Download the Study

Accountability and Flexibility in Public Schools: Evidence from Boston’s Charters and Pilots

How Effective Are Boston’s Charter and Pilot Schools?

School Reform, May 2011

Using stu­dent assign­ment lot­tery esti­mates, a study by Atila Abdulkadiroglu, Joshua Angrist, Susan Dynarski, Thomas Kane, and Parag Pathak unveils large and sig­nif­i­cant score gains for Boston’s char­ter stu­dents in mid­dle and high school, and in con­trast, mod­est or neg­a­tive effects for Boston’s pilot school students.

While char­ter schools and pilot schools both serve as alter­na­tives to tra­di­tion­al pub­lic edu­ca­tion, fur­ther inspec­tion of dif­fer­ences in their key fea­tures sug­gests pos­si­ble expla­na­tions for the dis­crep­an­cy in stu­dent achieve­ment shown by this study.  Pilot schools have some of the inde­pen­dence of char­ter schools but are in the Boston Public School dis­trict and are cov­ered by some col­lec­tive bar­gain­ing pro­vi­sions. Differences in school struc­tures may orig­i­nate from col­lec­tive bar­gain­ing agree­ments in Boston Public Schools which make it rel­a­tive­ly expen­sive for pilot schools to expand instruc­tion­al hours and staffing which favor teacher senior­i­ty over class­room effec­tive­ness.  Charter schools tend to embrace longer day and school year and young teach­ing staff.  In addi­tion, most of the char­ter schools in our lot­tery sam­ple embrace ele­ments of the No Excuses mod­el, an instruc­tion­al par­a­digm that is not com­mon in pub­lic schools, pilot or oth­er­wise.  Many of the char­ter schools in the study aspire to boost minor­i­ty achieve­ment, so a nat­ur­al bench­mark of char­ter effec­tive­ness exam­ined is the black-white test score gap.  Overall, while lot­tery esti­mates show sig­nif­i­cant gains for char­ter stu­dents, the gains for pilot stu­dents are insignif­i­cant.  Also, char­ter schools with bind­ing assign­ment lot­ter­ies appear to gen­er­ate larg­er gains than oth­er char­ters.  The con­clu­sions from the lot­tery study are sup­port­ed by the obser­va­tion­al study when using the same set of char­ter schools.