Number of adult cyclists counted at the same 25 intersections in Sydney between 1990 and 1993. Mandatory helmet law for adults was introduced between the 1990 and 1991 counts (data from Smith & Milthorpe)

As a consequence of the lacklustre performance of bikeshare schemes in Melbourne and Brisbane, there’s been a heated debate over the last couple of years about the wisdom of the country’s mandatory helmet law.

Much of the evidence offered against compulsory helmet wearing relies on a handful of ‘before and after’ studies that were done in Victoria and NSW at the time the law was introduced at the start of the 1990s. These “natural experiments” are cited so many times that I thought it would be an instructive exercise to go back and look at the original sources myself.

I started with a study done in Melbourne by the Monash Accident Research Unit in 1993. You can read my analysis here. In a nutshell, rather than showing the new law devastated cycling across-the-board, I was surprised to discover the level of cycling by adults and primary school children wasn’t seriously affected.

However there was a big fall in the high school-age youth segment (12-17 years). Youth numbers were flat in the years leading up to the law, but declined sharply in the first year after and remained at that lower level in the second year.

Still, that’s only one study. There’s also a group of four related NSW studies that are frequently cited in this debate. They proved so extraordinarily hard to find that I doubt many of the people who reference them have actually read the originals. However thanks to the kindness of a reader I was able to access all four and offer the following thoughts.

There were three surveys of cyclists done in 1990, 1991 and 1992 by Michael Walker from the Department of Psychology, University of Sydney, for the NSW Roads and Traffic Authority. He counted and described cyclists in urban and country NSW using essentially the same observation points and methodology in each year.

The first was done in September 1990 prior to introduction of the law and the second in March and April 1991 after helmets became mandatory for adults (but not riders aged under 16 years – they had a further six months grace). The third survey was undertaken in March and April 1992 when helmets were mandatory for all riders.

A fourth study was done for the Roads and Traffic Authority in March and April 1993 by Nariida Smith and Frank Milthorpe from the Institute of Transport Studies, University of Sydney. It also used essentially the same locations and methodology as the preceding studies.

As the exhibit shows, Smith and Milthorpe found the number of adult cyclists observed on roads in Sydney was not devastated by the law. Numbers were only 5% lower in 1993 than prior to introduction of the law in 1990. Indeed, numbers actually increased 22% in the first year after the law, probably because of better weather.

The number of adults cycling for recreational purposes was only counted in 1992 and 1993 (i.e. post-law in both cases) but showed a 48% increase in Sydney and a 148% increase in regional areas. Figures for recreational cycling by children were available from 1991 (prior to the law applying to children) and show only a small and insignificant decline over the period to 1993 (5%).

Yet it wasn’t by any means all good news. As was the case in Melbourne, there was also a dramatic fall in the number of students cycling to school. Between 1991 and 1993, numbers fell by a massive 64%. Unlike Melbourne where the fall was mainly limited to high school students, the drop in primary school numbers was almost as large.

Cycling levels also dropped significantly at observation sites in regional NSW between 1990 and 1993 (the Melbourne study didn’t look at regional areas). They fell on average by 37%, most dramatically in regional centres like Tamworth and Lismore, where the drop approached 60%.

So, the introduction of the mandatory helmet law in NSW was associated with a sharp and significant decline in cycling by students and regional adults. However, there appears to be no correlation between the law and the level of on-road cycling by adults in Sydney, or the level of recreational cycling by children. Moreover, recreational cycling by adults rose sharply in the years following the new law.

While these findings might give comfort to both sides of the debate, it’s important to consider the limitations of all four studies. Indeed, Smith and Milthorpe emphasise upfront that the findings of their report “should not be used to estimate total exposure or ridership”.

The trouble is the number of observation points is very small. For example, there was only one location at which adult cyclists were counted in each regional centre. Counts of on-road adult cyclists in Sydney were based on 25 locations, with just one in the CBD and two in the entire eastern suburbs. Smith and Milthorpe say the sites didn’t capture known cycling routes.

This constraint is highlighted by the high degree of variability between locations in terms of numbers of cyclists and other measures like helmet use and compliance with road laws. For example, counts of adult cyclists in Sydney rose by 33% at the two Ashfield observation sites over 1990-93 but fell by 50% at the two Parramatta sites.

Care should therefore be taken in extrapolating from these findings to make general points about cycling in NSW between 1990 and 1993. Come to think of it, similar caution should be exercised in relation to the Melbourne study too.

Update 30 September 2012: The NSW Roads and Maritime Services agency has made these studies, and others, available publicly.