Remote learning during the COVID-19 pandemic has exacerbated educational inequalities, resulting in long-term achievement gaps

What's this about?

People disagree about whether online school during COVID made school gaps worse.

They also ask if those gaps will last for many years.

What supporters say

• Students from poorer homes lost more learning than students from richer homes.
• Online learning worked better for children with good internet, computers, quiet rooms, and adult help.
• After schools reopened, test scores stayed below the levels experts expected before COVID.

What critics say

• Many students have started to catch up, so the gaps may not last forever.

How to read this

The number of points on each side does not show who is right; strong proof matters more.

The bottom line

The facts strongly show that poorer students lost more learning during the pandemic.

Online school likely made old school gaps wider, but we are not sure yet if it caused lasting harm.

The claim is that pandemic-era remote learning did more than interrupt schooling: it deepened existing education gaps and left some students behind even after schools reopened. The evidence supports much of that argument, but is less conclusive on whether remote learning alone caused permanent damage.

The case for

The strongest evidence is that learning losses were larger among disadvantaged students. A review of 42 studies across 15 countries found substantial declines in achievement during the pandemic, with students from lower-income backgrounds often suffering the biggest setbacks. Another major analysis also linked school closures to lower achievement and examined factors such as family income, students’ ages, subjects and the length of closures. 1

International agencies including the World Bank, UNESCO and UNICEF warned that children’s ability to learn at home depended heavily on their circumstances. Families differed in access to computers, reliable internet, quiet places to study and adults who could help. Research also found differences in how much contact students had with teachers and in the quality of online lessons. A study of 1,844 children in Ghana, for example, linked engagement in learning to household resources and caregiver support. 3

That creates a clear pathway by which remote learning could have widened inequality. When schools moved online, students’ access to effective teaching depended partly on what their households could provide, not simply on whether a lesson was officially available. In that sense, remote instruction may have turned existing differences in family resources into wider differences in learning.

Achievement also remained below where it was expected to be before the pandemic. U.S. NWEA data showed that scores in 2023–24 were still below projected pre-pandemic trends. The gap between pandemic-era and pre-COVID averages widened in most grades compared with the previous year—by about 36% in reading and 18% in mathematics. The Education Recovery Scorecard found that students continued to learn after schools reopened, but progress toward full recovery stalled in 2022–23, with large differences between districts and student groups. International testing likewise recorded broad falls in reading and mathematics between 2018 and 2022. 2 (see Figure 2)

The case against

The main weakness in the claim is that remote learning cannot be cleanly separated from the wider effects of the pandemic. School closures occurred alongside illness, financial stress, social isolation, disrupted routines, attendance problems and changes in how schools operated. Test scores show that achievement fell after the pandemic, but they do not prove that remote instruction was the sole cause—or that every resulting gap will last permanently.

Recovery evidence is particularly important. A review found that reopening schools, tutoring and some phone-based or in-person programs reduced learning losses. In Louisiana, district-level evidence showed full recovery in reading, even though other places continued to have serious gaps. These findings show that early losses could be reversed when schools reopened and support was focused on students who needed it.

The results therefore vary by location and group. Recovery depended on when schools reopened, the quality of teaching, students’ digital access and the size and quality of later interventions. The available research documents losses and unequal access to learning, but it does not establish one permanent outcome for all students.

The bottom line

The evidence strongly supports the core claim that pandemic-era remote learning was associated with major learning losses and that lower-resource students generally suffered more. It also offers meaningful evidence that achievement remained below pre-pandemic trends after reopening.

But the claim is less firmly established when it says remote learning created uniformly long-term achievement gaps. The pandemic combined online schooling with many other disruptions, making precise attribution difficult. Recovery in some places shows that initial, unequal losses can be real without becoming permanent. Overall, the evidence favors the claim with high confidence on unequal short- and medium-term harm, but only qualified confidence on the size and lasting nature of the effect specifically caused by remote learning.