Charter Schools

The New York Times has been giving a bit of publicity to a study by the American Federation of Teachers, which purports to show that charter schools actually do worse than public schools at all levels. The Times had a typical editorial announcing that the study was a "devastating setback" for the "Bush administration's education program." (Just the day before, a "news" story was phrased in similar terms, stating that the study "dealt a blow to supporters of the charter school movement, including the Bush administration.")

The problem is that the AFT study is misleading. The sad fact that we still have vast racial disparities in education in this country. It looks as if those disparities virtually completely explain the overall results.

The AFT study itself notes on page 10 that charter schools contain nearly twice the proportion of black students, and that this might affect the overall results:

In the NAEP sample, charter schools were approximately twice as likely as other public schools to enroll black students (33 percent compared to 18 percent) . . . . The question we turn to next is whether charter schools' disproportionate enrollment of minority students, whose achievement is generally low, may explain the low average performance of charter schools.

The study then states on page 11 that "[a]chievement gaps based on race/ethnicity in grade 4 are about the same in charter schools as in other public schools."

These gaps are substantial. For example, the AFT study's Table 6 shows that in public schools, 74% of white 4th-graders are reading competently but only 40% of black 4th-graders.

Thus, there is a straightforward reason that almost completely explains¹ why charter schools have slight lower scores on standarized tests than do public schools. It is unfortunate indeed that such racial disparities exist, but it is misleading to blame charter schools for the problem.

* * *

¹Indeed, a little bit of math demonstrates that the racial disparity may completely explain the end results.

Step 1: Look at Table 1. Notice that the overall difference between charter schools and public schools is 6 or 7 points for 4th graders.

Step 2: Look at Table 6. Notice that the charter schools have fewer whites and more blacks.

Step 3: Figure out the average score of a charter and a public school that has exactly the percentages of each minority shown.

I.e., multiply 47.1 by 241; multiply 33.4 by 213; multiply 14.6 by 219; then for the "other" column, assume that they had an average score of 220, and multiply 4.9 by 220. Add the results. Then divide by 100. The end result: An average math score of 227.

Repeat the same process for the public school figures. The end result: An average math score of 233.75.

In other words, if you look at the average scores of various racial groups, and combine that with the average demographics, it turns out that the predicted 4th grade math score for public schools is 6.75 points higher than the predicted score for charter schools. This is the more than the 6 point overall difference that was shown in Table 1 (as if it demonstrated a deficiency in charter schools per se).

Now try 4th grade reading:

Average charter school with racial proportions shown in Table 6: Average score of 210.72 (assuming that unknown students had score of 200)

Average public school with racial proportions shown in Table 6: Average score of: 216.12

Predicted disparity based on demographics: 5.4 points.

Actual disparity as shown in Table 1: 7 points.

So that nearly explains the entire overall reading disparity as well.