A few years ago I was cycling from Reading to London and struck me how busy the roads around the reservoirs near Heathrow were. In fact, I seemed to remember that they weren’t that busy when I first cycled there a few years earlier. This then got me thinking about how busy roads used to be in the late 80s and early 90s. I was beginning to wonder if I was having one of my first senior moments “where things used to be better when I was young”. Occasionally, I would remember this question which was difficult to answer due to the Small Data problem of finding out how many cars, or better vehicles, there used to be. In May 2013 I stumbled across UK government data whilst investigating UK housing stock. And to my surprise there was government data for the registered vehicles in the UK going back all the way to WWI (in the following “vehicle” refers to a registered vehicle).
I quickly plotted the data (top left plot) and it showed the familiar curve with car ownership taking up in the 50s and slowdowns for recessions and also the early noughties. But, on reflection I thought this doesn’t really answer my question. When there are more cars surely there are also more roads. Fortunately, the UK government also publishes data for the historic annual length of the road network in the UK. That’s shown in the top right figure. So, that’s growing as well, apart from the late 60s when it shrunk. And there is also a mysterious large growth a bit later on. But give or take this is probably fairly accurate. Is everything fine? Well, did the road network grow as quickly as the of vehicles? In trying to answer that we can work out the (naive) vehicle density, the number of vehicles per unit of road. That’s in the bottom left plot, showing the number of vehicles per kilometre. And that is going up steadily (apart from the short period where the UK road network grows mysteriously fast)! I could have worked out the number of vehicles per mile to use more familiar units for some, but given that the typical length of a car is close to five meters, the plot in the bottom right shows the number of vehicles per five meters, which is measures the vehicle density in a more natural unit, a car length. It is a shame the statistics for the years before 1950 do not match year for year, and perhaps more detailed analyses of vehicles and vehicle types per road type have interesting stories to tell.
So, the answer to my question is that roads have become more crowded (on average, if people actually use their vehicles). Whether that explains al aspects of my impression of today’s road traffic is another matter.
There are a few things to keep in mind before over interpreting these simple numbers.
- This is a crude average, there could be large local variations.
- Some roads have more than one lane, in fact (probably) most have one for each direction, so the density numbers need to be divided by two at least (probably).
- Is there a maximum car density? How do other countries compare? This could be an interesting measure for the wealth of an economy.
- A lot of cars are stored on driveways and in car parks which should bring down the number of cars on the road (until they all commute to work and do the school run).
- Can you fit all UK vehicles onto the UK motorways at the same time?
- … and there is probably more.
Thanks go to Gareth for listening to and criticising me in a Google hangout whilst “discovering” this.