Satisfaction of scientists during the COVID-19 pandemic lockdown

The paper investigates how the COVID-19 pandemic lockdown affected scientists’ life satisfaction in Austria, Germany, and Switzerland. After the WHO declared COVID-19 a pandemic on March 11, 2020, countries implemented social distancing, school and business closures, and a shift to home working. Public debate has emphasized negative impacts (mental health, domestic violence, educational disruptions). The authors leverage a survey that was in the field before and after the March 13–16, 2020 lockdowns, enabling a before–after quasi-experimental comparison. They focus on scientists, a relatively homogeneous professional group with comparatively secure employment and stable workloads, to assess whether the sudden transition to home office altered general and domain-specific satisfaction (health, work, work-life balance, leisure). The study aims to provide early empirical evidence and mechanisms for observed changes, with cautious consideration of generalizability beyond academia.

The article does not present a standalone literature review section but situates the work within several strands of prior research. It references concerns about negative social and health outcomes from lockdowns (e.g., increased depression, care burdens, educational inequalities) and ethical bases for social distancing. It draws on social physics, psychology, and sociology literature indicating that crowding and dense environments (e.g., commuting, large gatherings) reduce individual well-being. It also engages with gendered impacts of the pandemic, noting reports that mothers may experience disproportionate burdens, while highlighting evidence that dual home-office households can distribute childcare more equally. Comparative context is provided against UK panel evidence of decreased life satisfaction and increased depressive symptoms during the pandemic, with differences attributed to broader population coverage, economic vulnerability, and severity of the health crisis compared to the DACH region. Theoretical framing includes the notion of modern life’s acceleration and the potential for deceleration (via home-office and reduced crowding) to improve satisfaction.

Design and setting: Quasi-experimental before–after design using daily invitations and responses from a representative online survey of scientists (PhD students, postdocs, professors, and other academic staff) at higher education institutions in Austria, Germany, and Switzerland. Ethics approval: University of Zurich Ethics Commission (Approval 18.8.7). Field period: February to April 2020. Lockdown timing: March 13–16, 2020 across the three countries. Sampling and recruitment: Contact data were compiled via institutional webpages in collaboration with the German Centre for Higher Education Research and Science Studies. Invitations were emailed directly to scientists (institutions were not informed). All scientists in Austria and Switzerland were invited; in Germany, a 50% random sample was invited. To avoid spam filtering, up to ~12,000 invitations were sent daily over 77 days, with up to two reminders for non-respondents. Overall, ~141,500 invited; 15,972 participated (11.3% response rate), averaging ~207 responses/day. Analytic sample: For lockdown comparisons, responses from the weekend of March 13–16 were excluded. Responses before March 13 were coded pre-lockdown; after March 16 as post-lockdown. Final analytic N=13,316 (Germany N=7,476; Austria N=2,151; Switzerland N=3,689). Sample characteristics: 43% female; 39% PhD students; 39% postdocs; 18.6% professors; 3.4% other. Analyses were restricted to staff at regular full universities and technical universities for cross-country comparability. Measures: Primary outcome was general life satisfaction (“All in all, how satisfied or unsatisfied are you currently with your life?”). Domain-specific satisfaction included health, work, work-life balance, and leisure. Original response scales ranged from -5 (totally unsatisfied) to +5 (totally satisfied) and were recoded to 0–10. Additional variables: demographics (gender, age), cohabitation with partner, parenthood (child aged 0–17), contract type (permanent vs non-permanent), working hours (FTE), country, and scientific field. The exact response time stamp was recorded. Analytical strategy: The post-lockdown indicator served as the treatment variable in a quasi-natural experiment framework. Descriptive visualization of 7-day moving averages over time was combined with naive two-sample t-tests comparing pre- vs post-lockdown averages for each satisfaction measure; Cohen’s d effect sizes were reported, including subgroup contrasts by academic status and parenthood with approximate permutation tests (10,000 reps) to compare effect sizes. Multivariate linear regression models estimated treatment effects controlling for gender, age, academic status, parenthood, cohabitation, contract type, working hours, country, and field. Interaction models included treatment-by-covariate interactions to assess heterogeneous effects across subgroups; models were estimated separately with and without interactions. Additionally, the distribution of workload across administration, teaching, and research (shares of total workload) was compared pre- vs post-lockdown using t-tests.

