Econ Insights: Mapping nightlife in the UK with Yelp data
Updated: May 2
In this series, we look at interesting data sets to uncover hidden insights.
As part of a research project into the health of town centres and high streets I was interested in understanding the role of night life. The first challenge was to identify an appropriate data set with in order to understand the density and distribution of of nightlife in our towns and cities. There wasn't any official data so I decided to try using data from Google Maps, unfortunately google only makes basic information available for on its API. Instead I turned to Yelp which,, although less complete than google, has good coverage of most areas.
Extracting the data took a bit of trial and error as the API only returns 50 results at a time but after a few hours of playing around with python I built a program which looped through a list of towns with a population greater than 20,000 and returned all of the bars and clubs open at 1am on a Saturday.
That data set contained the co-ordinates of each listing and so I was able to map them in QGIS. Each point represents a listing and it is overlayed with a heatmap to highlight clusters.
As could be expected London has the highest number bars and clubs with over 300. The densest cluster is in the West End around Soho and Leicester Square. But there are hot spots throughout the city reflecting the distributed nature of London's night-time and social economy. In the east a sequence of spots runs up the A10 from Shoreditch towards Dalston and Stoke Newington, in the south there are clusters in Brixton and Clapham. The relationship between where people live and where they socialise is potentially an interesting area for further research.
In contrast, in Manchester which has the 96 locations nearly all the venues are concentrated in the city centre. is much more consolidated. This potentially has an impact on the strength of the local economies of the towns which make up Greater Manchester as the night time sees flows of people to the city centre. There are several reasons why this could be the case, the tram system in Manchester makes it relatively easy to access the centre but orbital travel remains difficult.
The top 10 areas in terms of the number of bars and clubs are:
Brighton and Hove 23
A slightly more interesting picture emerges when looking at the density of nightlife by population as shown in the chart below. In this view London has a relatively low density given the size of its population.
The chart uses the administrative population, which for Manchester is about 500,000, however, a better measure would be to determine a catchment population, i.e. the population who would be likely to travel to the city centre. Given the transport connections with the rest of Greater Manchester (population 2.8m) this is almost certainly too low.
The most interesting observation is the poor performance of Birmingham, the population of Birmingham is around 1.1m, with over 2.9m in the Greater Birmingham area. Why Birmingham does not enjoy the same nightlife as Manchester is difficult to explain and there may be an element of determinism about it - Manchester has a reputation for being a good night out in a way that Birmingham doesn't. However, I think the lack of a reliable, efficient transportation system may something to do with it.
Take Dudley, which is about 10 miles from Birmingham City Centre, by public transport the journey takes around 40 to 50 minutes and requires taking a bus and catching a train. The timings and route are inconsistent. In contrast Oldham, just over 8 miles from the centre of Manchester is served by the tram service which runs every 10 minutes (6 minutes during the week) from the centre of the town. Admittedly, some other areas such as Walsall and Sutton Coldfield are better served by direct trains but many other outer areas of Birmingham rely on bus services.