This paper illustrates whether the availability of online content of local interest affects the likelihood of individuals to connect to the internet and spend more time online. While the literature demonstrates a number of factors which push or enable individuals to spend more time online, we know little about the conditions that pull or attract individuals online. Although we know that individuals use the internet to access information, we do not know whether such attraction forces are relevant at the local scale too. Gaining a better understanding of how such mechanisms work at the local scale can assist our efforts to bridge digital divides, which tend to be geographically clustered. To explore this we utilise innovative data, which contain all the archived webpages under the UK top level domain name (.uk) and we calculate the volume of internet content of local interest at the neighbourhood level using the geolocation information included in the text of these webpages. Specifically, we calculate the radius of gyration for every archived website using the different postcodes included in the archived webpages and then we create an aggregated measure at the neighbourhood level discounting websites that have less of a local focus. We merge this measure of Local Internet Content (LIC) with a large population survey, which contains information about the frequency of internet usage in the UK and estimate the effect of LIC on the likelihood of an individual being a frequent internet user. Multilevel models are employed to utilise both individual and geographical level characteristics. Our results indicate that even after controlling for the individual and geographical characteristics, which according to previous studies affect internet usage, the availability of internet content of local interest still attracts individuals online.
Read full abstract