Technocratic Education Reform Is Dead. Long Live School Choice.

A recent study showing that test-score-driven school turnaround models fail to yield long-term improvements may be the final nail in the coffin for technocratic education reform strategies.

Going forward, education reform will thrive by empowering parents with choices, rather than by ordering them to obey self-appointed experts.

Technocratic reform strategies typically rely upon experts using student test scores to identify what schools and practices “work.” The technocrats then attempt to replicate what “works” by steering families into approved schools using favored techniques and to shut down what does not “work” by forbidding families from selecting disfavored schools and practices.

According to the technocrats, education can be improved if we just “follow the science.”

This all sounds great until you remember how those pushing us to follow the science during the COVID-19 pandemic produced enormous harm.

What if short-term changes in test scores are not good proxies for producing desired outcomes for students over the long term? What if the researchers examining those test scores are using flawed analytical methods that are prone to manipulation? What if the policy conclusions researchers draw are not even supported by their flawed methods?

As every central planner eventually learns, the world is too varied and complicated for even the best-intentioned researchers armed with the best research methods to know how everything is supposed to be run.

The technocrat approach to education reform reached its zenith during the Obama administration and its Race to the Top initiative.

Under that program, states and districts were directed to use billions of federal dollars to turn around the lowest-performing schools by adopting proven reform models.

The pretentiously titled What Works Clearinghouse of the U.S. Department of Education put together a report that was supposed to identify school turnaround models backed by rigorous evidence.

I was on the panel that prepared that report. The problem was that we could not find any school turnaround models supported by rigorous evidence.

I suggested that we write a very short report saying that science had nothing to offer about how to turn around failing schools, but I was told that was not our mandate and that we had to recommend turnaround models even if they had little or no evidence demonstrating their effectiveness.

So, despite my concerns, that is what we did.

Undeterred by the lack of evidence, some states jumped into school turnaround reforms with both feet. Chief among them was Tennessee.

Tennessee had been a pioneer in developing “value-added” student testing that focused on the gains students made from year to year. Focusing on gains was supposed to remove the influence of student background on achievement and isolate the quality of schools in improving performance.

In fact, focusing on test score gains does not eliminate the influence of student disadvantages, since those disadvantages also influence the rate at which students can make progress. Nonetheless, with this tool, Tennessee officials had confidence they could identify lower-performing schools and evaluate models for turning them around.

With a grant from Obama’s Race to the Top, Tennessee fully embraced the technocratic approach, seeking to close some low-performing schools, and to reassign other low-performers to new management equipped with school turnaround models.

Using the Tennessee Value-Added Assessment System, the state identified more than 200 schools as low-performers between 2012 and 2016 that it said were “priority schools” to be turned around.

Some of those schools were removed from the control of their local school district and reassigned to a newly created Achievement School District (ASD) that could provide turnaround models to low-performing schools statewide. Other priority schools were left under the control of their local districts but were designated as iZone schools to receive turnaround models. Still other priority schools did not receive turnaround models.

Schools that received turnaround models had their operation transferred to a charter management organization selected by state officials. The turnaround schools were not charters, but the state still contracted with these charter management organizations because they had a track record of running schools that had produced impressive test score gains.

While charter schools are chosen by parents, all of the students who attended the ASD and iZone schools were assigned to them based on where they lived and not because their parents chose those schools. The operators of these turnaround schools were chosen by education bureaucrats, not by parents.

The new study led by Lam Pham at North Carolina State University compared the long-term outcomes of students who lived in the catchment zones assigned to the ASD or iZone turnaround programs for middle schools relative to those who lived in areas with priority schools that did not receive a turnaround model.

Some earlier studies had found that students attending turnaround schools did better while they were in those schools, which those schools might be able to produce by narrowing their instruction to tested subjects and items even if that came at the expense of longer-term outcomes.

Pham’s study is exceptional in that it tracks students from turnaround schools for several years after they leave those schools to see if there are any enduring benefits. In addition, by looking at multiple outcomes, including end-of-course exams, college admissions tests, attendance, behavioral infractions, and graduation rates, his study was better able to capture the broad goals of education and avoid the narrow focus on student test scores that might be subject to manipulation.

Pham and his colleagues examine test scores for more than 70,000 students with multiple specifications of rigorous research designs and concluded: “We find little evidence to support improved long-run student outcomes—mostly null effects that are nearly zero in magnitude.”

In fact, if anything, the estimated effects tend to be negative, but most are not statistically significant. The smartest experts, armed with value-added test scores and turnaround models, were completely unable to improve outcomes for students assigned to low-performing schools in Tennessee.

The results from Tennessee also stand in sharp contrast to those produced by schools in New Orleans following Hurricane Katrina in 2005.

New Orleans, like Tennessee, had adopted a number of technocratic reforms, including using test scores to try to close lower-performing schools and force students to transfer to schools with higher test scores. But New Orleans also fully embraced school choice, turning virtually every school in the city into a charter school so that every family had to select their school rather than be assigned to one based on where they lived.

In a study similar to the one conducted by Pham, Doug Harris at Tulane University found significant long-term benefits to students in New Orleans as a result of these post-Katrina reforms. But whether those gains were produced by the technocratic aspects of New Orleans’ reforms or by its complete adoption of school choice was ambiguous.

Harris preferred to attribute the long-term gains made by students in New Orleans to technocratic measures.

The new results from Tennessee suggest Harris was mistaken. If technocratic reforms failed to yield long-term benefits in Tennessee, they were unlikely to be the explanation for gains in New Orleans.

Broad adoption of school choice likely produced the enduring benefits we saw in New Orleans and that are also becoming evident in Arizona, Florida, and other places that are pursuing school choice at scale.

Technocratic education reform has run its course. The evidence does not support it. The public and their elected representatives have grown wary of experts armed with test scores trying to tell us how to educate our children.

The future of education reform belongs to school choice and that future looks bright.