📄 social_science

Zooming In or Left Out?: How Covid Changed Schooling, Screen, and Social Time for High School and College Students by Household Income

Using 2017–2022 American Time Use Survey data, this study reveals that the pandemic caused high school and college students to significantly reduce educational class time and socialization while increasing screen time, with these negative shifts disproportionately affecting low-income students.

Original authors: Laura M. Crispin, Elizabeth Mitchell

Published 2026-07-14

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Laura M. Crispin, Elizabeth Mitchell

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine your daily life as a giant, 24-hour pie chart. Before the pandemic, this pie was sliced up into school, homework, hanging out with friends, sleeping, and maybe a little bit of working a part-time job. Then, in March 2020, the world hit the "pause" button, and schools flipped a switch to virtual learning.

This paper acts like a time-traveling detective, using a massive survey called the American Time Use Survey to look at how high schoolers and college students living at home sliced up their pie before the virus (2017 to early 2020) versus during the chaos (mid-2020 to late 2022). The authors, Laura and Elizabeth, wanted to see if the pandemic just changed where we learned, or if it actually changed how much we learned, and if that change was fair for everyone.

The Big Slice That Vanished: Class Time
The most dramatic change wasn't a mystery; it was a disappearance act. The data shows that students spent significantly less time in class during the pandemic. Specifically, on average, students lost about 33 minutes of educational time per day.

But here is the twist: it wasn't that they stopped doing homework. The paper explicitly rules out the idea that students simply swapped class time for more homework. Homework time stayed roughly the same. Instead, the "class time" slice of the pie just got smaller. For high schoolers, this was a drop of 22 minutes a day. For college students, it was a drop of 43 minutes.

If you do the math on a five-day school week, that adds up to a loss of about 1.5 hours of class time per week for the average student. And for students from lower-income households? The slice didn't just shrink; it nearly vanished. They lost a massive 75 minutes of educational time per day, with class time alone dropping by 50 minutes (or 4.2 hours a week).

The "Screen Time" Glitch
So, where did all that missing class time go? Did students suddenly become super-workers or marathon runners? The paper says no. There was no significant increase in paid work, exercise, or even sleep to fill the gap.

Instead, the missing time was replaced by a giant, glowing screen. The data suggests that students increased their leisure time by 19 minutes and their screen time by a whopping 37 minutes per day. For low-income students, this screen time jump was even steeper, rising by 53 minutes a day compared to their wealthier peers. It's as if the "class" slot in the schedule was swapped out for a "scrolling" slot.

The Great Divide: Rich vs. Poor Pies
The pandemic didn't affect everyone's pie chart equally. The authors found a stark difference based on household income.

Higher-income students saw their class time drop by about 10 to 29 minutes a day. They managed to keep their homework steady or even slightly increase it.
Lower-income students saw their class time plummet by 50 to 75 minutes a day. They also spent significantly more time socializing online but less time exercising (dropping by 48 minutes for low-income groups).

The paper suggests that lower-income students might have faced bigger hurdles, like not having reliable internet or the right gadgets to attend virtual classes, leading to this massive gap in learning time.

The "Alone" Factor
Another interesting finding is about who was in the room. For college students living at home, the pandemic made them more likely to do their education alone. The likelihood of studying at home by oneself increased by 11.6 percentage points, and the time spent doing so jumped by 65 minutes. It seems the classroom moved from a bustling hall to a quiet, isolated bedroom.

What the Paper Says It's NOT
It's important to note what this study didn't find. It didn't find that students worked harder on homework to make up for lost class time. It didn't find that students suddenly started working more jobs to support their families (though the labor market was messy, student work time didn't change significantly). It also didn't find that the changes were the same for everyone; the "low-income" group was hit much harder.

The Takeaway
The authors suggest that while students didn't stop trying, the shift to virtual learning meant they lost a lot of structured class time, especially those from lower-income backgrounds. They didn't replace that time with more studying; they replaced it with more screen time. The paper suggests that to fix this, schools might need to bring back more in-person learning and find ways to get students off screens and back together, because the "learning loss" might be real, and the gap between rich and poor students might have gotten wider.

The data is based on real surveys of thousands of students, so these aren't just guesses; they are measured trends. However, the authors note that because the sample of low-income students during the pandemic was smaller, we should be careful not to over-interpret every tiny detail, but the big picture of "less class, more screens, and a bigger gap for poor students" is clear.

