We have had quite a lot of back and forth on this blog about Boston charter schools, in anticipation of the vote this November in Massachusetts about lifting the charter cap and adding another 12 charter schools every year forever. Pro-charter advocates argue that the Boston charters are not only outstanding in test scores but that their attrition rate is no different from that of the public schools, or possibly even less than the public schools.

Jersey Jazzman (aka Mark Weber) is a teacher and is studying for his doctorate at Rutgers, where he specializes in data analysis.

In this post, he demolishes the claim that Boston charters have a low attrition rate. As he shows, using state data,

In the last decade, Boston’s charter sector has had substantially greater cohort attrition than the Boston Public Schools. In fact, even though the data is noisy, you could make a pretty good case the difference in cohort attrition rates has grown over the last five years.

Is this proof that the independent charters are doing a bad job? I wouldn’t say so; I’m sure these schools are full of dedicated staff, working hard to serve their students. But there is little doubt that the public schools are doing a job that charters are not: they are educating the kids who don’t stay in the charters, or who arrive too late to feel like enrolling in them is a good choice.

This is a serious issue, and the voters of Massachusetts should be made aware of it before they cast their votes. We know that charter schools have had detrimental effects on the finances of their host school systems in other states. Massachusetts’ charter law has one of the more generous reimbursement policies for host schools, but these laws do little more than delay the inevitable: charter expansion, by definition, is inefficient because administrative functions are replicated. And that means less money in the classroom.

Is it really worth expanding charters and risking further injury to BPS when the charter sector appears, at least at the high school level, to rely so heavily on cohort attrition?