The Learning Crisis: Three Years after COVID-19

The Learning Crisis: Three Years after COVID-19 | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Social Sciences - Article The Learning Crisis: Three Years after COVID-19 Harry Patrinos, Tomasz Gajderowicz, Maciej Jakubowski, Alec Kennedy, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6566438/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract The COVID-19 pandemic triggered widespread school closures affecting over one billion children worldwide, with potentially lasting effects on learning. Leveraging data from the 2023 Trends in International Mathematics and Science Study (TIMSS), covering 71 education systems and spanning two decades, we quantify the global impact of pandemic-related disruptions on student achievement in mathematics and science. Using mixed-effects models, we estimate deviations from pre-pandemic trends, adjusting for demographic and school-level factors. We find that average achievement declined by 0.11 standard deviations globally, with longer school closures associated with larger losses. Disadvantaged groups—including low performers, girls, and linguistic minorities—experienced greater setbacks, reaching up to 0.23 standard deviations. These findings reveal a persistent global learning crisis and highlight the urgent need for targeted policy responses to address educational inequalities exacerbated by the pandemic. Scientific community and society/Social sciences/Economics Scientific community and society/Social sciences/Education Figures Figure 1 Introduction Coronavirus (COVID-19) is an infectious disease caused by the SARS-CoV-2 virus. Most people infected with the virus will experience mild to moderate respiratory illness and recover without requiring special treatment. However, some will become seriously ill and require medical attention (WHO 2024 ). These effects were particularly concerning in the early stages of the pandemic and restrictions were established with the intention of reducing the spread of COVID-19, including lockdowns, stay at home orders, and school closures. One of the arguments for closing schools was that school children at that age might be drivers of the spread within countries and thereby also endangering the older part of the population. However, there is not much evidence that opening schools significantly increased infection rates (Vlachos et al. 2021 ). During the COVID-19 crisis, one billion children were affected by school closures. Many countries were ill-prepared for the extensive school closures, but the few with prior experience of remote learning, trained teachers, appropriate technology, and engaged learners provided continuity during the crisis. Globally, schools were closed for an average of 5.5 months (22 weeks) since the onset of the pandemic, equivalent to two-thirds of an academic year, when localized school closures are considered (UNESCO 2023 ). The duration varies by region, from just one month in Oceania, to 2.5 months (10 weeks) in Europe, to as many as 5 months (20 weeks) in Latin America and the Caribbean. Early on, the school closures were expected to contribute to what was already described as a learning crisis (Angrist et al. 2021 ). Distance learning during the school closures does not seem to have helped very much, with evidence that it increased dropout risk and lowered test scores (Lichand et al. 2022 a; Haelermans et al. 2022 ); only the duration of school closures led to variations. Most studies observe increases in inequality where certain demographics of students experienced learning losses that were more significant than others. These learning losses could translate to earnings losses and could cost this generation of students trillions of dollars (Psacharopoulos et al. 2021 ). This study explores the global extent of learning loss using successive rounds of TIMSS (Trends in International Mathematics and Science Study) – an international assessment of student achievement in mathematics and science at the fourth and eighth grades, conducted by the International Association for the Evaluation of Educational Achievement (IEA) – up to the latest, which is 2023. By incorporating school closure duration, we examine how the length and nature of these disruptions affected student outcomes. This is the largest and most comprehensive investigation of learning loss over time, offering a global perspective that includes both pre- and during-pandemic data. We investigate the variation in impacts based on school closure duration and demographic factors, providing a comprehensive view of global trends in educational recovery. Our findings offer valuable insights into the persistence of learning losses in the years following school closures. This study reveals a global decline of, on average, 0.11 standard deviations, with more severe losses linked to longer closures, particularly among girls and linguistic minorities. Low performers lost up to 0.23 standard deviations, or more than half a year’s worth of learning. We make several contributions to the existing literature. First, we improve the precision of estimates regarding the impact of school closures on learning outcomes. Previous reviews of national studies show large learning losses, with estimates ranging from one-third to half a year's worth of learning (Betthäuser et al. 2023 ; Hammerstein et al. 2021 ; Patrinos et al. 2023). Reviews of studies produce slightly lower losses than the analyses using the international student assessments. This could be because the reviews have limited samples. For example, Betthauser et al. (2023) review 42 studies and lacks evidence on lower income countries. The international assessments have many more countries and are more representative of countries at different income levels. International student assessments, such as TIMSS, offer more comprehensive and representative data across countries with varying income levels, allowing for better cross-country comparisons. The average learning losses identified through large, comparable, international standardized assessments are 0.19 of a standard deviation, or a half years’ worth of learning (Crato and Patrinos 2025 ). TIMSS 2023 offers data up to Spring 2023, providing the most recent measures of student achievement, as prior assessments like PIRLS and PISA were conducted before or during the pandemic. Another key contribution of this study is its inclusion of evidence on learning outcomes from both before and during school closures for most countries, with a longer duration of data that covers the entire period of closures for all countries. This allows for a more detailed understanding of the lingering impact of school closures, as prior research has shown long-lasting learning losses (Bollyky et al. 2023 ; Gambi and De Witte 2024 ), although some countries experienced quicker recovery (Miller et al. 2024 ). The study also highlights the disparities in recovery across countries, with ongoing challenges for many students, particularly those from disadvantaged backgrounds, in regaining lost learning. Finally, this paper contributes to the growing body of literature on the effects of pandemics and other disruptions to children’s schooling (Belot and Webbink 2010 ; Ichino and Winter-Ebner 2004 ; Jaume and Willén 2019 ). Previous research on past pandemics, such as the 1918 influenza (Almond 2006 ) and the 2013–2016 Ebola outbreak (Smith 2021 ), has shown that school closures result in long-term negative effects on educational attainment and earnings. By adding to this literature, our study contributed to our understanding of the lasting (side) effect of school closures on global education, emphasizing the need for targeted recovery efforts to mitigate these losses and address the widening educational inequalities exacerbated by the pandemic. Methods We utilize data from the Trends in International Mathematics and Science Study (TIMSS), an international large-scale, repeated cross-sectional study that randomly samples fourth- and eighth-grade students to assess their achievement in mathematics and science. The curriculum-based assessments measure the knowledge students have accumulated over four and eight years of schooling, respectively. Conducted in four-year cycles, the 2023 cycle marks the first administration following the onset of the COVID-19 pandemic. Drawing on data spanning 20 years, starting from 1999, this study examines long-term trends in student performance, allowing us to evaluate whether the 2023 results diverge from pre-pandemic patterns. Altogether, we use data from more than 2 million students across all TIMSS cycles, including 71 educational systems (58 for Grade 4 and 44 for Grade 8) that participated in TIMSS 2023, as well as data from earlier cycles (see Annex A for summary statistics by country). We only excluded data from 1995 as it relied on a different psychometric model for scaling achievement, so the scores may not be fully comparable to those from subsequent years. The achievement scores were standardized separately for the two outcomes and populations in the initial TIMSS administration, using a scale with an international mean of 500 and a standard deviation of 100. They are comparable over time. TIMSS not only captures student achievement but also contextual information and characteristics of students through background questionnaires. We use this information as control variables to explain differences in achievement scores while also capturing changes to country demographics or samples over time. Specifically, we use four pieces of information that were consistently captured across all cycles of TIMSS from the student background questionnaire: student age, sex, and how often the test language was spoken at home. Additionally, data on school closure durations are obtained from UIS (UNESCO’s Institute for Statistics ( https://covid19.uis.unesco.org/global-monitoring-school-closures-covid19/ ), detailing the number of weeks of full and partial closures in each country (see Annex A). Schools were considered fully closed in cases where government mandates required schools to be closed, affecting most or all students. In contrast, schools were considered partially open when schools were closed only in certain regions or for some grade levels. This also captured schools that were only partially open to in-person instruction (e.g., hybrid learning). This data allows us to assess how learning losses varied by the duration of school closures. To