Downtown Houston via Google Maps - if you don't believe it, click to look around

The Sydney Morning Herald reported this week that a “far smaller” proportion of Australians is cycling today than was the case in the mid 1980s. The number of cyclists dropped on a per capita basis “by 37.5% between 1986 and 2011.”

That must be unwelcome news for cycling advocates but music to the ears of those State Premiers who’re cutting back on funding and support for cycling. Victoria’s Ted Baillieu, who provided no new funding for cycling in his 2012-13 budget, could now defend his niggardliness by arguing that interest in cycling is waning.

The numbers come from an article by Chris Rissel and Chris Gillham published in the latest issue of World Transport Policy and Practice. While it’s more of an activist’s publication than a serious academic journal – e.g. it’s not peer reviewed – this is an interesting article.

They looked at the number of cycling trips per day in Australia in 1986 and 2011 by persons aged over nine years and compared them against the relevant population for those two years. They used Census data for their population statistics but had to rely on surveys for the cycling data.

On the face of it their numbers are sobering, but I think there are good reasons not to get overly stressed. Nor should they give Mr O’Farrell and Mr Baillieu much comfort.

First, I don’t think the data Rissel and Gillham rely on is good enough to draw definitive conclusions. They only have two data points – separate surveys done in 1986 and in 2011. Moreover, while the surveys have large samples, they use different methodologies.

The 2011 survey was done by telephone and seems OK. The 1986 survey, however, was a self-completed, mail-back questionnaire. It’s likely to have had a low response rate with a bias toward more committed respondents. It may have over-stated the level of cycling.

On top of that, Rissel and Gillham appear to get their arithmetic wrong. On the raw numbers they use, the per capita daily trip rate was 0.132 in 1986 and 0.101 in 2011. That’s a drop of 24%, quite a bit less than their claim of 37.5%. They appear to have arrived at 37.5% by the conceptually arguable process of subtracting the percentage change in trips from the percentage change in population.

Second, the findings are sensitive to the time frame used. They adopt a very long (27 year) time frame. That’s apparently because that was the only survey available, but had they gone back to 1955 (say), they’d have probably gotten a much bigger drop. On the other hand, if they’d gone back ten years to 2002, I expect they’d have gotten a more positive outcome for cycling.

Third, they choose to compare cycling levels against population. That can be useful, but it has its limitations. It doesn’t fully allow for the fact that, as I discussed recently, the rate of growth of travel in cities has been falling in Australia and many other developed countries for circa ten years for a range of reasons.

For example, total kilometres of passenger travel by all modes in Australia’s capital cities increased by just 2% between 2003-04 and 2008-09. Population however grew by 9% over the same period. That’s a key reason transport analysts focus on mode share.

It’s likely cycling’s performance would look much better if it were assessed in terms of its share of all travel (trips or kms), especially over the last 8 to 10 years, rather than compared with population. Unfortunately, there’s little data available to show this.

However, the Census collects data every five years on the use of bicycles for work travel. The journey-to-work question is the most reliable and consistent measure of cycling we have. While commuting accounts for a relatively small proportion of all trips, it’s one of the most important purposes.

The good news is there’s been no decline in cycling’s share of the journey-to-work – its remained steady at around 1% between 1976 and 2006. It’s share is of course much higher than this in the inner city, especially around the CBDs of our capitals, but lower in the suburbs.

The fourth and final reason why I’m not feeling too miserable in the face of Rissel and Gillham’s numbers is that cycling is still increasing in absolute terms. Even on the numbers provided by the authors, persons aged nine years and over made 1.65 million daily trips by bicycle in 1986 and 1.99 million in 2011. Cycling rates have very likely increased faster in recent years but unfortunately they don’t cite any more recent figures.

It’s probable that virtually all of the increase has come from adults rather than teenagers and is concentrated in key locations like the inner city. So it’s not surprising it seems like there’re more cyclists on the road – there are! The argument for better infrastructure and more supportive laws remains compelling and pressing.

1986 is ancient history. I’d like to see some numbers on the trend over the last ten years. It’s more relevant and I think it would show a more encouraging and optimistic outlook for cycling.