Paper on estimating casualty rates

[joeytalbot]

This could be submitted to Active Travel Studies

Questions could include:

• how does using rates vs counts of casualties change the picture?
• how can commuter data be used as a proxy for active travel levels?
• what % active travel is composed of commuting for walking and cycling?
• how can we define the 'rush hour' in an evidence based way using household travel survey data (the NTS)?
• what levels of geographic resolution how many crashes are needed to get meaningful estimates of road safety across zones (i.e. what is the smallest zone size that can be used for this kind of analysis)?
• which boroughs have the highest/lowest casualty rates and why?
• which boroughs have seen the greatest increase in walking and cycling safety and what are their characteristics/why?

[Robinlovelace]

Plus looking at link with IMD.

[Robinlovelace]

Could be split into sub-papers.

[Robinlovelace]

Synergies with #28

[Robinlovelace]

I think a starting point also could be facetted plots showing how crash numbers change over time is worthwhile.