capture changes following COVID-19, it is essential to take previous long-term trends into account. Our empirical strategy involves the estimation of deviations of the most recent 2023 result from long-term linear trends in average student mathematics and science achievement for each participating school system (see Jakubowski et al. 2025 ; Kennedy and Strietholt 2023 ). We model a separate linear trend for each country and include country-level fixed effects to control for unobserved time-invariant country characteristics. This is formalized in the following regression model: $$\:{Y}_{ijk}={{\Sigma\:}}_{k=1}^{n}{\alpha\:}_{k}+{{\Sigma\:}}_{k=1}^{n}{\beta\:}_{k}*time+\tau\:{D}_{2023}+\gamma\:{X}_{ijk}+{\epsilon\:}_{ijk}$$ 1 where \(\:{Y}_{ijk}\) represents the achievement of student i at school j in country k , with n being the number of countries in the sample being analyzed. The model is estimated on repeated cross-sections, with \(\:{\beta\:}_{k}\) capturing the slope of country-specific linear trends in student achievement across TIMSS cycles. \(\:{D}_{2023}\) is an indicator variable that is equal to one for the data collected during the 2023 cycle (after the onset of the pandemic) and is zero for all other cycles; \(\:\tau\:\) is our parameter of interest capturing the deviation from country-specific trends occurring after the onset of the pandemic after controlling for time-invariant country effects ( \(\:{\alpha\:}_{k}\) ) and student background characteristics ( \(\:{X}_{ijk}\) ). The model is estimated separately for both TIMSS subjects (i.e., mathematics and science) and both populations (i.e., grade 4 and grade 8). We next incorporate measures of national school closure policy duration to understand how our estimates of learning loss vary by the length of school closures. This is done through the introduction of an interaction term represented by the regression equation below: $$\:{Y}_{ijk}={{\Sigma\:}}_{k=1}^{n}{\alpha\:}_{k}+{{\Sigma\:}}_{k=1}^{n}{\beta\:}_{k}*time+\tau\:{D}_{2023}+\pi\:{D}_{2023}*week{s}_{k}+\gamma\:{X}_{ijk}+{\epsilon\:}_{ijk}$$ 2 with the model specified as it was in Eq. 1 , with an additional interaction term between \(\:{D}_{2023}\) and \(\:week{s}_{k}\) , the number of weeks that schools were closed in country k ; \(\:\tau\:\) can be interpreted as the deviation from the trend for a country that has no weeks of school closure; and \(\:\pi\:\) represents how those deviations vary by additional week of school closures for a country. The model is expanded further to test for differences in the impact of the pandemic and school closure policies by sex and home language. Each variable is dichotomized to provide three different contrasts: girls versus boys, and those who never or only sometimes speak the test language at home versus those who speak it most of the time. These contrasts are made by adding additional interaction terms between the individual student characteristics and both the D 2023 and D 2023 * weeks k from the model. Additionally, individual characteristics interact with the time trend variable, enabling trends to be estimated separately for each student group. Finally, we investigate the heterogeneity in the impact of school closures by achievement level. To check how learning losses vary among students of different proficiency, we estimate the main models using quantile RIF regressions, fitting models of unconditional of TIMSS achievement (Rios-Avila 2020 ). Data from TIMSS is collected using a complex stratified sampling scheme. To account for this, standard errors are calculated using a jackknife repeated replication (JRR) technique that accounts for sampling variance arising from the stratified class-based sampling design (Martin et al. 2020 ). Furthermore, TIMSS uses a rotated booklet design and employs the use of plausible values methodology to estimate students’ achievement (Mislevy et al. 1992 ). Therefore, all results presented are based on estimation accounting for the variation across the five plausible values for each subject scale (Rubin 2004 ). Findings The results highlight significant learning losses due to school closures, with variations across subjects, grade levels, and demographic factors such as sex and home language. For both Grade 4 and Grade 8 students, performance in mathematics and science fell significantly below expectations based on the linear trends observed in prior TIMSS cycles. Grade 8 students experienced greater declines across both subjects compared to Grade 4 students. We examine the impact of school closures on student learning in mathematics and science, analyzing data from 78 countries for Grade 4 students and 74 countries for Grade 8 students. Learning loss estimates Table 1 presents the estimated deviations from long-term trends by subject and grade. In Grade 4, the departure from trend was smaller for science than for mathematics. For Grade 8, the declines were larger in mathematics and science. To make the magnitudes of the linear trend departures comparable, we standardize them by the average within-country SD for each grade level and subject (SD Grade4,math =84, SD Grade4,science =85, SD Grade8,math =90, SD Grade4,science =91). The corresponding effect sizes (Cohen’s d) for Grade 4 were 0.11 for mathematics and 0.06 for science, while for Grade 8, they were slightly larger at 0.12 for mathematics and 0.14 for science. These results demonstrate significant negative effects on learning outcomes across all subjects and grade levels. Learning loss by school closure duration Table 2 presents how the deviations from the linear trend vary by duration of national school closures as captured in the UIS database considering both full and partial school closures. Note that some countries had to be omitted from the results due to them not having data collected in the UIS database. In all models, longer school closure durations are associated with greater deviations from long-term trends in TIMSS mathematics and science achievement. That is, countries where schools were closed for longer periods of time fell further below the expected linear trend than countries that closed schools for shorter periods of time. The duration of school closures was strongly associated with larger learning losses, particularly in mathematics. Within grade levels, longer closures were linked to more substantial losses in mathematics compared to science. Across grade levels, Grade 8 students experienced more significant learning losses per week of school closure compared to Grade 4 students, indicating that older students may have been more vulnerable to the extended disruptions in learning. This negative relationship ranged from about 0.16 points lost per week (grade 4 Science) up to 0.52 points lost per week (grade 8 mathematics). This indicates that if schools closed for one school year (approximately 36 weeks), losses would range from 6 to 18 points, depending on the grade level and subject. Table 2: Departure from linear trend and school closure duration Grade 4 Grade 8 Mathematics Science Mathematics Science Departure from linear trend -9.552*** -6.266*** -10.185*** -13.851*** (0.952) (0.969) (1.301) (1.318) School closure duration -0.247*** -0.156** -0.519*** -0.359*** (0.047) (0.049) (0.061) (0.060) N 1,287,355 1,287,355 1,409,298 1,409,298 N countries 74 74 72 72 Note. Standard errors in parentheses. * p < 0.05; ** p < 0.01; *** p < 0.001. Learning loss and school closure effects by student characteristics Table 3 next reports how both the deviations from long-term trends and school closure heterogeneity vary by sex and home language. When examining the impact by sex, it was found that girls experienced greater learning losses than boys across both subjects and grade levels. However, we find no evidence of significant differences in the effects of school closure duration by sex, suggesting that while girls experienced larger learning losses, the duration of closures did not disproportionately affect one sex over the other. The impact of speaking the test language at home was also explored. For Grade 4 students, those who did not speak the test language most often at home experienced greater learning losses than those who did. However, no such difference was found in Grade 8, where learning losses were similar for both groups. Furthermore, there were no significant differences in the impact of school closures based on home language when it came to the duration of school closures. Table 3 Departure from linear trend, school closure duration effects by sex and language Estimate Group G4 Math G4 Science G8 Math G8 Science Heterogeneity by sex Departure from linear trend Girls -12.510*** -7.790*** -16.155*** -19.871*** (1.025) (1.066) (1.399) (1.449) Boys -6.641*** -4.798*** -4.581** -8.075*** (1.138) (1.137) (1.606) (1.572) Difference (Girls – Boys) -5.869*** -2.992** -11.574*** -11.796*** (1.040) (1.067) (1.548) (1.544) School closure duration Girls -0.243*** -0.135* -0.511*** -0.383*** (0.053) (0.056) (0.067) (0.066) Boys -0.251*** -0.179** -0.531*** -0.338*** (0.059) (0.062) (0.076) (0.074) Difference (Girls – Boys) 0.008 0.044 0.020 -0.045 (0.063) (0.067) (0.078) (0.078) Heterogeneity by home language Departure from linear trend Do not speak test language most often -15.546*** -12.819*** -6.197* -17.651*** (1.592) (1.745) (2.662) (2.960) Speak test language most often -8.728*** -5.217*** -10.707*** -12.678*** (0.949) (0.958) (1.277) (1.249) Difference (not most often – most often) -6.818*** -7.602*** 4.510 -4.973 (1.524) (1.682) (2.484) (2.706) School closure duration Do not speak test language most often -0.226** -0.157 -0.371** -0.393** (0.079) (0.088) (0.125) (0.133) Speak test language most often -0.251*** -0.162*** -0.567*** -0.373*** (0.047) (0.048) (0.059) (0.056) Difference (not most often – most often) 0.025 0.005 0.195 -0.020 (0.077) (0.084) (0.115) (0.123) N 1,287,355 1,287,355 1,409,298 1,409,298 Note. Standard errors in parentheses. * p < 0.05; ** p < 0.01; *** p < 0.001. Learning loss by achievement level Analyses using unconditional quantile regressions reveal considerable losses among the lowest-achieving students in both subjects and grades. Figure 1 shows that learning loss estimates range from − 15 to -25 points for the lowest-achieving students (those at the 10th or 20th percentile of the achievement distribution) in grade 8 and − 10 to -20 in grade 4. Among the highest-achieving students, those in the 80th or 90th percentile of achievement, our