Get data on active travel levels associated with new developments

[Robinlovelace]

Potential sources of such data:

• Census data (assuming we can get a list of new developments that were completed before 2010)
• Healthy New Towns initiative: https://www.england.nhs.uk/ourwork/innovation/healthy-new-towns/
• NTS (assuming we can get data on which households were in new developments)

[Robinlovelace]

This could contribute to an indicator in terms of % trips by active modes.

[mvl22]

I think we should be wary of lumping together walking and cycling percentages as "active modes", because this will mean that in-town developments like those near stations and other transport hubs will end up with a very positive indicator, simply because they will naturally have high levels of walking, even though the cycling provision may be non-existent or of very low quality.

Cycling is almost always harder to achieve than walking, because pavement provision is almost universal, whereas cycling is not built into the psyche of developers yet, and as such involves extra thought.

I am not saying that walking doesn't have health/other benefits, but I suspect a lot of such walking trips will just be walking to the nearest station, which is of limited benefit.

One approach (mentioned by Robin) is weighting these differently.

This feeds into my general dislike of lumping together walking and cycling as if they were one thing. They are different transport modes and should be treated in policy terms as such. We need to get the UK away from the thinking of cyclists as walkers on wheels, travelling a bit faster. Cycling should be something that can be done at speed (as well as pootling along like me on a dutch bike), if it is to be a serious mode of transport, that can compete with the car.

Cycleways designed to join up settlements that consist of poor-quality stop-start shared use for 5 miles will never be used; by contrast, segregated cycleways that enable people to build up speed and provide good visibility and space become a real alternative to the car for journeys up to 5 miles.

[Robinlovelace]

Good point @mvl22, any further comments on this welcome. One way to create an evidence-based weighting would be to weight not arbitrarily differently between walking and cycling but to use METs (metabolic equivalent of task) for different modes (and potentially different simulated trips on the network to trip attractors, which would also account for hilliness and distance of trips) to weight the relatively crude counts or mode share of estimated trips.

[joeytalbot]

Yes, I agree cycling is harder to cater for than walking, and it would be helpful for us to be able to differentiate between cycling and walking, perhaps through using METs.

In terms of location though, I think there's a real need to create mixed-use communities where homes are close to schools, work and shops, and where the need to travel is minimised. It's only in these kind of urban settings that there's a real chance for people to live there without needing to own a car. So I wouldn't be in favour of a metric that discriminates against locations where most journeys are short in distance.

[Robinlovelace]

Data on METs from different modes:

• Walking MET = 3.20
• Running MET = 8.80
• Biking MET = 5.90

Source (one of many): https://github.com/EsriCanada-CE/ecce-app-challenge-2017/tree/master/Team_Earth_Link

[joeytalbot]

https://sites.google.com/site/compendiumofphysicalactivities/Activity-Categories/walking

The wording of the categories in here is very revealing (it is American). The most appropriate for us seem to be:

• Walking from house to car or bus, from car or bus to go places, from car or bus to and from the worksite = 2.5

• Bicycling, to/from work, self selected pace = 6.8

[joeytalbot]

I've got code now for importing the OA nomisweb data on method of travel to work, using the nomisr package.

[Robinlovelace]

Fantastic work @joeytalbot