This paper brings new evidence on the impact of income inequality on public decisions. Using a new panel database on French municipalities’ accounts, and on households’ income distribution at the local level, I estimate the impact of income distribution on municipal policy. This paper is the first to investigate this issue by simultaneously using a high number of comparable observations and identifying deciles of the income distribution which matter. After controlling for municipal fixed effects and for the dynamics of municipal incumbents’ decisions, I find no impact of income inequality on operating spending, but a strong positive impact on municipal infrastructures. Evidence suggests that an increase in income inequality by 1% leads on average to an increase in the value of municipal infrastructures between 0.06% and 0.18%. Importantly, I find that this result is driven by variations in bottom and top deciles. There is clear evidence that additive public facilities associated to more inequality are due to higher tax rates. When poorest individuals get poorer, or when richest ones get richer, municipal incumbents decide to increase the amount of infrastructures by increasing local taxation. These results suggest that what matter in public decisions are the extreme parts of voters’ income distribution, and that lower bottom incomes and higher top ones both lead to a higher size of government.