estimates suggest no loss for 4th-grade science and a relatively small decline of around 5 points in 4th and 8th-grade mathematics and 8th-grade science. These results confirm earlier findings that the at-risk students lost more than those at the top of the achievement distribution. Overall, the findings demonstrate significant learning losses in both mathematics and science due to school closures, with variations by grade level, sex, home language, and achievement level. The results are robust to different regression and sample specifications (see Annex B with robustness checks). On average, grade 8 students were more affected by the duration of school closures than grade 4 students. Additionally, girls faced larger learning losses than boys across both subjects and grade levels. The study also highlights that students who did not speak the test language at home experienced greater learning losses in Grade 4 but not in Grade 8. Despite these differences, the duration of school closures did not appear to have a significantly different impact based on sex or home language. Finally, the low-achieving students suffered significantly larger losses when compared to average or high-achieving students. These findings underscore the widespread and varied effects of school closures on student learning. Discussion The COVID-19 pandemic caused widespread disruptions in education, leading to lasting impacts on student performance, particularly in mathematics and science. Using TIMSS 2023 data in conjunction with TIMSS trends starting in 2003, this study highlights the long-term effects of the pandemic, showing a global decline in student achievement by an average of 0.11 standard deviations compared to pre-pandemic trends. Low performers, girls, and linguistic minorities experienced losses of up to 0.23 SD. Countries with longer school closures saw more significant declines, with disadvantaged groups—especially low achievers, girls, and students facing language barriers—suffering disproportionately. The findings suggest that the most severe losses occurred among lower-achieving students, emphasizing the need for targeted interventions to address these disparities and protect educational equity in the long term. Moreover, the study’s findings emphasize the importance of timely and targeted recovery strategies. Recovery must involve tailored policies that not only focus on restoring lost content but also on providing additional support to the most vulnerable students. These results further reinforce the call for coordinated global efforts to tackle the educational crisis created by the pandemic and to ensure that future disruptions do not disproportionately affect the most disadvantaged students (Singh et al. 2024 ). The findings from this study underscore the need for targeted policy interventions. These include motivational nudges, such as text messages sent to students or their caregivers, which have been shown to significantly increase standardized test scores during remote learning (Angrist et al. 2022 ). Furthermore, evidence from Australia highlights the importance of targeted funding aimed at promoting more equitable educational outcomes. These efforts, which focused on providing additional resources to the most disadvantaged schools and students, helped mitigate some of the learning loss experienced during the pandemic (Miller et al. 2024 ). Policy measures that combine direct support for students with efforts to engage parents and communities can be a powerful tool in accelerating recovery and narrowing achievement gaps (Gray-Lobe 2024). During the school closures, high-impact, online tutoring was shown to be cost-effective (Gortazar et al. 2024 ). While this study provides valuable insights into the global learning loss caused by school closures, there are several limitations to consider. First, the reliance on TIMSS 2023 data means that the analysis is limited to the countries that participated in the study, excluding nations with lower participation or those that did not assess mathematics and science achievement in the same way. Additionally, while the study adjusts for demographic factors such as sex, socioeconomic status, and home language, it may not fully capture all the complexities of individual student experiences, such as variations in access to online learning resources, teacher quality, or parental involvement during the pandemic. The data on school closure durations, while comprehensive, may also be imprecise in some cases, especially in countries with localized or partial closures. In Finland, students who studied online for longer periods performed equally well in the matriculation exam at the end of upper-secondary education than the students who experienced shorter school closures (Riudavets-Barcons and Uusitalo 2024 ). Moreover, school closures may be underestimated. In the case of Sweden, even though there were no school closures on a national level, children still became sick, and classes were sent home (Skolverket, 2021 ). Finally, the study focuses on the academic subjects of mathematics and science, which may not fully reflect the broader impact of school closures on other areas of learning, such as language arts or social-emotional development. These limitations suggest that while the findings provide a useful overview of the pandemic’s educational impact, further research is needed to explore a wider range of variables and outcomes. Declarations Data availability De-identified data and study procedures are available from the authors on request via email. Declaration of interests We declare no competing interests. Supplementary Materials (2) Annex A and B References Almond, D. 2006. Is the 1918 influenza pandemic over? Long-term effects of in utero influenza exposure in the post-1940 U.S. population. Journal of Political Economy 114(4): 672–712. Angrist, N, Djankov, S, Goldberg, PK, Patrinos, HA. 2021. Measuring human capital using global learning data. Nature 592: 403–408. Angrist, N, Bergman, P, Matsheng, M. 2022. Experimental evidence on learning using low-tech when school is out. Nature Human Behaviour 6: 941–950. Belot, M, Webbink, D. 2010. Do teacher strikes harm educational attainment of students? Labour 24(4): 391–406. Betthäuser, B, Bach-Mortensen, A, Engzell, P. 2023. A systematic review and meta-analysis of the impact of the COVID-19 pandemic on learning. Nature Human Behaviour 7: 375–385. Bollyky, TJ, Castro, E, Aravkin, A, Bhangdia, K, Dalos, J, Hulland, E, et al. 2023. Assessing COVID-19 pandemic policies and behaviours and their economic and educational trade-offs across US states from Jan 1, 2020, to July 31, 2022: an observational analysis. The Lancet 401(10385): 1341-1360. https://doi.org/10.1016/S0140-6736(23)00461-0. Crato, N, Patrinos, HA. 2025. PIRLS 2021 and PISA 2022 Statistics Show How Serious the Pandemic Losses Are. In: Crato, N, Patrinos, HA (eds) Improving National Education Systems After COVID-19. Evaluating Education: Normative Systems and Institutional Practices . Springer, Cham. https://doi.org/10.1007/978-3-031-69284-0_1 Gambi, L, De Witte, K. 2024. The resilience in test scores and the mitigating effect of summer schools one year after the COVID-19 school closures. Labour Economics 90: 102549. Gortazar, L, Hupkau, C, Roldán-Monés, A. 2024. Online tutoring works: Experimental evidence from a program with vulnerable children. Journal of Public Economics 232: p.105082. Gray-Lobe, G, Kremer, M, de Laat, J, Mbonu, O, Scanlon, C. 2024. Nudging Parents out the Door: The Impacts of Parental Encouragement on School Choice and Test Scores. Haelermans, C, Jacobs, M, Van Der Velden, R, Van Vugt, L, Van Wetten, S. 2022. Inequality in the Effects of Primary School Closures due to the COVID-19 Pandemic. American Economic Association Papers and Proceedings 112: 303-307. Hammerstein, S, König, C, Dreisörner, T, Frey, A. 2021. Effects of COVID-19-related school closures on student achievement: A systematic review. Frontiers in Psychology 12: 746289. Ichino, A, Winter-Ebner, R. 2004. The long-run educational cost of World War II. Journal of Labor Economics 22(1): 57–86. Jakubowski, M, Gajderowicz, T, Patrinos, HA. 2025. COVID-19, school closures, and student learning outcomes. New global evidence from PISA. Nature Science of Learning 10(1): 5. Jaume, D, Willén, A. 2019. The long-run effects of teacher strikes: evidence from Argentina. Journal of Labor Economics 37(4): 1097–1139. Kennedy, AI, Strietholt, R. 2023. School closure policies and student reading achievement: Evidence across countries. Educational Assessment, Evaluation and Accountability 35(4): 475–50. Lichand, G, Doria, CA, Leal-Neto, O, Fernandes, JPC. 2022. The impacts of remote learning in secondary education during the pandemic in Brazil. Nature Human Behaviour 6: 1079–1086. Martin, MO, von Davier, M, Mullis, IV. 2020. Methods and procedures: TIMSS 2019 Technical Report. International Association for the Evaluation of Educational Achievement . Miller, A, Fray, L, Gore, J. 2024. Was COVID-19 an unexpected catalyst for more equitable learning outcomes? A comparative analysis after two years of disrupted schooling in Australian primary schools. Australian Educational Researcher 51(2): 587-608. Mislevy, RJ, Beaton, AE, Kaplan, B, Sheehan, KM. 1992. Estimating population characteristics from sparse matrix samples of item responses. Journal of Educational Measurement 29(2): 133–161. Psacharopoulos, G, Collis, V, Patrinos, HA, Vegas, E. 2021. The COVID-19 cost of school closures in earnings and income across the world. Comparative Education Review 65(2): 271-287. Rios-Avila, F. 2020. Recentered influence functions (RIFs) in Stata: RIF regression and RIF decomposition. Stata Journal 20(1): 51-94. Riudavets-Barcons, M, Uusitalo, R. 2024. School closures and student achievement, evidence from a high stakes exam. Journal of the Finnish Economic Association 5(1): 30-57. Rubin, DB. 2004. Multiple Imputation for Nonresponse in Surveys . John Wiley & Sons. Singh, A, Romero, M, Muralidharan, K. 2024. COVID-19 Learning loss and recovery: Panel data evidence from India. Journal of Human Resources . Skolverket. 2021. COVID-19-pandemins påverkan på skolväsendet: Delredovisning 4 - Regeringsuppdrag . Skolverket. https://www.skolverket.se/publikationer Smith, WC. 2021. Consequences of school closure on access to education: Lessons from the 2013-2016 Ebola pandemic. International Review of Education 67: 53-78. UNESCO. 