Analysing the distribution of walking in the Swiss population

In many countries including Switzerland, public policy encourages people to walk for reasons linked to health, the environment, and transport. However, the distribution of walking in the population is not known. People who walk great distances have not been investigated, nor have people who do no walking in public spaces although they drive a motorised vehicle (and who, arguably, may be seen as not conforming to certain public policy objectives). Nothing is known about the proportions or their distribution in space of these groups. In order to answer these questions, this project uses the Swiss transport micro-survey (MRMT2010), a complex database whose 13 inter-related sub-files include information on transport behaviour on a randomly selected reference day for 62’686 individuals. Each person was interviewed by telephone, in a representative stratified sample covering the whole of Switzerland. Rather than investigating mode shares, this study concentrates on the people involved in the survey. Preliminary analyses allowed the selection of walking bouts in one data file, which were then aggregated and linked to the characteristics of the people which were in another file. Detailed investigations on the distribution of walking in the population were carried out. The results show that walking in Switzerland is not normally distributed. The curve representing kilometres walked on the reference day is strongly skewed towards the left because a substantial proportion of the population walked very little or not at all. This finding is illustrated using histograms, and its implications are discussed. We then sub-divided the population into several groups, with different levels of walking and other transport behaviours. Altogether, 12% of the sample stayed at home on the reference day. A further surprise was that 23% of the sample succeeded in driving a mechanised vehicle, without any walking in public space (transport within buildings or facilities is not covered in MRMT2010, nor are any bouts < 25 metres). Other groups of interest identified were small walkers (who walked less than 2 km on the reference day) representing 27% of the sample), average walkers (2-5 km, 22%) and big walkers (>5km, 13%), as well as non-walking cyclists (4%) and outliers (>20 km of walking, 0.6% of the total sample). The implications of such a wildly unequal distribution of walking in Switzerland are discussed, and preliminary maps are shown, suggesting that people with widely different transport behaviours on a given day may well live next to each other. A suggestion is made to start tailoring public policy information in order to target the aforementioned groups. This has been done with success in sectors such as tobacco control, so in our view there is potential for such an approach if it is suitably adapted and applied to the promotion of walking.