Since the second half of the twentieth century, ethnicity has come to play an increasingly important role in political phenomena, especially in the justification of armed conflicts. To explain this particular role that ethnic identities seem to play, recent research highlights the strategic mobilization of ethnic identities by elites to obtain and legitimize positions of power. Based on this work, this research aims to answer two main gaps that characterize quantitative studies on the subject and which prevent a better understanding of the role of ethnicity in the acceptance of leaders’ authority. First, quantitative research on ethnicity typically fails to take the social constructivist stance seriously as shown by the use of measurements (i.e. "fractionalization" or "polarization" indices) that treat ethnic identity as a descriptive characteristic, regardless of its subjective relevance for individuals. Second, research generally focuses on either the societal (national) level or the individual level when trying to understand the relationship between ethnicity and violence, and therefore confuses dynamics that happen at the national level with those occurring at more local scales. Relying on the spiral of silence theory and the social representation approach, I propose the following hypothesis to explain how strong leadership may become uncontested: local contexts where the importance of ethnic identities substantially changes are characterized by a questioning of the political norms (i.e. what political stances can be publicly enacted) and constitutes therefore places where otherwise censored political views (e.g. authoritarian) may come to dominate the public sphere. To test this hypothesis, I use data from the first two rounds of the Afrobarometer survey collected in 10 African countries. Using multilevel logistic models, I examine whether regional change in the salience of ethnic identities interacts with the political attitude of individuals (authoritarian vs Democrats) to predict their political participation. As hypothesized, results show that regional volatility selectively affects the enactment of political views. However, the pattern is more complex than predicted and suggests that the spiral of silence framework is be too simplistic to explain processes occurring in these contexts.