Traditionally, bureaucrats have been viewed as a stereotypical example of employees with flat pay schedules and low-powered incentive schemes. This paper challenges that view by providing evidence that the wages of a particular group of senior bureaucrats - city managers - are tightly connected to their performance as measured by city growth. Additional tests indicate that these results reflect reward for performance, rather than rent extraction, as exogenous shocks to city growth do not affect city managers’ wage. First, I show that the salaries of city managers do not react to observable exogenous shocks to city performance. Next, I demonstrate that performance affects city managers’ wages not only in the city in which they are currently employed, but also in the city in which they work afterwards. Finally, I find that in cities with council-manager forms of government, the wages of mayors - who do not play important roles in running the cities with council-manager form of government - are not sensitive to city performance.