Journeys to work by bicycle (source: Charting Transport)

The newly released Census 2011 data on the journey to work shows cycling’s mode share increased in all major cities bar one. Thanks to Charting Transport once again crunching the numbers, we can compare the figures across Australia for bicycle-only trips.

The journey to work is one of the most promising purposes for cycling as a means of transport i.e. for non-recreational cycling. That’s partly because other modes are less competitive in peak hour due to congestion and over-crowding. Also, the regularity of commuting means problems like guarding against theft and showering/changing are easier to manage.

However as the first exhibit shows, there’s no avoiding the obvious. When looked at from the point of view of the total metropolitan transport commuting task, cycling is apparently still a bit player.

Less than 1% of workers in Sydney currently travel by bicycle. Even in Melbourne with its stronger tradition of cycling, it’s the mode of choice of just 1.5% of commuters. Cycling does best in Canberra, but it’s a very small place with an atypical workforce profile.

Still, cycling increased its mode share over 2006-2011 in all mainland capitals bar Adelaide. But the improvement was slight – its share went up by 0.2 percentage points in Sydney, from 0.7% of all commutes to 0.9%.

It also increased by 0.2 percentage points in Melbourne, Perth and Canberra, and by 0.1 percentage points in Brisbane. It dropped 0.2 percentage points in Adelaide.

This seems like a pretty miserable return given the increase in cycling infrastructure investment since 2006. But there are a number of other factors to consider.

One is that cycling does much better in inner city areas (see second exhibit). In the Melbourne suburb of North Fitzroy, for example, getting on for 15% of commuters cycle to work.

As with Canberra, that probably reflects special characteristics of this suburb. It has reasonably good cycling infrastructure (e.g. Canning Street), is close to the CBD, and has plenty of young and well educated residents.

Another factor to take into account is that the numbers cycling to work grew very strongly in absolute terms over the period. In Melbourne, cycling grew from 20,598 to 28,606, or by 38%. In Sydney the percentage increase was even bigger, with the numbers growing from 12,128 to 17,838, i.e. by 47%.

Averaged over all mainland capital cities, the numbers cycling to work increased 36%, or an average of 7% p.a. Although cycling’s mode share fell 0.2 percentage points in Adelaide, there was nevertheless a small rise in the number cycling.

That these large absolute increases resulted in only a marginal change in mode share is an object lesson in the dangers of only paying attention to growth rates. Regard must be given to the size of the base (very small in this case) and to the fact that competing modes are also growing in absolute terms.

Cycling’s low mode share might also be partly a function of the Census being taken on a single day every five years. The 2011 Census only captures how workers travelled on Tuesday 9 August.

That doesn’t account for seasonal variations. In cities like Melbourne with a Mediterranean climate (i.e. cold & wet winters, hot & dry summers) there are likely to be many fewer cyclists in winter when the Census is conducted than in summer.

Weather also varies from day to day. For example, Census day 2006 in Melbourne was relatively balmy, but in 2011 it was forecast to be cold and wet (update: it was cold and wet).

Most of these objections can be minimised by looking at all cities and by focussing on the inter-Census change. I think it has to be acknowledged that cycling hasn’t made strong headway so far.

That suggests a number of issues worth thinking about.

It may be that cycling has already attracted most of the workers who’re prepared or able to cycle to work. Any really large gains from here may require far-reaching but politically difficult changes.

For example, it might be necessary to construct a comprehensive network of fully segregated paths before it really takes off. Lower speed limits and changes to the law to positively support cycling might also be a pre-condition.

I expect cycling in Australian cities will continue to be a choice largely made by city centre workers for quite a few years yet. As at present, most of them will live in the inner city – see second exhibit.

We need to think further about the ‘bar’ for cycling. Copenhagen and Amsterdam are inspiring examples, but it should be borne in mind that they’re outliers, even in the context of Europe. Cycling’s commuting mode share doesn’t need to hit double figures for it to be a key mode of transport.

Since someone will surely raise it if I don’t, I have to say I don’t think the mandatory helmet law has much to do with cycling’s share of commuting.

I doubt there are many would-be bicycle commuters who are so averse to helmets they choose not to cycle. Since commuting mostly involves mixing with traffic, I expect virtually all positively want to wear a helmet.

There’s more data on the journey to work at Charting Transport. There’re charts here and some delightful maps here. Also see my previous post on what Census 2011 tells us about commuting by public transport.

Animation: % journey to work by bicycle 2006 & 2011, Melbourne (source: Charting Transport)