Let's think for a bit about the available school choice evidence. Milton Freidman's argument includes treating parents as if they were consumers of education. In this claim, schools are said to be failing children. Therefore, if parents are consumers, they ought to have a choice in where to send their children to school. In short, parents should have the option to remove their children from low-performing schools and send them to high-performing schools.
School performance is defined as average student performance on standardized tests. While this is a measurable outcome, it leaves open many questions. Are standardized tests measuring performance or something else? Are standardized tests the only way to understand student performance? Is teachers' professional judgment out of bounds for determining student performance? Does race, gender, culture, or income level impact student performance? These questions will be discussed below.
Available School Choice Evidence: Failing Schools Claim
When thinking about available school choice evidence is important. There are several things that are well understood about standardized testing. One is that these tests do not generally correlate to student performance or growth. They do correlate well with IQ testing where one expects IQ to not vary over time. In many of the reporting schemes for standardized tests, growth as Grade Level Equivalency (GLE) is a way of reporting a mean or average score adjusted to the student's grade. It is also well-settled that standardized tests should never be used in isolation as a diagnostic tool for evaluating students. In the United States, standardized tests are used to measure school performance, something for which these tests were never designed to measure. A school in which average scores fall below the 20th percentile are defined as failing. There is no other tool available to measure school performance. It is the law.
Downey, von Hippel, and Hughes (2008) looked at measures other than testing that could be used to measure school performance. They tried a unique approach by measuring the extent of student learning over an entire school year. Using data gathered from the Early Childhood Longitudinal Study (ECLS) conducted by the National Center for Educational Statistics, which tracked a cohort of students from grades k-5. 4,217 children were tracked in the ECLS coming from 287 school districts across the United States. They looked at performance testing and at learning rate statistics.
Their findings were unexpected, some have even called them shocking. They found that less than 50% of the schools that are classified as failing based on the US Code would be considered as failing when measured by student learning or educational impact. They also found that around 20% of the schools with satisfactory ratings were failing when alternative measures were applied.
Available School Choice Evidence and The Causes of Failure
Downey and his colleagues also reported that when testing alone is used to determine success or failure there are clear causes that emerge. Failing schools tend to serve racial minority students and poor students, often overlapping. The Downey study showed that when measured for learning rates, the socioeconomic differences either shrink or disappear completely. What happens is that the gap closes between high- and low-performing schools as opposed to when testing alone is used.
Available School Choice Evidence: A Final Word
The available school choice evidence points to factors other than learning and impact. By relying on standardized testing alone, a skewed result further disadvantages the already disadvantaged. To look more closely at the learning and the impact of learning on students, a one creates a sharper picture. This picture seems to argue against the idea that parents act as surrogates for their children regarding school choice.