A quick post on a common urban theme. Answering the question of how many people live in a city is very tricky. Do you use the political boundaries of the local authority? Do you use the extent of the urban area? Do you use some other definition? In a completely unscientific experiment, I've developed a different approach - and it turns out that the populations I come up with are not that different to those given for the city-regions of the places on my list.
Here's what I did... I took the eight core cities in England, plus London and chose the main central train stations in each of them as the central location (e.g. Lime Street in Liverpool). Then I used a bit of GIS wizardry to calculate the population within a 15 mile radius of said stations (statistically significant distance determined after much spatial analysis... and a bit off the top of my head). Here's the results (using ancient 2001 Census data):
Obviously, these figures differ a lot from the local political entities with the same names (e.g. Manchester the district has about 430,000 people). Even though I've used a fairly arbitrary technique, the populations are a much better reflection of the urban population than others often cite. So, let's compare the figures. On Census day in 2001 London had 7,172,091 people, Birmingham city-region had 2,693, 917, Manchester city-region had 2,482,328 and the other cities on the list show similarities to the 15 mile figure. What does this tell us? Not much really, except that if we use a 15 mile buffer from central train stations we get a good approximation of city-region populations for the major English cities. Here's how it looked on the map: