Teenagers in the United States continue to perform significantly worse in math compared to before the pandemic, but they have held relatively steady in reading and science and lost less ground across subjects than most of their international peers, according to results released this week from an international assessment of teenagers鈥� academic performance.
U.S. 15-year-olds performed 13 points lower than students their age did on the last round of the Program for International Student Assessment, or PISA, in 2018. Their score of 465 is on par with the global average, but 18 points lower than it was for the first PISA in 2003.
Average scores on PISA range from 0 to 1,000. A difference of 25 points represents roughly a year of learning.
The average U.S. student scored statistically flat in reading (504) and science (499) compared to 2018, which is above the international average.
As with scores on the National Assessment of Educational Progress, the gaps between the highest- and lowest-performing students on the PISA in reading and science (though not in math) have widened over time.
The Organization for Economic Development and Cooperation administers PISA to 15-year-olds every three years in reading, math, and science, with a special focus on a different subject each time. The pandemic delayed the 2021 test to 2022, when it focused on math. Some 620,000 students in 38 mostly industrialized countries and a total of 81 education systems, including four in China, participated in PISA in 2022.
Of all education systems in OECD, The United States ranked 6th in reading, 10th in science, and 26th in math.
鈥淭hese results are another piece of evidence showing the crisis in mathematics achievement, only now can we see that it is a global concern,鈥� said Peggy Carr, commissioner of the National Center for Education Statistics, which administers the PISA in the United States. 鈥淵ou can consider that as good news, perhaps, that we have rankings that are higher than they were before, but those rankings are due to others suffering more because of this pandemic than the United States. So I think it鈥檚 relative.鈥�
High and low achievers diverge
In the United States, fewer than 1 in 10 students scored at an advanced level, considered the two highest of six achievement levels, in math, which is roughly around the average for OECD countries. By contrast more than 1 in 5 in Japan, Korea, and three Chinese education systems tested performed at the highest levels. Singapore led countries with 4 in 10 students performing at the highest levels in math.
More than 1 in 3 U.S. students did not meet basic achievement levels in PISA in math, also on par with the OECD average. This means they could not, for example, use multiple strategies, including ratios and one-variable simulations, to solve problems.
鈥淢ath is a very school-dependent subject, at least in the United States, and obviously COVID has been a wake-up call for many,鈥� said Bob Hughes, director of the Bill and Melinda Gates Foundation鈥檚 K-12 education in the United States program. But, he added, 鈥渕ath has been stagnant for years, and so these declines shouldn鈥檛 be surprising. They should actually just be a call to action and give us a sense of urgency about the types of math reform that need to be in place to be effective.鈥�
Compared to the global average, the United States had a larger share of its students score at the highest levels in reading (14 percent) and science (11 percent), and a smaller share of low performers, 1 in 5 in reading and slightly more than that in science.
U.S. students who attended schools serving three-quarters or more students in poverty scored 50 scale-score points lower than the U.S. average in math, 48 scale points lower in reading, and 53 scale points lower in science. That means teenagers in the highest poverty schools in the United States performed on average below basic proficiency on PISA, and about two years behind the U.S. average.
A senior Education Department official pointed to disparities in access to higher-level math courses, particularly for students of color, as a reason for the dramatic gaps between the low- and high-income schools. New civil rights data show half of U.S. high schools don鈥檛 offer calculus or computer science.
While U.S. schools remained closed longer than average compared to other countries during the pandemic, students鈥� performance declined less than OECD countries that reopened sooner, such as Finland. Carr noted that differences in school reopening times explained only about 11 percent of the variation in scores among different countries.
U.S. Secretary of Education Miguel Cardona credited $350 billion in federal emergency spending during the pandemic for preventing steeper declines in the country. Data from the School Pulse Panel, an ongoing federal survey of schools鈥� instruction since the pandemic, show that more than half of districts have used American Rescue Plan money for tutoring and one-on-one interventions for students who are struggling academically, including in math. A senior Education Department official noted that states like Washington have also used relief money to adopt new math curriculum.
鈥淲hile we saw even more dramatic declines in the math scores in other countries, our math scores are still declining and they remain stubbornly low,鈥� Cardona said. 鈥淲e have what it takes to be number one in PISA, so today we need to fight complacency with the same urgency with which we fought COVID.鈥�
Cardona announced the Education Department would award $90 million in grants on Tuesday to identify and adopt effective instruction in science, technology, engineering, and math areas.
