A few months ago, I published a post about the charter schools in Minneapolis, which are expanding rapidly in that city, replacing unionized teachers with young and inexperienced Teach for America teachers.

The burgeoning of charters in Minneapolis has something to do with a very powerful family named Kramer. EduShyster reported that the family is a powerful organization for corporate reform. She writes:

Readers: meet the Minneapolis Kramers. Father Joel is the former publisher of the Minneapolis Star Tribune and took home $8 million when the paper was sold to McClatchy. These days he presides over Minnpost.com and a brood of young rephormers. Son Matt is the president of Teach for America, in charge of TFA’s “overall performance, operations, and effectiveness.” Son Eli, another former TFAer, is the executive director of Hiawatha Academies, a mini charter empire in Minneapolis. Meanwhile, daughter-in-law Katie Barrett-Kramer is a former TFAer who now serves as director of academic excellence at Charter School Partners, a non-profit organization devoted to expanding the number of charters in Minneapolis, including the ones her brother-in-law runs.

Now I have acquired a deep thirst just writing about the Kramer siblings and their dedication to the civil right$ i$$ue of our time. But there’s still more. Matt, who with his brother attended the tony Breck School (which I suspect is likely not a ‘no excuses’ school), also sits on numerous rephorm boards. Matt is the chair of the board of 50Canand a member of the board of Students for Education Reform.

EduShyster returned to the charter-TFA empire of the Kramer family in another post.

Who knew that one family could create a separate school system in a major American city?

In my post on the Minneapolis charters, I noted a study by Myron Orfield and the Institute on Metropolitan Opportunity at the University of Minnesota Law School, which was very critical of the charters, saying that they were more segregated than public schools and had worse results. John Bloomberg, an education writer for Bloomberg News, is also cited in the same post, expressing astonishment at the hyper-segregation in the charter schools of Minneapolis.

Now charter advocates have challenged these claims.

Orfield responds to them here.

Charter School Partners’ (CSP) January 6 blog post titled Minnesota Charters 2014: Part I: Building a high-impact charter sector, Closing the opportunity/achievement gap cited recent work by the Institute on Metropolitan Opportunity (IMO) updating our work on charter schools in the Twin Cities. The post seriously misrepresents some of our findings and we request that you make this response available on your web site along with your original post. (Readers are invited to download IMO’s original study on charters in the Twin Cities and two updates at http://www.law.umn.edu/metro/school-studies/school-choice.html).

Most importantly, the post implies that our work does not properly control for the fact that charters serve higher percentages of low-income students. This is simply not true. Our work shows that charters still under-perform their traditional counterparts even after controlling for the effects of school poverty rates on reading and math performance.

Two methods were used. A multiple regression analysis demonstrated that charter elementary schools have lower achievement rates on average after controlling for student poverty, race, special education needs, limited language abilities, student mobility rates and school size. Indeed, contrary to the claims in the January 6 post, the most recent statistical results imply that all else equal, the gap between charters and traditional schools widened between 2010-11 and 2012-13. In 2012-13 the proficiency rate for charters was 11.2 percentage points lower than traditional schools for math and 5.9 percentage points lower for reading. In 2010-11 the gaps were just 7.5 for math and 4.4 for reading.

Consistent with the earlier studies and other research, the multiple regression analysis showed that student poverty (measured by eligibility for free or reduced price lunch) was the dominant factor in the performance of schools in 2012-13. In 2012-13, the math performance of students in only 31 percent of charter schools was better than what would be expected given their poverty rate alone. The rest, 69 percent, under-performed expectations. Consistent with the regression analysis, this represented a significant step back from 2010-11 when 51 percent of charters out-performed expectations. Similarly, the reading performance of students in just 36 percent of charter schools was better than expected (compared to 39 percent in the 2010-11 analysis).

These results and other analysis in the report also refute the statement that the IMO “study ignores the impact of that high-achieving charters are having on these critical populations.” IMO’s study does in fact acknowledge that a few high-poverty charters are performing very well. But it also shows that a larger number of high-poverty charters are performing very poorly, even when accounting for the fact that they serve mostly low-income students. In fact, Charts 5 and 6 in the report clearly show that there are more students in under-performing charters than in the high-achieving ones.