Technical Summary: Zooming In or Left Out?: How Covid Changed Schooling, Screen, and Social Time for High School and College Students by Household Income

Problem Statement
The abrupt shift to virtual and hybrid learning models during the Covid-19 pandemic fundamentally altered the daily time allocation of students. While existing literature has documented learning loss and academic performance declines, there is a gap in understanding the specific mechanisms of time reallocation—specifically, how students shifted time between skill-building activities (education, work) and leisure (screen time, socialization, sleep). Furthermore, prior research has not comprehensively analyzed whether these shifts were heterogeneous across household income levels, despite evidence suggesting lower-income students faced greater barriers to remote learning (e.g., lack of internet access). This paper addresses the need to quantify changes in time use for high school and live-at-home college students before and during the pandemic, with a specific focus on income-based disparities.

Methodology
The authors utilize data from the American Time Use Survey (ATUS) covering the period from 2017 through 2022. The sample is restricted to high school students and college students living at home, excluding those over age 25, resulting in a final sample of 2,495 observations (1,368 high school and 1,127 college students).

The study employs a two-pronged empirical strategy to analyze time allocation:

Extensive Margin (Participation): Probit models are estimated to determine the likelihood of a student engaging in a specific activity (education, work, leisure, screen time, sleep) on a given diary day.
Intensive Margin (Duration): Ordinary Least Squares (OLS) models are estimated for the subset of students who participated in an activity, predicting the minutes spent on that activity.

The analysis controls for demographic characteristics (age, race, sex, school status), household characteristics (income, with a specific "low-income" indicator defined as households earning $\le$ $34,999), and location variables (Census region, metro fixed effects). The "pandemic" period is defined as mid-May 2020 through December 2022, excluding the data collection gap between March 18 and May 9, 2020. The authors also exploit the hierarchical nature of ATUS data to analyze the location of educational activities (school, home, elsewhere) and, for college students, the social context (alone, with parents, etc.).

Key Contributions
This paper is the first to use ATUS data to provide comprehensive, income-disaggregated estimates of the pandemic's effect on student time allocation. Key contributions include:

Decomposition of Educational Time: The study separates educational time into "class time" and "homework time," revealing that declines were driven almost entirely by reduced class time rather than changes in homework.
Income Heterogeneity: The analysis explicitly tests for differential effects between low-income and mid-to-high-income households, finding that low-income students experienced significantly larger reductions in educational time and larger increases in screen time.
Location and Context Analysis: By utilizing "where" and "who with" data, the paper quantifies the shift of education from school to home and the increase in solitary learning, particularly for college students.

Results

Overall Time Use: Relative to the pre-pandemic period, students reduced total educational time by approximately 33 minutes per day. This reduction was driven by a 23-minute decline in class time, while homework time remained statistically unchanged. Conversely, leisure time increased by 25 minutes, and screen time increased by 40 minutes. Socialization likelihood decreased by 11 percentage points.
Income Disparities: The impact was highly heterogeneous. Low-income students (<$35k) saw a 75-minute decline in total educational time compared to a 20-minute decline for mid-to-high-income students. Specifically, low-income students reduced class time by 75 minutes per day, whereas higher-income students reduced it by only 10 minutes. Low-income students also increased screen time by 53 minutes relative to their peers and decreased exercise time.
Temporal Dynamics: The most severe disruptions occurred during the early pandemic period (May 2020–March 2021). While some metrics (like total education time) began to normalize in the later pandemic period, screen time increases and reductions in socialization persisted.
Subsample Differences: High school students experienced more drastic changes in the early pandemic period (e.g., a 76-minute reduction in education time) compared to college students, though both groups saw persistent increases in screen time.
Location Shifts: Educational time shifted significantly from school to home. For college students, the likelihood of engaging in education alone increased by 11.6 percentage points, and time spent in solitary education increased by 65 minutes.

Significance and Claims
The authors claim that their findings suggest a mechanism for pandemic-era learning loss: the reduction in instructional class time was not compensated for by increased homework or other productive activities. Instead, this time was largely substituted with leisure and screen time.

The paper argues that these shifts pose significant risks to human capital accumulation, particularly for low-income students who faced the largest declines in educational time and the largest increases in screen time. The authors suggest that post-pandemic policy interventions should focus on:

Providing additional academic supports to address learning loss.
Reducing reliance on online/asynchronous assignments in favor of in-person learning.
Creating opportunities for in-person socialization and extracurriculars to counteract isolation.
Targeting support specifically toward low-income students, whose time use patterns did not fully return to pre-pandemic norms by the end of the study period.

The authors maintain a modest tone regarding causality, noting that while they control for observables, unobserved factors may influence time allocation. They also acknowledge limitations, such as the inability to link time use directly to academic outcomes in the ATUS and the lack of "who with" data for high school students.

More like this