Left out and invisible? : exploring social media representation of ‘left behind places’

‘Left behind places’ and regional inequalities have garnered increasing attention from policy makers, with new data needed to support further analysis. Social media data has the potential to act as a real-time barometer of local opinions, supplementing traditional time-intensive public opinion surveys to inform public policy insights. This study evaluates the scope of potential use of social media data to investigate left behind places. Twitter is used as a case study, with the volume of tweets measured across England and Wales. Linear regression is employed to identify under-represented places in the data. The residuals of the model are then compared across classifications of left behind places, to explore whether left behind places are more likely to post fewer tweets, rendering them ‘invisible’ in the data. This study provides a valuable foundational assessment of the potential suitability of Twitter data for this purpose, engaging with concepts relating to spatial bias and social inequalities.