Institutions, Parents, Communities and Student Achievement.

There is a contentious debate over the relationship between institutional design and student academic achievement (Smith 2005). We further the school choice debate by considering the relationship between different types of school institutions, parental and community variables and third grade student test scores using the Early Childhood Longitudinal Study (ECLS-K) data. Though it has been infrequently used by political scientists (Cox and Witko 2008), this panel data set allows us to overcome problems that plague many school choice studies by examining the academic performance of the same students over time, holding unobserved child-level variables constant. It is also unique in that it follows younger children. Since it is large and nationally representative, this data set also allows us to make generalizable conclusions. We demonstrate that academic performance is unlikely to be improved by simply attending a school of choice, public or private, and instead argue that parental variables and other aspects of a school's institutional environment and community are more important, though still modest, determinants of academic achievement. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]