Would more emphasis on high-stakes testing help address America’s learning woes?

Welcome to Chalkbeat Ideas. I’m glad you’re here and I’ve appreciated the support from so many of you as I’ve launched this new section. Chalkbeat counts on reader donations and I hope you’ll consider making one today. Our journalism is free to read, but of course not free to produce. Gifts will be matched through this year.

A striking wave of nostalgia is rippling through the education world these days. 

A number of commentators and advocates have harkened back to the early aughts when test scores were rising and No Child Left Behind was the law of the land.

“This was an era when elected officials from both political parties came together around a reform agenda,” Harvard University professor Marty West said at a U.S. Senate hearing earlier this year. “That approach had plenty of critics and, as implemented, some real flaws, but it produced results.”

Federally imposed high-stakes testing kept the heat on schools, but after the repeal of No Child Left Behind, we’ve let them off the hook, say these accountability hawks. This is one reason, they argue, why test scores were stagnant before the pandemic and have fallen since. 

Some policymakers are beginning to respond. More governors touted accountability in education in their state of the state speeches this year. Oregon recently enacted stricter oversight of schools after years of disappointing test scores.

It’s because of this new fervor that I wanted to take a deep dive into the research evidence on school accountability. Should we really have a case of No Child Left Behind nostalgia? Would putting more pressure on schools improve learning?

The short answer is perhaps — with several asterisks. There’s a good case to be made that NCLB produced real learning gains in math and could again. However, the law also led to numerous unintended consequences and quickly proved unsustainable. 

No Child Left Behind, enacted in 2002, was a dramatic shift in American education. Schools across the country were required to give tests every year in reading and math in grades three through eight. Those scores could then be used to sanction schools that didn’t meet annual progress benchmarks. Although this didn’t happen often, schools with persistently low scores could be taken over by the state or shuttered altogether. The law set an audacious goal of reaching nearly 100% proficiency on these tests by 2014. 

Researchers Thomas Dee and Brian Jacob developed a clever strategy to study the impact of NCLB. Since some states already had accountability laws in place, Jacob and Dee reasoned that if the law was working, it would especially make a difference in the states that had to make bigger changes.

That’s exactly what they found in math. The share of students who reached basic proficiency in fourth and eighth grade math jumped by several percentage points in these states over the first five years of the law. In reading, however, there were no clear effects. A subsequent study, comparing public schools to private schools, also found gains from NCLB in math, as well as “possible but distinctly smaller” improvements in reading. 

These studies have an important advantage: They rely on the National Assessment of Educational Progress or NAEP, a respected test that does not come with any stakes for students or schools. 

One caveat to keep in mind: Dee and Jacob also found that states spent additional money on their schools in response to NCLB. This means that score gains likely came from a combination of funding and accountability. 

Although there are exceptions, other research has generally found that accountability pressure on schools boosts test scores, particularly in math. Reading gains appear less consistently. A handful of studies have found longer-term benefits from accountability, including higher high school graduation rates, college attendance, and adult income.

“The body of evidence is overwhelmingly consistent with the idea that test-based accountability meaningfully improves student outcomes,” says Morgan Polikoff, an education professor at the University of Southern California.

Jennifer Jennings, a Princeton University professor who has researched NCLB, notes that this research came from when schools shifted from limited accountability to quite a lot. Today, she says, is more of a “medium accountability” environment. 

The gains from prior research “are not reproducible in the reality we have now,” she says. “It’s certainly not the case that people aren’t worried about tests.”

The irony of the current nostalgia is that when No Child Left Behind was in place, much of the country was growing angry at how the law was changing schools. By 2007, one of the law’s supporters said it “may be the most tainted brand in America."

Teachers and parents increasingly complained that schools focused too much on math and reading, while art, music, field trips, science, and social studies were sometimes sidelined. Some schools emphasized test-taking strategies or moved their best teachers to tested grades. High-profile cheating scandals erupted. Researchers said the law’s focus on absolute proficiency was a misleading way to judge schools. Families who liked their local public school recoiled when it was labeled a failure. The reality that 100% proficiency was unrealistic began to set in.

“While accountability appealed in the abstract, its allure curdled pretty quickly once voters saw it in practice,” wrote Rick Hess of the American Enterprise Institute.

There was also emerging evidence that after a burst of initial improvement, scores didn’t respond much to further pressure, a phenomenon some dubbed the “accountability plateau.”

It was in this context that policymakers pivoted to holding individual teachers, rather than schools, accountable for test scores. This strategy, embraced by the Obama administration, also drew a fierce backlash — and did little to raise achievement nationally, according to a major study.

In 2015, a bipartisan coalition in Congress approved the Every Student Succeeds Act, which scaled back federal pressure, even as regular testing continued. The main provision of the law says states only have to identify 5% of their worst Title I schools, whereas NCLB put virtually every school on notice with specific threatened sanctions. Although some policy wonks objected at the time, there was no grassroots opposition to ending NCLB. 

Accountability hawks now say this shift in law bears some of the blame for recent declines in learning, although there’s relatively little research on this point specifically.

So here’s where we stand: There’s a solid argument that more school-level accountability could help improve math scores. The evidence is murkier when it comes to reading, where scores have been particularly resistant to improvements of late.

Almost everyone acknowledges that prior efforts had a number of problems, though. For those who want to revive such policies, now would be a good time to shift from demanding more accountability generally to thinking about how to build better systems that reflect what we’ve learned from the success and failure of NCLB.

Have thoughts about designing accountability systems for schools? Share them with me at [email protected].