State by state, it is hard to draw a line between school closures and learning loss, since some states that stayed closed longest fared best, and vice versa. Earlier research showed a clearer relationship between school closures and learning loss at the district level, but at a news conference announcing the latest N.A.E.P. report, the commissioner of the National Center for Education Statistics said, “There’s nothing in this data that tells us there is a measurable difference in the performance between states and districts based solely on how long schools were closed.”
In New York City, the nation’s largest school district, schools reopened in September 2020. There, average scores for reading fell by about a point for fourth graders and improved by about a point for eighth graders; in math, fourth-grade scores fell by nine points (statewide scores fell by 12) and eighth-grade scores fell by four points (statewide scores fell by six). In Los Angeles, the second-largest district, schools stayed closed through January 2021. There, average scores actually improved in fourth-grade reading, eighth-grade math and eighth-grade reading, where they improved by a robust nine points (to 257 from 248). Scores fell only in fourth-grade math (to 220 from 224).
In a vacuum, the pandemic declines look like bad news, if at a relatively small scale. But none of this happened in a vacuum. I’ve mentioned the million deaths not to fearmonger about how much higher those numbers might have been without school closures — the scale of that impact is, I believe, an open question — but just to point out the enormous and widespread human impact of the disease itself. And that impact was much larger than measured simply by mortality. More than 3.5 million Americans were hospitalized, according to one estimate, and probably at least as many suffered from long Covid. In the spring of 2020, the country’s unemployment rate exploded, jumping to nearly 15 percent from about 4 percent; for a brief period in April, six million new jobless claims were filed each week. In a single quarter, U.S. GDP fell by 9 percent. Murder rates grew by 30 percent; deadly car crashes spiked, too. Overdose deaths rose 30 percent in 2020 and 15 percent in 2021. According to some research, rates of depression tripled in the United States when the pandemic first hit. Some 600,000 teachers left the profession.
This is the world in which American students — most of them learning remotely for many months, many of them for close to a year, and some for longer — fell off by a handful of points, on their reading and math exams, compared with their prepandemic peers.
“The sudden onset of the pandemic has been the most catastrophic event in recent American history, making the expectation that there would not be something called ‘learning loss’ bizarre,” Keeanga-Yamahtta Taylor wrote recently in The New Yorker. “The idea that life would simply churn on in the same way it always has only underscores the extent to which there have been two distinct experiences of the pandemic,” she went on, emphasizing how much harder the pandemic was for the poor and marginalized to navigate, compared with those for whom its secondary effects were buffeted by wealth.
International comparisons offer another bit of context for test score declines. In England, schools closed in the spring of 2020, opening again in some places in early summer and across the country in the fall (with an Omicron interruption of about a month that winter of 2021). In retrospect, that would have been a plausible but relatively aggressive school reopening approach in the United States, where many schools stayed remote well into the 2020-2021 school year. It also resulted in a drop of six percentage points in proficiency scores, roughly comparable to the American experience. In other words, in England, with a close-to-optimal school reopening, they fared no better.
In the Netherlands, where schools were even less disrupted than in Britain, student performance fell by three percentage points — a bit better, but still below the standards set in prepandemic years. At the most extreme end of the spectrum, there is Sweden, which did not close schools at all, and which, some reporting has suggested, experienced no such declines. But the country also suspended its testing program, which means the data on which such claims might be based is pretty shaky.