While IMO acknowledges that a few charters are doing well, CSP essentially ignores all of the charters that are doing poorly. Only the “Strategic Framework” schematic in the post acknowledges poorly performing charters at all and it does so with an illustration that badly understates the size of the problem. Even in CSP’s own chart later in the post (“Minneapolis Charter Schools – 3 Year Average”), poorly performing charters greatly outnumber high-performing charters.

CSP is correct when they note that the data now available to IMO does not allow us to fully compare the impact of charter and traditional schools on individual students. However, if positive impacts on individual students are greater in charters then one would expect the proficiency gap between charters and traditionals to narrow over time. Once again, this is not what is happening. Despite the fact that a few charters seem to have shown improvement, the gap has actually widened in recent years. It is also worth noting that the method used in the CREDO study cited in the post does not represent the gold standard of analysis in this field, raising complicated research questions well beyond the scope of these comments that IMO addressed in its original study.

The post also says that, because the IMO study is region-wide, the work “ignores compelling data that show that charters in Minneapolis and St. Paul are significantly outperforming district schools…” We believe the region-wide scale is appropriate for a variety of reasons. Most importantly, charters are growing rapidly in the suburbs and suburban charters now represent about 40 percent of total regional charter school enrollment. In addition, school choice programs—the Choice is Yours Program in particular—make a number of suburban schools available to low-income city students, implying that the proper comparison group is in fact regional. (IMO’s work shows that, in contrast with the results for charters, schools available to low-income students through the Choice is Yours Program tend to out-perform their traditional and charter peers after controlling for poverty and other school characteristics.)

Nor does restricting the IMO analysis to Minneapolis and St. Paul provide the kind of “compelling” evidence of superior charter performance that CSP claims. When the statistical analyses described above are repeated for city schools alone the results show no advantages for charters. Not surprisingly, the multiple regression analysis (which controls for poverty and other school characteristics) shows that the gap is narrower in the cities but the results still imply that, all else equal, charters in the cities under-perform their traditional peers. (The gaps are roughly three percentage points for both tests. The coefficients are negative, but not statistically significant. While this is admittedly only weak evidence that city charters do worse, it is certainly not “compelling” evidence that they do better.)

The simpler comparisons controlling only for low-income percentages show similar results. City charters are slightly more likely to perform more poorly than expected given their poverty rates than traditional schools.

Finally, it is important to note that IMO’s studies look at much more than simply school performance (however defined). IMO’s original study and two updates have also examined:

• Whether the charter system is more or less segregated than traditional schools. They are much more segregated and the situation is not improving. A disturbing proportion of charters in the Twin Cities are essentially single-race schools. In sharp contrast with the traditional system, where the percentage of schools which are integrated has increased steadily, the share of integrated charter schools has been stagnant. As a result, charter school students of all races are much more likely to be attending segregated schools than their counterparts in traditional schools.
• The effects on the Minneapolis and St. Paul school districts of enrollment declines resulting from charters. Rapid enrollment declines hurt the districts because they require costly actions like school closures, teacher and staff cutbacks, and administrative reorganizations. As a result, districts losing students must devote effort and resources to deal with the costs of decline, often to the detriment of other educational priorities. Minneapolis and St. Paul now lose between 15 and 20 percent of their resident students to charters and more than half of total enrollment losses in the last decade have been to charters. As a result, both districts have teetered from one financial crisis to another.
• Charter school closures. Although data for this is hard to come by, press coverage of closures suggests that charters are more likely to close as the result of financial mismanagement or, in some cases, malfeasance than because of the poor performance that many exhibit.
• The growing, negative effects of suburban charters on the ability of traditional districts to pursue pro-integrative reforms. Increasing numbers of predominantly white suburban charters are locating near significantly more diverse traditional schools—schools which are often unstable and vulnerable to rapid racial and economic transitions. Whether by intent or not, more and more suburban charters are facilitating white flight from increasingly diverse traditional schools in the suburbs.

CSP seems to have little to say about these topics, only promising to discuss closures in a future post. Test results are not the only thing that schools produce. CSP should abandon its blinkered view and widen its work to include the very significant problems with the state’s current charter model.