Charter School Demand and Effectiveness, A Boston Update

New Findings, School Reform, October 2013

As in the 2009 report, Informing the Debate, we find that attend­ing a char­ter school in Boston sig­nif­i­cantly boosts MCAS scores and pro­fi­ciency lev­els. Positive test  score effects from the most recent years where our lot­tery sam­ple cov­er­age is nearly com­plete are  of sim­i­lar mag­ni­tudes. Non-lottery results con­firm the lot­tery results for char­ters from which we were able to col­lect lot­tery records, and point to lower per­for­mance for closed  char­ters and those with­out com­plete records. However, test scores are only one part of the story. This report also pro­vides evi­dence on the demand for char­ter schools.

Many stu­dents in Boston apply to a char­ter, with appli­ca­tion rates ris­ing in the past few years,  espe­cially for mid­dle schools. A major­ity of stu­dents who apply get an offer to at least one  school, but not all stu­dents accept these offers. A third of mid­dle school stu­dents and 60 per­cent  of high school stu­dents choose other options. Many of these offers arrive after the lot­tery, a  con­tribut­ing fac­tor to low take up rates, along with the many school options avail­able in Boston,  espe­cially for high school. Offer rates at Boston char­ters are broadly sim­i­lar to the offer rates for first choice schools in the BPS assign­ment mechanism.

Charter school stu­dents tend to have some­what higher early test scores than the gen­eral BPS  pop­u­la­tion. This most reflects that higher  scor­ing stu­dents are more likely to apply in the first  place. The pro­por­tion of stu­dents with spe­cial needs and English lan­guage learn­ers is also lower in  the appli­cant group than in the gen­eral pop­u­la­tion. Importantly, how­ever, gaps between char­ter  appli­cants and non-applicants are shrink­ing. In the most recent year, we see almost as many spe­cial  edu­ca­tion stu­dents apply­ing as exist in the BPS pop­u­la­tion. At the same time, some gaps remain.  This is impor­tant because our analy­sis of char­ter effec­tive­ness (here, as in ear­lier work) uncov­ers  sub­stan­tial dif­fer­ences in impact. Students from groups least likely to apply, includ­ing English  lan­guage learn­ers and stu­dents with low achieve­ment scores, are those for which achieve­ment gains are likely to be the largest.