Skip down to main content

Understanding the geographical variation of Internet activity across Britain

Published on
10 Sep 2013

Britain has the largest Internet economy in the industrial world (measured as a percent of GDP; 8.3% in 2012), and the Internet is an increasingly important part of British society and economy. Social life is increasingly mediated and influenced by online interactions that take place through email or social media.

However, despite the inportance of the Internet both for British society and the future of the economy, there remains a major area about which little is known: the geographical variation of Internet use and online participation across the UK. This is true both for both policy-makers and scholars. Although the Office of National Statistics (ONS) produces regular reports on the British population and economy, it produces nothing about the Internet that is more detailed than reports on the 12 official regions. Even Ofcom produces little beyond broadband penetration reports. Similarly, scholarly work (Blank and Reisdorf 2012; Blank 2013) is mostly at the level of the UK as a whole.

Geographical data is crucially important because there is evidence of major geographic inequalities in access and use (e.g. Internet use in Scotland is 20 percentage points below the East Midlands). Government support for organisations like Go ON UK (and its predecessor Race Online 2012) signals that it understands the importance of mitigating digital inequalities in order to promote growth and employment. But policy currently can do little about local-scale geographic inequality because there are almost no data on the geography of the Internet. Currently no one knows how Internet use differs between Edinburgh, Manchester, London, or Cardiff. Outside of Ofcom reports about broadband penetration (an important, but limited topic), no government, private or scholarly entity has local-level geographic data on the Internet.

As part of an OII project looking at the geography of digital inequality, we will combine data from existing datasets to produce the first dataset with detailed estimates of Internet use. Specifically, we will combine the OxIS survey with the 2011 Census for England and Wales, the Scotland Census 2011, and several special-purpose datasets on important metrics of Internet use and participation (e.g. tweets, Wikipedia articles, photo uploads). This will give us a rich dataset with hundreds of measures of Internet use that we can analyze at any desired geographic level, including wards, counties (or Welsh and Scottish Councils, or Unitary Authorities), or cities.

More specifically, we intend to investigate the geography of several types of use. First is simple use or non-use of the Internet. However, we recognise that there are many ways of interacting with, using, and communicating through the Internet. We will also examine online buying and selling, social networking, banking and finance, information and entertainment seeking, politics, and communication. Multiple uses suggest greater intensity of use, which we can investigate by looking at the amount of time spend online. The Internet is a unique medium in that it allows ordinary users to create and distribute content; we will therefore explore online content production in the form of blogs, personal websites, uploading music or videos, and others.

Our research will begin with descriptive statistics and maps of the geographic distribution of uses. We will then move beyond descriptive work to multivariate, inferential studies that predict the geographic inequalities in digital Britain; using spatial statistical analyses we can examine the relative importance of issues like broadband use, technology attitudes, trust in e-commerce, or Internet experience as predictors of geographic stratification.


Blank, G. (2013) Who creates content? Stratification and content creation on the Internet. Information, Communication & Society 16 (4).

Blank, G., and Reisdorf, B.C. (2012) The participatory web: A user perspective on Web 2.0. Information, Communication & Society 15 (4).