In the last few weeks, there have been some nasty cycling accidents in Sydney. I’ve been a cyclist commuter in Sydney off and on since 1996, and my first reaction (particularly as one of the cyclists killed was a neighbour of mine) is to think that I should stop cycling, because it is just so risky these days.
But is it? I decided to see what I could find out, and in particular, whether the number of kilometres of cycling had increased by more than the number of deaths. And I thought I would see what I could find out about injuries, too. The first place I looked was a report entitled Cycling in NSW – What the Data tells us. According to this report:
There is no single central source, repository or clearing-house for data about cycling in NSW. Data is collected by a variety of NSW Government agencies, local councils and cycling organisations. Available bicycle usage data is recorded at varying levels of detail, reliability and quality. Datasets on cycling infrastructure are updated intermittently and standard terminology is often applied inconsistently. Richer datasets are available on commuter cycling than recreational cycling.
My favourite comment:
Injury data may be both under-reported and double-counted.
For anyone interested in cycling statistics this is a fantastic report (despite the initial disclaimers about the data), and well worth reading in its entirety. But for my purposes, since it was written in 2008, it doesn’t really tell me what I want to know, which is what is happening in the trend. To understand this, we need to know two things. First, we need to have some view of the number of serious accidents and deaths involving cyclists, with a trend. Second, we need to have some view of how much cycling is happening. As the report notes above, this data isn’t great.
- Looking at the census measure of proportion of trips to work that were cycling trips (in 2001, 2006 and 2011) and adjusting the population for that change – this proportion grew by 3% more than the population growth in NSW over the same time period, a total of 34% over 10 years (but this is only a commuter measure).
- Looking at the number of trips over the Harbour bridge and Anzac bridge per day and how it changed between 2007 and 2013 – the number of trips (averaging the two bridges) increased by 10% per annum over that time period, a total of 80% over six years – this measures weekday and weekend use, so might be a better measure of exposure.
- Looking at the information from the NSW Household Travel Survey of the proportion of trips that were made by bicycle (this proportion was so small it only could be seen in the detailed tables, and reduced from 0.64% of all trips to 0.55% of all trips between 2001 and 2011). In other words, this survey shows the opposite to the other two.
So for two of my measures, the proportion of the population cycling went up since 2001, and for one, the proportion went down. On balance, I do believe that there is more cycling happening now than there was when I first started riding a bike on Sydney’s streets nearly 20 years ago. The census has the biggest coverage (which matters, with the small numbers we are talking about), and is backed up by real numbers on the streets.
So am I safer now than I was then?
In 1996 there were 218 reported cycling injuries and 2.1 deaths per million population in NSW. In 2012, my three different possible measures give between 81 and 149 injuries per million population (after adjusting for the amount of cycling happening). I don’t have injury figures for 2013, but for deaths (which went up between 2012 and 2013) there were between 1.0 and 2.0 deaths per million population (with various adjustments for exposure to cycling deaths). The injury statistics have more chance of changes in reporting in that time, but (fortunately) the numbers of deaths are small enough that any individual year is quite volatile.
Even if I believe the NSW Transport study, I still seem to be slightly less likely to die on the streets, and quite a lot less likely to get injured.
Judging by the number of deaths, cycling does seem to have got a bit riskier in NSW in 2013 and 2014 than it was in the previous couple of years. But overall, it is still safer than it was when I started.
I will mourn my neighbour, who used to wave cheerily at our family on our Sunday morning bike rides, while he worked on his bike. But I’m going to keep riding Sydney’s streets.
NSW Household Travel Survey 2011/12
Australian Bureau of Statistics (Census data about mode of travel to work on census night)