Download the Study

Explaining Charter School Effectiveness

Explaining Charter School Effectiveness

School Assignment, School Reform, December 2012

Estimates using admis­sions lot­ter­ies sug­gest that urban char­ter schools boost stu­dent achieve­ment, while char­ter schools in oth­er set­tings do not. Using the largest avail­able sam­ple of lot­teried appli­cants to char­ter schools, we explore stu­dent-lev­el and school-lev­el expla­na­tions for this dif­fer­ence in Massachusetts. In an econo­met­ric frame­work that iso­lates sources of char­ter effect het­ero­gene­ity, we show that urban char­ter schools boost achieve­ment well beyond that of urban pub­lic school stu­dents, while non-urban char­ters reduce achieve­ment from a high­er base­line. Student demo­graph­ics explain some of these gains since urban char­ters are most effec­tive for non-whites and low-base­line achiev­ers. At the same time, non-urban char­ter schools are uni­form­ly inef­fec­tive. Our esti­mates also reveal impor­tant school-lev­el het­ero­gene­ity with­in the urban char­ter sam­ple. A non-lot­tery analy­sis sug­gests that urban char­ters with bind­ing, well-doc­u­ment­ed admis­sions lot­ter­ies gen­er­ate larg­er score gains than under-sub­scribed urban char­ter schools with poor lot­tery records. Finally, we link char­ter impacts to school char­ac­ter­is­tics such as peer com­po­si­tion, length of school day, and school phi­los­o­phy. The rel­a­tive effec­tive­ness of urban lot­tery-sam­ple char­ters is account­ed for by those schools’ embraces of the No Excuses approach to urban edu­ca­tion.