Pay attention to whatever Yong Zhao writes. He is among the very top tier of educational thinkers in the world. I always learn when I read his work.

This post warns parents, teachers, and policy makers to beware the “learning loss” rhetoric. It is a trap, he says.

He writes:


A dangerous trap exists for educators and education policy makers: the learning loss. This trap comes with a large amount of data and with sophisticated projection methods. It presents a stunningly grim picture for education and it invites educators and policy makers to make wrong decisions and invest in wrong things. The article identifies a number of undesirable outcomes that their concerns could lead to. It also suggests several productive actions when the pandemic is controlled and schools reopen.

The trap is the so-called learning losses during the Covid-19 pandemic. A number of organizations and individuals have put out various estimates about what students have lost due to school closures and remote learning during the pandemic. For example, the global consulting firm McKinsey produced two reports about these learning losses. As late as December 8, 2020, McKinsey said, “Students, on average, started school about three months behind where we would expect them to be in mathematics” and “Students of color were about three to five months behind in learning; white students were about one to three months behind” (Dorn et al. 2020). The Center for Research on Educational Outcomes (CREDO 2020) at Stanford University issued a press release stating that “the average estimates of how much students lost in the Spring of 2020 ranged from 57 to 183 days of learning in Reading and from 136 to 232 days of learning in Math” (para. 2). Other organizations, such as the assessment company NWEA (Kuhfeld and Tarasawa 2020) and the Annenberg Institute at Brown University (Santibanez and Guarino 2020), have also published reports about learning losses. The Organization for Economic Cooperation and Development (OECD) published a projection for the economic losses due to learning losses as $14 trillion over the next 80 years (Hanushek and Woessmann 2020).

These estimates have caught the attention of policy makers and educators. Governments, school leaders, and teachers are all concerned about the learning loss students may experience due to the Covid-19 pandemic. After all, schools have been seriously disrupted, as have students and their families. The pandemic has, in more ways than one, significantly affected learning and school operations. It seems only natural to want to know the extent of the learning loss students have experienced and then take actions to hopefully make up for the losses.

Possible mistakes

This is wherein the trap lies. There is nothing wrong with making estimates about learning losses, but the possible actions these projections can induce are worrisome because they can, at best, waste resources and, at worst, lead post-pandemic education in the wrong direction. The concerns of educators and policy makers are to be expected, but these policy makers could end up investing in unproductive educational efforts. Below are a number of undesirable outcomes that their concerns could lead to.

Governments may decide to launch standardized assessments to track students’ learning losses. It is possible that educational policy makers may be so interested in learning the extent of loss experienced by students that they will use standardized testing to assess all students. The desire to know the overall extent of loss and what achievement gaps may exist between different groups of students is completely understandable, but standardized testing can be the worst way to collect such data for two major reasons.

First, any standardized testing given to all students will have a typically limited scope, with a focus on math and reading. In other words, what will be measured is not the entirety of students’ learning but a small piece of their overall education. Even assuming that the assessments are highly accurate (which they are not), they would miss other equally and perhaps more important aspects of learning, such as confidence, self-determination, creativity, entrepreneurial thinking, and other subjects.

Education has many desirable outcomes (Zhao 2017, 2018b). These outcomes can be short term or long term, cognitive and non-cognitive, and instructional and educational. Short-term, cognitive, and instructional outcomes do not necessarily translate directly into long-term, non-cognitive, and educational outcomes. For example, test scores have often been found to have a negative correlation with students’ confidence and well-being (Loveless 2006; OECD 2019; Zhao 2018b). Test scores have also been found to have a negative correlation with economic development and entrepreneurial confidence and activities across (Baker 2007; Tienken 2008; Zhao 2012). Test scores do not predict the future of an individual’s success very well, and non-cognitive skills may play a bigger role than cognitive skills play (Brunello and Schlotter 2010; Levin 2012). Some assessments show successes that are only productive in the short term, while failures may actually be more productive in the long term (Dean and Kuhn 2007; Kapur 2014, 2016).

That’s the beginning. Read it all.