During the COVID-19 pandemic, many schools moved to some form of online learning to keep students and teachers safe. But those measures took a significant toll on student achievement, according to a paper in the American Economic Review: Insights. 

Authors Dan Goldhaber, Thomas J. Kane, Andrew McEachin, Emily Morton, Tyler Patterson, and Douglas O. Staiger found that students who had more remote instruction during the pandemic improved their achievement from fall 2019 to fall 2021 much less than prior cohorts. 

Using data on over two million students across 49 states, the authors compared the achievement growth of individual students before and during the pandemic. The decline in growth was especially steep for students in high-poverty districts that were remote for much of the 2020–21 school year. Where schools returned to in-person learning, the losses were smaller, and gaps between high and low-poverty districts did not widen.

Figure 2 from the authors’ paper shows how student achievement growth between fall 2019 and 2021 declined in relation to the degree of remote instruction and poverty level of school districts.

Figure 2 from Goldhaber et al. (2023)

The x-axis indicates the percentage of the 2020–2021 school year that a school was in remote instruction. And the y-axis indicates the difference, in standard deviations, between student achievement growth from fall 2019 to fall 2021 and the predicted growth based on student growth from fall 2017 to fall 2019. The green triangles, red squares, and blue circles represent values for low-, middle-, and high-poverty school districts, respectively.

The chart shows that when the share of remote instruction was low, losses were similar for low-, middle-, and high-poverty schools—about 0.17 standard deviations. However, the gap between high- and low-poverty schools widened significantly as the percentage of time spent in remote learning rose. For schools that spent more than 50 percent of the year in remote instruction, students in high-poverty schools lost roughly 0.44 standard deviations relative to prepandemic growth, while students in low-poverty schools lost 0.26 standard deviations.

The authors calculate that high-poverty districts that went remote in 2020–2021 would need to spend nearly all of their aid from the American Rescue Plan to help students recover from pandemic-related learning losses.