Representative Bureaucracy and Democracy: A Comparative Analysis

Research Question: Does democracy facilitate bureaucratic gender representation? Motivation: The theory of representative bureaucracy argues that a diverse and representative bureaucracy contributes to a more equitable distribution of social benefits. Extensive research has supported this claim by showing the positive relationship between descriptive representation and policy outcomes for underrepresented groups. Most studies on this issue, however, have been conducted in the United States and European countries with high levels of democratic values and effective institutions relative to the rest of the world. This paper aims to contribute to this line of literature by testing whether democratic deficits in countries hinder the transition from descriptive representation to substantive representation. Design: We first examine the basic notion of representative bureaucracy and then test the role of democracy in the mechanism. Focusing on gender representation in education, we test the effects of having more female teachers on girls’ performance at the individual level and then investigate how the relationship would vary depending on the different levels of democracy across forty-four countries. Data: We use individual-level panel data from the Trends in International Mathematics and Science Study (TIMSS) from 1995 to 2019. The dataset includes at least 4,500 to 5,000 students from sixty-three countries per year and their exam scores in math and science. The exams are translated and taken in each country’s language, which enables us to compare scores between countries. We also obtain various democracy measures from Cheibub, Gandhi, and Vreeland, Freedom House, Polity, World Bank Governance Indicators, Bertelsmann Transformation Index, and the Economist Intelligence Unit. Methods: We use OLS regression for our analysis. Given that our data includes multiple levels, such as individuals (students), schools, and countries, ignoring different levels in the dataset can bias our estimates downward. To account for this issue, we employ regression with clustered robust standard errors by country. We also include year-fixed effects. Implications: Primary findings suggest that gender representation is more effective in countries with high levels of democracy. This result highlights the important role of democracy in facilitating the effects of bureaucratic representation. Our study sheds light on how institutional structure can play a role in the translation of descriptive representation into substantive representation. This study also advances our understanding of the effect of representative bureaucracy from a comparative perspective. Authors: Kenneth J. Meier, Distinguished Scholar in Residence, American University Seung-Ho An, Assistant Professor, University of Arizona Miyeon Song, Assistant Professor, Rutgers University – Newark