Competitive Effects of MeansTested School Vouchers测试手段的竞争效应.pptVIP

Competitive Effects of Means-Tested School Vouchers David Figlio, Northwestern and NBER Cassandra Hart, Northwestern December 2009 Introduction School choice options have become increasingly prevalent in recent years Considerable attention paid to potential competitive effects of choice, both positive (efficiency) and negative (cream-skimming) Challenging to gauge competitive effects because of interrelationship between private school supply and public school performance Prior literature: cross-sectional studies of private school penetration in US and international; Milwaukee vouchers; Florida school grades; Sweden voucher program introduction; Chile voucher cross-section This paper Study introduction of large new school voucher program; use introduction of this program as source of plausibly exogenous variation that increased demand for private school options after 2001 Look at quantity and variety of nearby private school options in year prior to program announcement, which could generate variation in access to the program Florida is large and varied in its pre-program private school supply Identifying off of a policy change; use student data from 99-00 through 06-07 How vouchers might affect public schools Competition effect Composition effect Resource effect First year of program was before any students left the public schools – but were applying Work in progress: still trying to tease out three effects in the “mobility” years of the program Florida’s Corporate Tax Credit Scholarship Program Funded by fully tax creditable corporate contributions to one of three Scholarship Funding Organizations, each with geographic range; total contributions capped by Legislature Began with 20,000+ students, now at 27,000+ students Students below 185% of poverty line and attending public school in prior year (or entering grades K/1) eligible; renewal requires income below 200% of poverty line Initial voucher was $3,500; now, it’s $3,950-$4,100 (around 90% of average “rack ra