2023. UNESCO figures show two thirds of an academic year lost on average worldwide due to Covid19 school closures. https://www.unesco.org/en/articles/unesco-figures-show-two-thirds-academic-yearlost-average-worldwide-due-covid-19-school-closures Vlachos, J, Hertegård, E, Svaleryd, HB. 2021. The effects of school closures on SARS-CoV-2 among parents and teachers. Proceedings of the National Academy of Sciences 118(9): p.e2020834118. WHO. 2024. Coronavirus Disease (COVID-19), https://www.who.int/health-topics/coronavirus#tab=tab_1, 30 November 2024. Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryMaterialsAnnexAandB.docx Supplementary Materails Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6566438","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Social Sciences - Article","associatedPublications":[],"authors":[{"id":453404499,"identity":"e44ec42a-8342-4258-a3ad-a59f8b8df45c","order_by":0,"name":"Harry Patrinos","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2ElEQVRIiWNgGAWjYBACAwglAcTMB0AMGeK0HABrYUsAaeEhVgsI8IDZhLWYs589+PnjHos8fv4zn1/dqLHgYWA/fHQDPi2WPXnJEgeeSRRLzsjdZp1zDOgwnrS0G3gddiDHQOLAAYnEDTd4txnnsAG1SPCY4ddy/o3xD7CW82eeGef8I0bLjRwziC0Hcpgf57YRpeVdmsUZoJaZM9LMmHP7JHjYCPrlfO7hGxUH6hL7+Q8//pzzrU6On/3wMbxakCOCTQJM4leOqoX5A2HVo2AUjIJRMBIBAFVxS8HO9O4CAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-4939-3568","institution":"University of Arkansas","correspondingAuthor":true,"prefix":"","firstName":"Harry","middleName":"","lastName":"Patrinos","suffix":""},{"id":453404500,"identity":"b3b9fe73-1996-437c-b14e-bbcccd301089","order_by":1,"name":"Tomasz Gajderowicz","email":"","orcid":"","institution":"University of Warsaw","correspondingAuthor":false,"prefix":"","firstName":"Tomasz","middleName":"","lastName":"Gajderowicz","suffix":""},{"id":453404501,"identity":"da44c78c-e2ff-4c3d-9399-f1ea87ec038c","order_by":2,"name":"Maciej Jakubowski","email":"","orcid":"https://orcid.org/0000-0002-2512-3133","institution":"University of Warsaw","correspondingAuthor":false,"prefix":"","firstName":"Maciej","middleName":"","lastName":"Jakubowski","suffix":""},{"id":453404502,"identity":"19adc835-cbcb-473c-a149-3c8b27a569b8","order_by":3,"name":"Alec Kennedy","email":"","orcid":"","institution":"International Association for the Evaluation of Educational Achievement","correspondingAuthor":false,"prefix":"","firstName":"Alec","middleName":"","lastName":"Kennedy","suffix":""},{"id":453404503,"identity":"11494f39-7db0-4a5b-a498-f1cbfb4155dd","order_by":4,"name":"Christian Kjeldsen","email":"","orcid":"","institution":"Aarhus University","correspondingAuthor":false,"prefix":"","firstName":"Christian","middleName":"","lastName":"Kjeldsen","suffix":""},{"id":453404504,"identity":"1239c963-9b8d-4a5f-a332-7a2a97934c04","order_by":5,"name":"Rolf Strietholt","email":"","orcid":"","institution":"International Association for the Evaluation of Educational Achievement","correspondingAuthor":false,"prefix":"","firstName":"Rolf","middleName":"","lastName":"Strietholt","suffix":""}],"badges":[],"createdAt":"2025-04-30 15:46:04","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6566438/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6566438/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":83813266,"identity":"88b85c00-18ec-46f5-a2c2-f2295ef52694","added_by":"auto","created_at":"2025-06-03 07:20:59","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":35881,"visible":true,"origin":"","legend":"\u003cp\u003eLearning loss by achievement percentile\u003c/p\u003e\n\u003cp\u003eNote: Results of unconditional RIF quantile regressions with 95% confidence intervals.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6566438/v1/564750b998955a695da67ecf.png"},{"id":83815072,"identity":"8d41ee03-e41e-4164-b297-5fd4cb90bd91","added_by":"auto","created_at":"2025-06-03 07:36:59","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":662589,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6566438/v1/d3fa2922-f227-4704-8657-c93c0fa6a29d.pdf"},{"id":83814538,"identity":"da077c73-acdc-4a5a-b1df-b86429e1037c","added_by":"auto","created_at":"2025-06-03 07:28:59","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":70411,"visible":true,"origin":"","legend":"Supplementary Materails","description":"","filename":"SupplementaryMaterialsAnnexAandB.docx","url":"https://assets-eu.researchsquare.com/files/rs-6566438/v1/a7b412b8aa5c62522c30ed96.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"The Learning Crisis: Three Years after COVID-19","fulltext":[{"header":"Introduction","content":"\u003cp\u003eCoronavirus (COVID-19) is an infectious disease caused by the SARS-CoV-2 virus. Most people infected with the virus will experience mild to moderate respiratory illness and recover without requiring special treatment. However, some will become seriously ill and require medical attention (WHO \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). These effects were particularly concerning in the early stages of the pandemic and restrictions were established with the intention of reducing the spread of COVID-19, including lockdowns, stay at home orders, and school closures. One of the arguments for closing schools was that school children at that age might be drivers of the spread within countries and thereby also endangering the older part of the population. However, there is not much evidence that opening schools significantly increased infection rates (Vlachos et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). During the COVID-19 crisis, one billion children were affected by school closures. Many countries were ill-prepared for the extensive school closures, but the few with prior experience of remote learning, trained teachers, appropriate technology, and engaged learners provided continuity during the crisis. Globally, schools were closed for an average of 5.5 months (22 weeks) since the onset of the pandemic, equivalent to two-thirds of an academic year, when localized school closures are considered (UNESCO \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The duration varies by region, from just one month in Oceania, to 2.5 months (10 weeks) in Europe, to as many as 5 months (20 weeks) in Latin America and the Caribbean. Early on, the school closures were expected to contribute to what was already described as a learning crisis (Angrist et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDistance learning during the school closures does not seem to have helped very much, with evidence that it increased dropout risk and lowered test scores (Lichand et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2022\u003c/span\u003ea; Haelermans et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2022\u003c/span\u003e); only the duration of school closures led to variations. Most studies observe increases in inequality where certain demographics of students experienced learning losses that were more significant than others. These learning losses could translate to earnings losses and could cost this generation of students trillions of dollars (Psacharopoulos et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThis study explores the global extent of learning loss using successive rounds of TIMSS (Trends in International Mathematics and Science Study) \u0026ndash; an international assessment of student achievement in mathematics and science at the fourth and eighth grades, conducted by the International Association for the Evaluation of Educational Achievement (IEA) \u0026ndash; up to the latest, which is 2023. By incorporating school closure duration, we examine how the length and nature of these disruptions affected student outcomes. This is the largest and most comprehensive investigation of learning loss over time, offering a global perspective that includes both pre- and during-pandemic data. We investigate the variation in impacts based on school closure duration and demographic factors, providing a comprehensive view of global trends in educational recovery. Our findings offer valuable insights into the persistence of learning losses in the years following school closures. This study reveals a global decline of, on average, 0.11 standard deviations, with more severe losses linked to longer closures, particularly among girls and linguistic minorities. Low performers lost up to 0.23 standard deviations, or more than half a year\u0026rsquo;s worth of learning.\u003c/p\u003e \u003cp\u003eWe make several contributions to the existing literature. First, we improve the precision of estimates regarding the impact of school closures on learning outcomes. Previous reviews of national studies show large learning losses, with estimates ranging from one-third to half a year's worth of learning (Betth\u0026auml;user et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Hammerstein et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Patrinos et al. 2023). Reviews of studies produce slightly lower losses than the analyses using the international student assessments. This could be because the reviews have limited samples. For example, Betthauser et al. (2023) review 42 studies and lacks evidence on lower income countries. The international assessments have many more countries and are more representative of countries at different income levels. International student assessments, such as TIMSS, offer more comprehensive and representative data across countries with varying income levels, allowing for better cross-country comparisons. The average learning losses identified through large, comparable, international standardized assessments are 0.19 of a standard deviation, or a half years\u0026rsquo; worth of learning (Crato and Patrinos \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). TIMSS 2023 offers data up to Spring 2023, providing the most recent measures of student achievement, as prior assessments like PIRLS and PISA were conducted before or during the pandemic.\u003c/p\u003e \u003cp\u003eAnother key contribution of this study is its inclusion of evidence on learning outcomes from both before and during school closures for most countries, with a longer duration of data that covers the entire period of closures for all countries. This allows for a more detailed understanding of the lingering impact of school closures, as prior research has shown long-lasting learning losses (Bollyky et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Gambi and De Witte \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), although some countries experienced quicker recovery (Miller et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The study also highlights the disparities in recovery across countries, with ongoing challenges for many students, particularly those from disadvantaged backgrounds, in regaining lost learning.