The next session at the 2026 International Communication Association conference in Cape Town is on influencers and politics, and starts with a paper by Christian Pipal. Influencers are of course often also political communicators now, and especially reach young audiences who do not follow the mainstream news; usually non-political influencers are especially influential when they post political content, in fact.
But we still don’t know nearly enough about the vast bulk of influencers, especially at the micro- and nano-level of influence activity. It is also important to do more work on how such content travels, and how audiences engage with it, across what platforms. Audience engagement is also a reward signal, of course, and creators learn from this for their further activities.
This paper addresses these questions by investigating how content patterns are distributed across the population of influencers, and how engagement differs between political and other topics. It worked with some 760,000 Instagram posts from some 1,800 social media influencers in Germany and Austria across nano-, micro-, and macro-levels of influence as defined by their followers numbers. Posts were scraped from their profiles, and were converted from audiovisual formats to text; they covered both formal and lifestyle politics categories.
The political nature of posts was identified using LLMs, and in the end some 17,000 or 2.25% of posts were identified as distinctly political. These were divided into 23 different topic categories, and coded for direct references to political systems; the 23 topics were collapsed into a broad distinction between formal and lifestyle politics, too.
Overall, there were roughly double the number of posts in Germany compared to Austria; the volume over time was stable in Austria, and rose substantially in Germany ahead of the 2025 snap election. Micro-creators in Germany produced more political content than macro-influencers, but this was not the case for nano-influencers; political posts received some 20% more likes and 30% more comments than non-political posts by the same creator; this boost is stronger for smaller creators.
Such engagement rewards also predict further political posts from the same creator: one standard deviation of higher engagement increases the chance of further political posts by 17%. Higher lifestyle engagement also predicts and escalation from lifestyle politics into more formal political discussion; this phenomenon is specific to lifestyle content engagement.
Lifestyle-anchored political talk is not a substitute, but a route into formal political discussion, then; this is driven by audience reward for individual creators. This is a learnt behaviour for creators, then: it is shaped by an incentive structure rather than simply by intrinsic strategic choices. However, it is also important to note that this study observed only overall engagement with posts that were still available at the time of data gathering: posts deleted due to negative engagement could not be considered, but indeed negative engagement might also have been a factor in the move towards formal political posting.
