How do charters do? Depends on the standard

As I reported last week, Ohio charter schools received a bad rap in recent articles by The Economist. After singing the praises of charters in some of America’s largest cities, The Economist went on to disparage Ohio’s charters, stating that they “have done badly.” I didn’t disagree with their appraisal.

Why the agreement? It’s because the standard matters.

So in Ohio, charters are "bad" compared to what standard? To answer, I take a slice of data from Cleveland to look at the performance of its charter schools relative two comparison groups. First, I compare how Cleveland’s charters stack up against Cleveland Municipal School District (the city’s traditional public school). Second, I compare Cleveland's charters against a broader set of public districts--all districts in Cuyahoga County, which includes Cleveland Municipal, poorer inner-ring suburban districts, and some affluent suburban districts.

I use the fourth grade math proficiency rate—essentially, the proportion of students who “pass” Ohio’s annual standardized test in a given grade and subject—for the 2010-11 school year. And by using what’s called a “z-score” in statistics, I calculate how far each school's proficiency rate is above or below the average proficiency (pass) rate.[1] A school with a positive score has an above-average proficiency rate; vice-versa, a school with a negative score has a below-average rate.

Figure 1 shows how charters compare against their district peers. Each bar indicates a school: charters are shown in red and district schools in grey. The vertical axis indicates schools’ z-scores—again, indicating how far their proficiency rate is from the group average proficiency rate.

On the left chart (figure 1A), Cleveland charters are pretty evenly distributed above and below the average. Conclusion: Cleveland’s charters do just about the same as their district peers. So far so good; but remember, Cleveland Municipal is one of Ohio’s worst traditional public school districts and consistently one of the nation’s worst urban districts. For now, put the champagne on ice—performing on par with one of the worst districts in Ohio and the nation should be no cause for celebration for charters.

When I expand the geographic scope to all Cuyahoga County (figure 1B), charters, as a group, do worse. More charters fall below the average line and fewer remain above the average line. Note the greater density of the red lines below zero. The rise in the average proficiency rate when higher-performing suburban schools are included causes this downward shift. In other words, when the standard gets higher, charters do worse. (Consider, though, that a few charters compete with the best schools in this group. But these remain the “Needles in the Haystack.”)


Figure 1: Fourth grade math proficiency rates, scaled to the average rate, 2010-11. (A) Cleveland charters versus Cleveland Municipal School District schools. (B) Cleveland charters versus Cuyahoga County public school districts, inclusive of Cleveland Municipal. Data source: Author’s calculations based on Ohio Department of Education data.

Have Ohio’s charters “done badly?” Depends on your standard. If you favor low standards, then perhaps there’s no reason for concern. As a group, Cleveland’s charters fail just the same as their peers in Cleveland Municipal. But if the standard is high—for charters to outclass schools across their county and state—then charters, as a whole, disappoint. I’ll stick with high standards.


[1] The z-score calculation is (proficiency rate of building x – average proficiency rate) ÷ standard deviation. Z-scores are in standard deviation units and assume a normal distribution (bell-shaped curve). The shape of the curve is determined by the standard deviation.

More By Author

Related Articles