> The combined effect on engagement with organisations such as political parties, unions and professional associations was a 6% reduction in participation from

> The combined effect on engagement with organisations such as political parties, unions and professional associations was a 6% reduction in participation from 2010 to 2017 for each 1.8km closer to the local exchange someone lived. This doesn't link to the study for some reason, so it's hard to tell, but it looks as if they are using proximity to telco exchanges as a 1:1 proxy metric for internet bandwidth, which doesn't make any sense for a whole pile of reasons, and then trying to connect that over a time span of 2010 to 2017 while assigning 100% of the change to this as the only possible variable.