Reform UK voters are the least likely to see posts from friends and family on social media and the most likely to be shown content from brands and news organisations, according to a new study.
The research, carried out by the Institute for Public Policy Research (IPPR), examined the feeds shown to social media users on Instagram, Facebook, X, Bluesky and TikTok. The thinktank said its findings suggest that algorithms are helping to fuel isolation and division by shaping the kind of material people are exposed to online.
Among the clearest differences identified in the study was the share of content coming from people users knew personally. Only 13% of Reform UK voters saw posts from someone they knew, the IPPR found. That compared with 23% among Green party voters.
The thinktank said this gap points to a broader shift in how social media works, with platforms increasingly pushing users toward institutional content rather than updates from their own networks. In the sample analysed, Reform UK voters were also most likely to see content from brands and news organisations.
Algorithms and online isolation
The IPPR argued that recommendation systems used by major platforms are not neutral. Instead, it said they can amplify division by narrowing the range of content people encounter and reducing the amount of interaction they have with posts from friends, family and acquaintances.
The study focused on what users were shown in their feeds across five platforms: Instagram, Facebook, X, Bluesky and TikTok. By comparing the mix of posts delivered to different political groups, the thinktank said it found evidence that social media can reinforce separate online environments for different kinds of users.
The findings add to wider concerns about the role of algorithms in shaping political and social experience online. According to the IPPR, the imbalance between personal posts and professionally produced content may leave some users more isolated from their own networks and more dependent on platform-driven recommendations for what they see.
Reform UK voters’ feeds stood out in the analysis because they were less likely than those of Green party voters to include posts from known contacts, and more likely to include material from media outlets and commercial accounts.
The report’s central warning is that these patterns may not simply reflect individual preferences, but the way platform systems sort and prioritise content. By doing so, the IPPR said, social media can deepen division rather than connect people to the communities around them.
The study was published after analysing posts shown to users across the platforms, and the thinktank said the results show how online feeds can differ sharply depending on the audience.
