Abstract

While democratization and enfranchisement are known to benefit minority groups in the long run, sudden increases in political representation can disrupt existing power balances, provoke resistance, and lead to worse policy outcomes in the short run. We document and explain this pattern. In our theoretical model, conflict and backlash are triggered by a sufficient increase in political power if preferences are sufficiently different. We exploit the introduction of an affirmative action measure in Italian local elections, which led to an exogenous increase in female political representation in small municipalities. Using a Difference in Discontinuity design, we document that, in line with the theory, moderate increases in female representation led to higher day care spending, while large increases resulted in lower spending on this gender-sensitive issue. Higher council dissolution rates and null effects on non-gender-related policy areas support the interpretation of the evidence suggested by the theory. Several robustness checks and evidence from Spanish data also support the internal and external validity of our findings.