Designed to free educators from bureaucratic regulations and encourage innovations, charter schools are public schools that have been granted autonomy over their operations. Although charter schools have been authorized in most states and serve more than one million students, evidence is lacking on whether these publicly funded institutions promote student learning adequately.

Mathematica's charter school evaluations are helping to inform the debate. We are conducting studies of KIPP (Knowledge Is Power Program), a national network of open-enrollment college preparatory charter schools in underserved communities; charter management organizations (CMOs), nonprofit organizations that start and manage new charter schools; a national set of charter middle schools that hold admissions lotteries; and charter school outcomes in several states.

A key question in the debates about school choice is what happens to students' educational achievement when families are given vouchers that cover at least part of the cost of private school tuition. Working with researchers at Harvard University, Mathematica conducted an experiment to test a school choice program in New York City, offering vouchers to low-income families. The results showed positive impacts on test scores for African Americans, but no impacts on test scores for Latinos.