Keele Research Repository

Collective action is a key driver of social and political change within societies. So far, the main factor mobilizing individuals into collective action remains the extent to which they feel identified with a protesting group (i.e., social identification). Although the link between social identification and collective action is well-established, current evidence relies mostly on self-report data. To tackle this issue, we combined real-life protest counts in the US (2017-2020) with online search data (Google Trends) for pronouns indicating a ‘group’ mindset (first-person plural pronouns; e.g. ‘we’, ‘us’). Time series analyses indicated that weekly fluctuations in searches (N = 164) predict both protest and protester counts over time. Confirmatory mixed-models then showed that a 1% increase in pronoun searches was linked with +13.67% protests 95%CI[4.02,23.32] and +47.45% protesters 95%CI[26.54, 68.36] the following week. These original results have important implications for the ecological study and quantification of collective action dynamics in psychology.