\u003c/p\u003e \u003cp\u003eFinally, this paper contributes to the growing body of literature on the effects of pandemics and other disruptions to children\u0026rsquo;s schooling (Belot and Webbink \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Ichino and Winter-Ebner \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Jaume and Will\u0026eacute;n \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Previous research on past pandemics, such as the 1918 influenza (Almond \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2006\u003c/span\u003e) and the 2013\u0026ndash;2016 Ebola outbreak (Smith \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), has shown that school closures result in long-term negative effects on educational attainment and earnings. By adding to this literature, our study contributed to our understanding of the lasting (side) effect of school closures on global education, emphasizing the need for targeted recovery efforts to mitigate these losses and address the widening educational inequalities exacerbated by the pandemic.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eWe utilize data from the Trends in International Mathematics and Science Study (TIMSS), an international large-scale, repeated cross-sectional study that randomly samples fourth- and eighth-grade students to assess their achievement in mathematics and science. The curriculum-based assessments measure the knowledge students have accumulated over four and eight years of schooling, respectively. Conducted in four-year cycles, the 2023 cycle marks the first administration following the onset of the COVID-19 pandemic. Drawing on data spanning 20 years, starting from 1999, this study examines long-term trends in student performance, allowing us to evaluate whether the 2023 results diverge from pre-pandemic patterns. Altogether, we use data from more than 2\u0026nbsp;million students across all TIMSS cycles, including 71 educational systems (58 for Grade 4 and 44 for Grade 8) that participated in TIMSS 2023, as well as data from earlier cycles (see Annex A for summary statistics by country). We only excluded data from 1995 as it relied on a different psychometric model for scaling achievement, so the scores may not be fully comparable to those from subsequent years. The achievement scores were standardized separately for the two outcomes and populations in the initial TIMSS administration, using a scale with an international mean of 500 and a standard deviation of 100. They are comparable over time. TIMSS not only captures student achievement but also contextual information and characteristics of students through background questionnaires. We use this information as control variables to explain differences in achievement scores while also capturing changes to country demographics or samples over time. Specifically, we use four pieces of information that were consistently captured across all cycles of TIMSS from the student background questionnaire: student age, sex, and how often the test language was spoken at home.\u003c/p\u003e \u003cp\u003eAdditionally, data on school closure durations are obtained from UIS (UNESCO’s Institute for Statistics (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://covid19.uis.unesco.org/global-monitoring-school-closures-covid19/\u003c/span\u003e\u003cspan address=\"https://covid19.uis.unesco.org/global-monitoring-school-closures-covid19/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), detailing the number of weeks of full and partial closures in each country (see Annex A). Schools were considered fully closed in cases where government mandates required schools to be closed, affecting most or all students. In contrast, schools were considered partially open when schools were closed only in certain regions or for some grade levels. This also captured schools that were only partially open to in-person instruction (e.g., hybrid learning). This data allows us to assess how learning losses varied by the duration of school closures.\u003c/p\u003e \u003cp\u003eTo capture changes following COVID-19, it is essential to take previous long-term trends into account. Our empirical strategy involves the estimation of deviations of the most recent 2023 result from long-term linear trends in average student mathematics and science achievement for each participating school system (see Jakubowski et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Kennedy and Strietholt \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). We model a separate linear trend for each country and include country-level fixed effects to control for unobserved time-invariant country characteristics. This is formalized in the following regression model:\u003c/p\u003e\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:{Y}_{ijk}={{\\Sigma\\:}}_{k=1}^{n}{\\alpha\\:}_{k}+{{\\Sigma\\:}}_{k=1}^{n}{\\beta\\:}_{k}*time+\\tau\\:{D}_{2023}+\\gamma\\:{X}_{ijk}+{\\epsilon\\:}_{ijk}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003cp\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{Y}_{ijk}\\)\u003c/span\u003e\u003c/span\u003e represents the achievement of student \u003cem\u003ei\u003c/em\u003e at school \u003cem\u003ej\u003c/em\u003e in country \u003cem\u003ek\u003c/em\u003e, with \u003cem\u003en\u003c/em\u003e being the number of countries in the sample being analyzed. The model is estimated on repeated cross-sections, with \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\beta\\:}_{k}\\)\u003c/span\u003e\u003c/span\u003e capturing the slope of country-specific linear trends in student achievement across TIMSS cycles. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{D}_{2023}\\)\u003c/span\u003e\u003c/span\u003e is an indicator variable that is equal to one for the data collected during the 2023 cycle (after the onset of the pandemic) and is zero for all other cycles; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\tau\\:\\)\u003c/span\u003e\u003c/span\u003e is our parameter of interest capturing the deviation from country-specific trends occurring after the onset of the pandemic after controlling for time-invariant country effects (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\alpha\\:}_{k}\\)\u003c/span\u003e\u003c/span\u003e) and student background characteristics (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{X}_{ijk}\\)\u003c/span\u003e\u003c/span\u003e). The model is estimated separately for both TIMSS subjects (i.e., mathematics and science) and both populations (i.e., grade 4 and grade 8).\u003c/p\u003e \u003cp\u003eWe next incorporate measures of national school closure policy duration to understand how our estimates of learning loss vary by the length of school closures. This is done through the introduction of an interaction term represented by the regression equation below:\u003c/p\u003e\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\:{Y}_{ijk}={{\\Sigma\\:}}_{k=1}^{n}{\\alpha\\:}_{k}+{{\\Sigma\\:}}_{k=1}^{n}{\\beta\\:}_{k}*time+\\tau\\:{D}_{2023}+\\pi\\:{D}_{2023}*week{s}_{k}+\\gamma\\:{X}_{ijk}+{\\epsilon\\:}_{ijk}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003cp\u003e\u003c/p\u003e \u003cp\u003ewith the model specified as it was in Eq.\u0026nbsp;\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, with an additional interaction term between \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{D}_{2023}\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:week{s}_{k}\\)\u003c/span\u003e\u003c/span\u003e, the number of weeks that schools were closed in country \u003cem\u003ek\u003c/em\u003e; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\tau\\:\\)\u003c/span\u003e\u003c/span\u003e can be interpreted as the deviation from the trend for a country that has no weeks of school closure; and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pi\\:\\)\u003c/span\u003e\u003c/span\u003e represents how those deviations vary by additional week of school closures for a country.\u003c/p\u003e \u003cp\u003eThe model is expanded further to test for differences in the impact of the pandemic and school closure policies by sex and home language. Each variable is dichotomized to provide three different contrasts: girls versus boys, and those who never or only sometimes speak the test language at home versus those who speak it most of the time. These contrasts are made by adding additional interaction terms between the individual student characteristics and both the \u003cem\u003eD\u003c/em\u003e\u003csub\u003e\u003cem\u003e2023\u003c/em\u003e\u003c/sub\u003e and \u003cem\u003eD\u003c/em\u003e\u003csub\u003e\u003cem\u003e2023\u003c/em\u003e\u003c/sub\u003e*\u003cem\u003eweeks\u003c/em\u003e\u003csub\u003e\u003cem\u003ek\u003c/em\u003e\u003c/sub\u003e from the model. Additionally, individual characteristics interact with the time trend variable, enabling trends to be estimated separately for each student group.\u003c/p\u003e \u003cp\u003eFinally, we investigate the heterogeneity in the impact of school closures by achievement level. To check how learning losses vary among students of different proficiency, we estimate the main models using quantile RIF regressions, fitting models of unconditional of TIMSS achievement (Rios-Avila \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eData from TIMSS is collected using a complex stratified sampling scheme. To account for this, standard errors are calculated using a jackknife repeated replication (JRR) technique that accounts for sampling variance arising from the stratified class-based sampling design (Martin et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Furthermore, TIMSS uses a rotated booklet design and employs the use of plausible values methodology to estimate students’ achievement (Mislevy et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e1992\u003c/span\u003e). Therefore, all results presented are based on estimation accounting for the variation across the five plausible values for each subject scale (Rubin \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2004\u003c/span\u003e).