the two modes calculated by the authors is too big. They assume cyclists in Copenhagen average 15 kph and drivers 60 kph across all trips. I admit my knowledge of traffic conditions in contemporary Copenhagen is limited, but that seems a large difference. It also looks like the methodology doesn’t take account of parking costs; I expect they would have a significant impact on the attractiveness of driving.
Looking at where cycling does well in Australian cities – the inner suburbs – I think the key explanation is it’s time competitive with both cars and public transport for certain trip purposes i.e. work and education (see How important is cycling in Australian cities?). It’s main drawback is the cost imposed by the low sense of subjective safety on Australian roads (that’s much less of an issue in Copenhagen and wasn’t measured by the researchers).
I think we can be confident the social costs of cycling are significantly lower than those of driving in Copenhagen and elsewhere, but I’m not confident this study makes the case as convincingly as it should have.
The value used for noise by Gossling and Choi doesn’t look right. Elsewhere in the paper the authors put the cost at €0.007 per km, which is closer to the values in the Danish Ministry of Transport’s External Costs of Transport Review than the €0.048 in the authors’ table. Using the lower value reduces the social cost of car travel in Copenhagen by €0.04/km.