People Killed or Seriously Injured (KSI) on UK Roads, 2004-2021

For various reasons I became interested in how many people had not died on UK roads over the period of Coronavirus "lockdowns", and so I pulled some numbers from the relevant UK Department for Transport website and reduced them using the python code shown below.

Data Selection

The data chosen is the Killed or seriously injured (KSI) adjusted data set, per the description:

For analysis of trends over time across the country, using the experimental adjusted series is recommended. They are based on what we estimate the totals would be if all police forces were using injury-based severity reporting systems.

Data Extract Process

Visit: https://roadtraffic.dft.gov.uk/custom-downloads/road-accidents
Select: Casualties
Select: Killed or seriously injured (KSI) adjusted
Select: All Years
Select: Great Britain, countries and regions
Select: Great Britain
Select: Casualty age
Select: Confirm
Result: Your report reference is 14fa598f-ea9c-4e5d-8db4-d2a6d32ac70c
Link: Data

Chart

Overall Results

across all age groups, 2020 saw 6810 fewer KSI than 2019
across children ages 0 thru 17, 2020 saw 2666 fewer KSI than 2019

Per the ITF-OECD report:

Similarly, large fatality reductions were seen during the recession in the early 1990s. However, traffic levels have risen again in recent years, surpassing pre-recession levels. This may be why casualty levels have plateaued at best, except in 2020.

Some of this is visible in the chart above, that since 2010 mortality statistics have been almost flat, only gradually declining until the lockdowns of 2020 occured; surprisingly this is in spite of rising numbers of cars on the road (ITF numbers)

year 2000: 29.629 million cars
year 2010: 35.170 million cars
year 2019: 39.891 million cars

...so it's quite impressive that numbers continue to decline in the face of expanding car usage. Reductions likely include the results of crackdowns on drink-driving and on speed, but also economic effects such as young people being "priced out" of insurance, or even simply moving to the cities and using more and greater variety of public transport, e.g. Uber. It would be hard to generalise.

Also from the ITF come interesting statistics re: the numbers of deaths per 100,000 population:

year 2000: 6.1 road deaths per 100k
year 2010: 3.0 road deaths per 100k
year 2019: 2.7 road deaths per 100k
year 2020: 2.3 road deaths per 100k

...which are fairly comparable to national rates for homicide, averaged for all genders:

year ending March 2000: 1.29 homicide deaths per 100k
year ending March 2010: 1.07 homicide deaths per 100k
year ending March 2019: 1.08 homicide deaths per 100k
year ending March 2021: 0.99 homicide deaths per 100k

For Discussion

After a considerable period of "plateau" in improvements for child protection on the roads, if it's possible to save considerably more than 2000 children (and nearly 7000 people overall) from being killed or seriously injured each year, why not re-impose national lockdowns to achieve that goal?

alecmuffett/uk-roads-ksi-2004-thru-2021