With the goals of improved test scores, lower absenteeism, and higher graduation rates, the 1988 Chicago School Reform Act decentralized school governance by forming Local School Councils (LSCs). LSCs are an elected group of parents, community members, and school personnel who are responsible for the hiring, evaluation, and contract renewal decisions of principals. Ongoing concerns that LSCs have not substantially improved schools, especially in high poverty areas, have led some to push for more centralized control. In this paper we develop a conceptual model of the relationship between the LSC, the Board of Education, and the principal that incorporates differences in incentives, management capacity, and information. We then use Chicago Public School administrative and survey data and accountability ratings to examine the factors associated with principal contract renewal. We are particularly interested in the extent to which the probability a principal is retained despite evidence of poor job performance as measured by student test scores, graduation rates, other student outcomes, and survey responses differs for schools with LSCs and those managed centrally. Among schools managed by LSCs, we will examine differences by school demographic compositions including the rate of poverty and share of students who speak English as a second language as well as differences between elementary and high schools. Finally, we will simulate the cumulative impact of retaining an ineffective principal on student and teacher turnover and student academic outcomes.