\u003c/p\u003e \n\n\n"},{"header":"Findings","content":"\u003cp\u003eThe results highlight significant learning losses due to school closures, with variations across subjects, grade levels, and demographic factors such as sex and home language. For both Grade 4 and Grade 8 students, performance in mathematics and science fell significantly below expectations based on the linear trends observed in prior TIMSS cycles. Grade 8 students experienced greater declines across both subjects compared to Grade 4 students. We examine the impact of school closures on student learning in mathematics and science, analyzing data from 78 countries for Grade 4 students and 74 countries for Grade 8 students.\u003c/p\u003e\u003ch3\u003eLearning loss estimates\u003c/h3\u003e\u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e presents the estimated deviations from long-term trends by subject and grade. In Grade 4, the departure from trend was smaller for science than for mathematics. For Grade 8, the declines were larger in mathematics and science. To make the magnitudes of the linear trend departures comparable, we standardize them by the average within-country SD for each grade level and subject (SD\u003csub\u003eGrade4,math\u003c/sub\u003e=84, SD\u003csub\u003eGrade4,science\u003c/sub\u003e=85, SD\u003csub\u003eGrade8,math\u003c/sub\u003e=90, SD\u003csub\u003eGrade4,science\u003c/sub\u003e=91). The corresponding effect sizes (Cohen’s d) for Grade 4 were 0.11 for mathematics and 0.06 for science, while for Grade 8, they were slightly larger at 0.12 for mathematics and 0.14 for science. These results demonstrate significant negative effects on learning outcomes across all subjects and grade levels.\u003c/p\u003e\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\"\u003e\u003c/p\u003e\u003ch3\u003eLearning loss by school closure duration\u003c/h3\u003e\u003cp\u003eTable\u0026nbsp;2 presents how the deviations from the linear trend vary by duration of national school closures as captured in the UIS database considering both full and partial school closures. Note that some countries had to be omitted from the results due to them not having data collected in the UIS database. In all models, longer school closure durations are associated with greater deviations from long-term trends in TIMSS mathematics and science achievement. That is, countries where schools were closed for longer periods of time fell further below the expected linear trend than countries that closed schools for shorter periods of time.\u003c/p\u003e\u003cp\u003eThe duration of school closures was strongly associated with larger learning losses, particularly in mathematics. Within grade levels, longer closures were linked to more substantial losses in mathematics compared to science. Across grade levels, Grade 8 students experienced more significant learning losses per week of school closure compared to Grade 4 students, indicating that older students may have been more vulnerable to the extended disruptions in learning.\u003c/p\u003e\u003cp\u003eThis negative relationship ranged from about 0.16 points lost per week (grade 4 Science) up to 0.52 points lost per week (grade 8 mathematics). This indicates that if schools closed for one school year (approximately 36 weeks), losses would range from 6 to 18 points, depending on the grade level and subject.\u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003ctable float=\"No\" id=\"Taba\" border=\"1\"\u003e\u003ccolgroup cols=\"6\"\u003e\u003c/colgroup\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e \u003cp\u003eTable\u0026nbsp;2: Departure from linear trend and school closure duration\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003eGrade 4\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e \u003cp\u003eGrade 8\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMathematics\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eScience\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMathematics\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eScience\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eDeparture from linear trend\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-9.552***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-6.266***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-10.185***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-13.851***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.952)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.969)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.301)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(1.318)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eSchool closure duration\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.247***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.156**\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.519***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.359***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(0.047)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.049)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.061)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.060)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1,287,355\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1,287,355\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,409,298\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1,409,298\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN countries\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e74\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e74\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e72\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e72\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c6\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cem\u003eNote.\u003c/em\u003e Standard errors in parentheses. * p \u0026lt; 0.05; ** p \u0026lt; 0.01; *** p \u0026lt; 0.001.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003c/div\u003e\u003ch3\u003eLearning loss and school closure effects by student characteristics\u003c/h3\u003e\u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e3\u003c/span\u003e next reports how both the deviations from long-term trends and school closure heterogeneity vary by sex and home language.\u003c/p\u003e\u003cp\u003eWhen examining the impact by sex, it was found that girls experienced greater learning losses than boys across both subjects and grade levels. However, we find no evidence of significant differences in the effects of school closure duration by sex, suggesting that while girls experienced larger learning losses, the duration of closures did not disproportionately affect one sex over the other.\u003c/p\u003e\u003cp\u003eThe impact of speaking the test language at home was also explored. For Grade 4 students, those who did not speak the test language most often at home experienced greater learning losses than those who did. However, no such difference was found in Grade 8, where learning losses were similar for both groups. Furthermore, there were no significant differences in the impact of school closures based on home language when it came to the duration of school closures.\u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDeparture from linear trend, school closure duration effects by sex and language\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e\u003ccolgroup cols=\"6\"\u003e\u003c/colgroup\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEstimate\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGroup\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eG4 Math\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eG4 Science\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eG8 Math\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eG8 Science\u003c/p\u003e \u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cem\u003eHeterogeneity by sex\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e \u003cp\u003eDeparture from linear trend\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eGirls\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-12.510***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-7.790***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-16.155***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-19.871***\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.025)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.066)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(1.399)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(1.449)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eBoys\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-6.641***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-4.798***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-4.581**\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-8.075***\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.138)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.137)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(1.606)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(1.572)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eDifference\u003c/p\u003e \u003cp\u003e(Girls – Boys)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-5.869***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-2.992**\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-11.574***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-11.796***\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.040)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.067)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(1.548)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(1.544)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e \u003cp\u003eSchool closure duration\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eGirls\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.243***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.135*\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.511***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.383***\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.053)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.056)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.067)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.066)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eBoys\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.251***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.179**\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.531***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.338***\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.059)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.062)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.076)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.074)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eDifference\u003c/p\u003e \u003cp\u003e(Girls – Boys)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.008\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.044\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.020\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.045\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.063)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.067)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.078)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.078)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cem\u003eHeterogeneity by home language\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e \u003cp\u003eDeparture from linear trend\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eDo not speak test language most often\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-15.546***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-12.819***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-6.197*\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-17.651***\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.592)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.745)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(2.662)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(2.960)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eSpeak test language most often\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-8.728***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-5.217***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-10.707***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-12.678***\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.949)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.958)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(1.277)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(1.249)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eDifference (not most often – most often)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-6.818***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-7.602***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.510\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-4.973\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.524)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(1.682)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(2.484)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(2.706)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e \u003cp\u003eSchool closure duration\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eDo not speak test language most often\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.226**\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.157\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.371**\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.393**\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.079)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.088)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.125)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.133)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eSpeak test language most often\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.251***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.162***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.567***\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.373***\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.047)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.048)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.059)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.056)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eDifference (not most often – most often)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.025\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.005\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.195\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0.020\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.077)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e(0.084)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e(0.115)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e(0.123)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eN\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1,287,355\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,287,355\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1,409,298\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1,409,298\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cem\u003eNote.\u003c/em\u003e Standard errors in parentheses. * p \u0026lt; 0.05; ** p \u0026lt; 0.01; *** p \u0026lt; 0.001.\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003c/div\u003e\u003ch3\u003eLearning loss by achievement level\u003c/h3\u003e\u003cp\u003eAnalyses using unconditional quantile regressions reveal considerable losses among the lowest-achieving students in both subjects and grades. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e shows that learning loss estimates range from − 15 to -25 points for the lowest-achieving students (those at the 10th or 20th percentile of the achievement distribution) in grade 8 and − 10 to -20 in grade 4. Among the highest-achieving students, those in the 80th or 90th percentile of achievement, our estimates suggest no loss for 4th-grade science and a relatively small decline of around 5 points in 4th and 8th-grade mathematics and 8th-grade science. These results confirm earlier findings that the at-risk students lost more than those at the top of the achievement distribution.\u003c/p\u003e\u003cp\u003eOverall, the findings demonstrate significant learning losses in both mathematics and science due to school closures, with variations by grade level, sex, home language, and achievement level. The results are robust to different regression and sample specifications (see Annex B with robustness checks). On average, grade 8 students were more affected by the duration of school closures than grade 4 students. Additionally, girls faced larger learning losses than boys across both subjects and grade levels. The study also highlights that students who did not speak the test language at home experienced greater learning losses in Grade 4 but not in Grade 8. Despite these differences, the duration of school closures did not appear to have a significantly different impact based on sex or home language. Finally, the low-achieving students suffered significantly larger losses when compared to average or high-achieving students. These findings underscore the widespread and varied effects of school closures on student learning.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe COVID-19 pandemic caused widespread disruptions in education, leading to lasting impacts on student performance, particularly in mathematics and science. Using TIMSS 2023 data in conjunction with TIMSS trends starting in 2003, this study highlights the long-term effects of the pandemic, showing a global decline in student achievement by an average of 0.11 standard deviations compared to pre-pandemic trends. Low performers, girls, and linguistic minorities experienced losses of up to 0.23 SD. Countries with longer school closures saw more significant declines, with disadvantaged groups\u0026mdash;especially low achievers, girls, and students facing language barriers\u0026mdash;suffering disproportionately. The findings suggest that the most severe losses occurred among lower-achieving students, emphasizing the need for targeted interventions to address these disparities and protect educational equity in the long term.\u003c/p\u003e \u003cp\u003eMoreover, the study\u0026rsquo;s findings emphasize the importance of timely and targeted recovery strategies. Recovery must involve tailored policies that not only focus on restoring lost content but also on providing additional support to the most vulnerable students. These results further reinforce the call for coordinated global efforts to tackle the educational crisis created by the pandemic and to ensure that future disruptions do not disproportionately affect the most disadvantaged students (Singh et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe findings from this study underscore the need for targeted policy interventions. These include motivational nudges, such as text messages sent to students or their caregivers, which have been shown to significantly increase standardized test scores during remote learning (Angrist et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Furthermore, evidence from Australia highlights the importance of targeted funding aimed at promoting more equitable educational outcomes. These efforts, which focused on providing additional resources to the most disadvantaged schools and students, helped mitigate some of the learning loss experienced during the pandemic (Miller et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Policy measures that combine direct support for students with efforts to engage parents and communities can be a powerful tool in accelerating recovery and narrowing achievement gaps (Gray-Lobe 2024). During the school closures, high-impact, online tutoring was shown to be cost-effective (Gortazar et al. \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWhile this study provides valuable insights into the global learning loss caused by school closures, there are several limitations to consider. First, the reliance on TIMSS 2023 data means that the analysis is limited to the countries that participated in the study, excluding nations with lower participation or those that did not assess mathematics and science achievement in the same way. Additionally, while the study adjusts for demographic factors such as sex, socioeconomic status, and home language, it may not fully capture all the complexities of individual student experiences, such as variations in access to online learning resources, teacher quality, or parental involvement during the pandemic. The data on school closure durations, while comprehensive, may also be imprecise in some cases, especially in countries with localized or partial closures. In Finland, students who studied online for longer periods performed equally well in the matriculation exam at the end of upper-secondary education than the students who experienced shorter school closures (Riudavets-Barcons and Uusitalo \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Moreover, school closures may be underestimated. In the case of Sweden, even though there were no school closures on a national level, children still became sick, and classes were sent home (Skolverket, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Finally, the study focuses on the academic subjects of mathematics and science, which may not fully reflect the broader impact of school closures on other areas of learning, such as language arts or social-emotional development. These limitations suggest that while the findings provide a useful overview of the pandemic\u0026rsquo;s educational impact, further research is needed to explore a wider range of variables and outcomes.