"Rich-Get-Richer"? Analyzing Content Creator Earnings Across Large Social Media Platforms

This paper examines whether monthly content creator earnings follow a power law distribution, driven by compounding 'rich-get-richer' dynamics (Barabasi and Albert 1999). Patreon creator earnings data for 2018, 2021, and 2024 for Instagram, Twitch, YouTube, Twitter, Facebook, and Patreon exhibit a power law exponent around $\alpha = 2$. This suggests that algorithmic systems generate unequalizing returns closer to highly concentrated capital income and wealth, rather than labor income. Platforms governed by powerful and compounding recommendation systems, such as Instagram and YouTube, exhibit both a stronger power law relation (lower $\alpha$) and lower mean, median, and interquartile earnings, indicating algorithms that disproportionately favor top earners at the expense of a 'middle class' of creators. In contrast, Twitter and Patreon have a more moderate $\alpha$, with less earnings inequality and higher middle class earnings. Policies which incentivize the algorithmic promotion of longer-tail content (to explore more and exploit less) may help creator ecosystems become more equitable and sustainable.