Skip Navigation

On Learning and Loss in a Time of Covid

Mark Schneider, Director of IES | September 28, 2020

I begin this blog with a short report from my own personal encounter with schooling in the COVID era—but this will lead to a broader set of issues affecting all of us as we enter a new school year structured by responses to the pandemic.

My older daughter lives in Brookline, Massachusetts. She is a single mom, a high-powered, busy attorney, with two sons, one 12, the other 10. The last time I visited them was in early March and I flew back to DC just a few days shy of when the world shut down. She refused to let me visit for the past six months, because she feared for my safety. But, lately, she realized that the risks of COVID were beginning to pale compared to the psychological costs of not seeing her and her kids. So, after appropriate quarantining, I flew for the first time in six months and spent several days witnessing firsthand how "school" is being conducted via video conference.

Every child is different and there is great variation across school districts in the quality of the online education delivered. But here I am in Brookline, a well-funded school system (indeed—at around $20,000 per pupil annual expenditures—very well-funded), watching two boys trying to learn in a (to use a highly technical term) "weird" learning environment.

The physical and attentional burdens of coordinating so many different career, education, and life demands are enough to drive anyone crazy. I am sitting in the kitchen writing this. Each of the boys are in their own bedrooms, while my daughter has commandeered the dining room as her office. The demand on her internet is extreme, accompanied by frequent and unexplained disconnects from the school's platform.

There's a 10-minute break after every lesson. Sometimes the breaks don't align, which means 20 minutes of free-roaming children. Sometimes the breaks do align, which means quick fights over video game access. The demand for snacks is ever present. Both boys are suffering from the lack of personal contact with their teachers and their friends. What a mess!

This personal saga leads me to the more professional (indeed societal) issue that motivated me to write this blog. I want to call your attention to three disturbing reports that have crossed my desk recently, highlighting the damage to learning and the economy that school closures have inflicted on all of us.

The first report is from the Annenberg Center at Brown University. Published in June and entitled "Projecting the potential impacts of COVID-19 school closures on academic achievement," the authors, extrapolating from analyses of learning loss due to absenteeism and summer breaks, estimate that students are likely to have lost about one-third of the learning gains of a typical school year in reading as a result of school shutdowns. The losses in math are expected to be even greater, with somewhere between half and two-thirds of the typical gains lost. Not surprisingly, the losses are not evenly distributed, but concentrated among the lowest-performing students. The authors anticipated a return to school this fall, so these estimates are likely to be low, given that returning to school now looks radically different. It's unlikely that the losses they estimated through June will abate during this year—I fear just the opposite.

Two related papers highlight the economic costs associated with these learning losses. On June 1, McKinsey Insights published "COVID-19 and student learning in the United States: The hurt could last a lifetime." In a scenario in which large-scale, in-person instruction resumes in January 2021, McKinsey estimates that, for the average K–12 student in the U.S., losses could total between $61,000 and $82,000 in lifetime earnings in constant 2020 dollars—more than a year of full-time work for the average American. The impact is more severe for Black and Hispanic students. McKinsey translates these individual losses into societal ones: an annual loss of $110 billion!

The third paper is an OECD report by Eric Hanushek and Ludger Woessmann. Published in September, "The Economic Impacts of Learning Losses" looks at the economic catastrophe stemming from Covid-related learning losses both in the U.S. and across G20 countries. Hanushek and Woessmann argue that the learning loss experienced by students in grades 1–12 affected by the global shutdown of schools will lead to lower long-term growth—an average of around 1.5 percent lower annual GDP for the remainder of the century.

This estimate is based on students losing a third of a year of schooling. The Annenberg report suggests that losses could be higher, somewhere around half to two-thirds of a year of lost learning (if not more). According to Hanushek and Woessmann, for the U.S., a loss of a third of a year of learning will lead to a $14 trillion loss in GDP—and two-thirds lost would produce a loss in GDP of over $27 trillion—through the rest of the century (these are both present value estimates). These scenarios fit within the Annenberg estimates of learning loss—but remember, those were released before the start of this school year. With no end to the pandemic and return of in-person schooling for large numbers of students in sight, learning losses (and the corresponding loss in wealth) could be even greater.

The shutdown of schools and resulting reliance on online education caught many of us, including IES, flat footed. IES is trying to figure out how to fund more high-quality research to identify and support ways in which students can learn when they are not able to attend school in person, including online and hybrid learning models. One place where we are moving fastest is our support for research and development pertaining to digital learning platforms (DLPs).

We are working on two initiatives that, hopefully, will accelerate the accumulation of knowledge about DLPs. First, we expect to announce a prize competition to identify best in class DLPs. Second, we are working on an RFA to bring together networks of researchers, developers, and platform owners to support R&D of DLPs.

As many of you know from previous blogs, the impetus for this work, especially the prize competition, is our desire to support ways of increasing the rate at which experiments, and subsequent replications, can be delivered. That is still a driving force, but I think the stakes have grown substantially since these efforts were first conceived. To be sure, supporting DLPs and identifying best in class platforms is important as a way of delivering high-quality experiments and replications faster and cheaper. These innovations in research methods matter, but these investments should also reveal opportunities to improve learning outcomes via online education—a critical need that has only grown with the pandemic.

We will continue to support innovations in online education through our annual research competitions. But we will also continue to find (and fund) special efforts to accelerate the accumulation of knowledge about how best to identify and support what works best for whom in online education.

As always, please feel free to contact me at