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDe-identified data and study procedures are available from the authors on request via email.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eDeclaration of interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eSupplementary Materials (2)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAnnex A and B\u003c/p\u003e\n"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAlmond, D. 2006. Is the 1918 influenza pandemic over? Long-term effects of in utero influenza exposure in the post-1940 U.S. population. \u003cem\u003eJournal of Political Economy\u003c/em\u003e 114(4): 672\u0026ndash;712.\u003c/li\u003e\n\u003cli\u003eAngrist, N, Djankov, S, Goldberg, PK, Patrinos, HA. 2021. Measuring human capital using global learning data. \u003cem\u003eNature\u003c/em\u003e 592: 403\u0026ndash;408.\u003c/li\u003e\n\u003cli\u003eAngrist, N, Bergman, P, Matsheng, M. 2022. Experimental evidence on learning using low-tech when school is out. \u003cem\u003eNature Human Behaviour\u003c/em\u003e 6: 941\u0026ndash;950.\u003c/li\u003e\n\u003cli\u003eBelot, M, Webbink, D. 2010. Do teacher strikes harm educational attainment of students? \u003cem\u003eLabour\u003c/em\u003e 24(4): 391\u0026ndash;406.\u003c/li\u003e\n\u003cli\u003eBetth\u0026auml;user, B, Bach-Mortensen, A, Engzell, P. 2023. A systematic review and meta-analysis of the impact of the COVID-19 pandemic on learning. \u003cem\u003eNature Human Behaviour\u003c/em\u003e7: 375\u0026ndash;385.\u003c/li\u003e\n\u003cli\u003eBollyky, TJ, Castro, E, Aravkin, A, Bhangdia, K, Dalos, J, Hulland, E, et al. 2023. Assessing COVID-19 pandemic policies and behaviours and their economic and educational trade-offs across US states from Jan 1, 2020, to July 31, 2022: an observational analysis. \u003cem\u003eThe Lancet\u003c/em\u003e 401(10385): 1341-1360. https://doi.org/10.1016/S0140-6736(23)00461-0.\u003c/li\u003e\n\u003cli\u003eCrato, N, Patrinos, HA. 2025. PIRLS 2021 and PISA 2022 Statistics Show How Serious the Pandemic Losses Are. In: Crato, N, Patrinos, HA (eds) \u003cem\u003eImproving National Education Systems After COVID-19. Evaluating Education: Normative Systems and Institutional Practices\u003c/em\u003e. Springer, Cham. https://doi.org/10.1007/978-3-031-69284-0_1\u003c/li\u003e\n\u003cli\u003eGambi, L, De Witte, K. 2024. The resilience in test scores and the mitigating effect of summer schools one year after the COVID-19 school closures. \u003cem\u003eLabour Economics\u003c/em\u003e 90: 102549.\u003c/li\u003e\n\u003cli\u003eGortazar, L, Hupkau, C, Rold\u0026aacute;n-Mon\u0026eacute;s, A. 2024. Online tutoring works: Experimental evidence from a program with vulnerable children. \u003cem\u003eJournal of Public Economics\u003c/em\u003e 232: p.105082.\u003c/li\u003e\n\u003cli\u003eGray-Lobe, G, Kremer, M, de Laat, J, Mbonu, O, Scanlon, C. 2024. Nudging Parents out the Door: The Impacts of Parental Encouragement on School Choice and Test Scores.\u003c/li\u003e\n\u003cli\u003eHaelermans, C, Jacobs, M, Van Der Velden, R, Van Vugt, L, Van Wetten, S. 2022. Inequality in the Effects of Primary School Closures due to the COVID-19 Pandemic. \u003cem\u003eAmerican Economic Association Papers and Proceedings\u003c/em\u003e 112: 303-307.\u003c/li\u003e\n\u003cli\u003eHammerstein, S, K\u0026ouml;nig, C, Dreis\u0026ouml;rner, T, Frey, A. 2021. Effects of COVID-19-related school closures on student achievement: A systematic review. \u003cem\u003eFrontiers in Psychology\u003c/em\u003e 12: 746289.\u003c/li\u003e\n\u003cli\u003eIchino, A, Winter-Ebner, R. 2004. The long-run educational cost of World War II. \u003cem\u003eJournal of Labor Economics\u003c/em\u003e 22(1): 57\u0026ndash;86.\u003c/li\u003e\n\u003cli\u003eJakubowski, M, Gajderowicz, T, Patrinos, HA. 2025. COVID-19, school closures, and student learning outcomes. New global evidence from PISA. \u003cem\u003eNature Science of Learning\u003c/em\u003e 10(1): 5.\u003c/li\u003e\n\u003cli\u003eJaume, D, Will\u0026eacute;n, A. 2019. The long-run effects of teacher strikes: evidence from Argentina. \u003cem\u003eJournal of Labor Economics\u003c/em\u003e 37(4): 1097\u0026ndash;1139.\u003c/li\u003e\n\u003cli\u003eKennedy, AI, Strietholt, R. 2023. School closure policies and student reading achievement: Evidence across countries. \u003cem\u003eEducational Assessment, Evaluation and Accountability\u003c/em\u003e 35(4): 475\u0026ndash;50.\u003c/li\u003e\n\u003cli\u003eLichand, G, Doria, CA, Leal-Neto, O, Fernandes, JPC. 2022. The impacts of remote learning in secondary education during the pandemic in Brazil. \u003cem\u003eNature Human Behaviour\u003c/em\u003e 6: 1079\u0026ndash;1086.\u003c/li\u003e\n\u003cli\u003eMartin, MO, von Davier, M, Mullis, IV. 2020. Methods and procedures: TIMSS 2019 Technical Report. \u003cem\u003eInternational Association for the Evaluation of Educational Achievement\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eMiller, A, Fray, L, Gore, J. 2024. Was COVID-19 an unexpected catalyst for more equitable learning outcomes? A comparative analysis after two years of disrupted schooling in Australian primary schools. \u003cem\u003eAustralian Educational Researcher\u003c/em\u003e 51(2): 587-608.\u003c/li\u003e\n\u003cli\u003eMislevy, RJ, Beaton, AE, Kaplan, B, Sheehan, KM. 1992. Estimating population characteristics from sparse matrix samples of item responses. \u003cem\u003eJournal of Educational Measurement\u003c/em\u003e 29(2): 133\u0026ndash;161.\u003c/li\u003e\n\u003cli\u003ePsacharopoulos, G, Collis, V, Patrinos, HA, Vegas, E. 2021. The COVID-19 cost of school closures in earnings and income across the world. \u003cem\u003eComparative Education Review \u003c/em\u003e65(2): 271-287.\u003c/li\u003e\n\u003cli\u003eRios-Avila, F. 2020. Recentered influence functions (RIFs) in Stata: RIF regression and RIF decomposition. \u003cem\u003eStata Journal\u003c/em\u003e 20(1): 51-94.\u003c/li\u003e\n\u003cli\u003eRiudavets-Barcons, M, Uusitalo, R. 2024. School closures and student achievement, evidence from a high stakes exam. \u003cem\u003eJournal of the Finnish Economic Association\u003c/em\u003e 5(1): 30-57.\u003c/li\u003e\n\u003cli\u003eRubin, DB. 2004. \u003cem\u003eMultiple Imputation for Nonresponse in Surveys\u003c/em\u003e. John Wiley \u0026amp; Sons.\u003c/li\u003e\n\u003cli\u003eSingh, A, Romero, M, Muralidharan, K. 2024. COVID-19 Learning loss and recovery: Panel data evidence from India. \u003cem\u003eJournal of Human Resources\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eSkolverket. 2021. \u003cem\u003eCOVID-19-pandemins p\u0026aring;verkan p\u0026aring; skolv\u0026auml;sendet: Delredovisning 4 - Regeringsuppdrag\u003c/em\u003e. Skolverket. https://www.skolverket.se/publikationer\u003c/li\u003e\n\u003cli\u003eSmith, WC. 2021. Consequences of school closure on access to education: Lessons from the 2013-2016 Ebola pandemic. \u003cem\u003eInternational Review of Education\u003c/em\u003e 67: 53-78.\u003c/li\u003e\n\u003cli\u003eUNESCO. 2023. UNESCO figures show two thirds of an academic year lost on average worldwide due to Covid19 school closures. https://www.unesco.org/en/articles/unesco-figures-show-two-thirds-academic-yearlost-average-worldwide-due-covid-19-school-closures\u003c/li\u003e\n\u003cli\u003eVlachos, J, Herteg\u0026aring;rd, E, Svaleryd, HB. 2021. The effects of school closures on SARS-CoV-2 among parents and teachers. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e 118(9): p.e2020834118.\u003c/li\u003e\n\u003cli\u003eWHO. 2024. Coronavirus Disease (COVID-19), https://www.who.int/health-topics/coronavirus#tab=tab_1, 30 November 2024.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6566438/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6566438/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe COVID-19 pandemic triggered widespread school closures affecting over one billion children worldwide, with potentially lasting effects on learning. Leveraging data from the 2023 Trends in International Mathematics and Science Study (TIMSS), covering 71 education systems and spanning two decades, we quantify the global impact of pandemic-related disruptions on student achievement in mathematics and science. Using mixed-effects models, we estimate deviations from pre-pandemic trends, adjusting for demographic and school-level factors. We find that average achievement declined by 0.11 standard deviations globally, with longer school closures associated with larger losses. Disadvantaged groups\u0026mdash;including low performers, girls, and linguistic minorities\u0026mdash;experienced greater setbacks, reaching up to 0.23 standard deviations. These findings reveal a persistent global learning crisis and highlight the urgent need for targeted policy responses to address educational inequalities exacerbated by the pandemic.\u003c/p\u003e","manuscriptTitle":"The Learning Crisis: Three Years after COVID-19","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-03 07:20:54","doi":"10.21203/rs.3.rs-6566438/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"4a2b285d-3603-48db-b4ec-4950cf54c40b","owner":[],"postedDate":"June 3rd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":48218250,"name":"Scientific community and society/Social sciences/Economics"},{"id":48218251,"name":"Scientific community and society/Social sciences/Education"}],"tags":[],"updatedAt":"2025-06-03T07:20:54+00:00","versionOfRecord":[],"versionCreatedAt":"2025-06-03 07:20:54","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6566438","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6566438","identity":"rs-6566438","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}