Inequality in frontline communication : Bureaucrats talk differently to men and women

Abstract

While gender biases in public service delivery are well-established, we still lack empirical insights on the underlying behavioral micro-mechanisms. This paper contributes towards closing this perennial gap by investigating gender-differences in the complexity and emotionality of verbal bureaucrat-client communication. We build on a dataset comprising 154 dialogs recorded across different local public services in Germany. Combining rule-based and machine learning classification, we analyze differences in verbal administrative communication across 20,000 utterances. We find no association between bureaucrats’ gender and their communication. Conversely, clients’ gender yields a significant difference, with officials communicating more complex and emotional when interacting with male clients. No differences prevail for gender-matching. As the first study to systematically examine implicit (gender) biases in bureaucrats’ communication, the paper advance our existing understanding of the micro-mechanisms of administrative inequality: The findings contradict expectations from gender socialization theory, they confirm expectations linked to gender stereotypes, and they challenge the idea that in-group settings reduce stereotypical biases at the level of communication.