Descriptive trends and t-tests: - Across all five dimensions, average satisfaction was higher during lockdown than pre-lockdown. - General life satisfaction: 7.48 (pre) vs 7.67 (lockdown), t(12,786)=5.23, p<0.001. - Health: 7.53 vs 7.98, t(12,514)=11.08, p<0.001. - Work: 6.84 vs 7.15, t(12,711)=6.91, p<0.001. - Work-life balance: 6.56 vs 6.96, t(12,665)=8.16, p<0.001. - Leisure: 6.63 vs 6.94, t(12,565)=6.77, p<0.001. - Cohen’s d effect sizes by status showed small positive effects (roughly d=0.10–0.22) for doctoral students, postdocs, professors, and other staff across outcomes (mostly p<0.001), with some variation by domain. - Work-life balance effects differed by parenthood: no significant change among researchers with children aged 0–17 (d≈0.04, ns) versus a positive effect among those without children (d≈0.19, p<0.001); the difference between these effects was significant (permutation test p<0.001). Multivariate regression (controls included): - Estimated average increases associated with lockdown (unstandardized coefficients, 95% CI): - General life satisfaction: β=0.229 (0.159, 0.298), t(12,231)=6.439, p<0.001. - Work satisfaction: β=0.376 (0.290, 0.462), t(12,178)=8.567, p<0.001. - Work-life balance: β=0.401 (0.305, 0.497), t(12,148)=8.215, p<0.001. - Leisure: positive main effect reported; see table (β≈0.291; model constants and Ns reported). - Health: positive main effect reported; see table (β≈0.442; model constants and Ns reported). - Heterogeneity (interaction models): Most interaction terms were not significant, indicating broadly similar effects across subgroups. Notable exceptions: - Parenthood: Weaker work-life balance effect for researchers with children vs without (interaction β≈-0.482; 95% CI -0.733 to -0.232; t(12,114)=3.781; p<0.001). - Employment level (FTE): Higher FTE associated with a slightly weaker work-life balance effect (interaction β=-0.007; 95% CI -0.012 to -0.001; t(12,114)=2.462; p<0.02). - Field differences: - Work satisfaction: geologists (β=-0.646; 95% CI -1.126 to -0.167; t=2.642; p<0.01) and mathematicians (β=-0.703; 95% CI -1.197 to -0.210; t=2.793; p<0.01) showed weaker effects than humanities. - Work-life balance: economists showed a weaker effect (β=-0.519; 95% CI -0.994 to -0.045; t(12,114)=2.144; p<0.04) than humanities. - Leisure: geologists showed a weaker effect (β=-0.536; 95% CI -1.049 to -0.023; t(12,020)=2.046; p<0.05) than humanities. Workload composition: - Administrative workload share decreased by 6.9% (25.25% pre vs 23.51% post), t(12,379)=5.57, p<0.001. - Research workload share increased by 3.8% (48.87% vs 50.73%), t(12,657)=4.10, p<0.001. - Teaching workload share did not change significantly (27.18% vs 27.43%), t(12,247)=0.73, p=0.463.

The findings indicate that, contrary to widespread expectations of declines, scientists’ overall and domain-specific life satisfaction did not decrease during the initial COVID-19 lockdown; instead, small but significant increases occurred, particularly in health-, work-, and work-life-related satisfaction. The authors argue that the abrupt transition to home-office is a key mechanism: increased flexibility in organizing work and non-work activities, reduced daily crowding (e.g., commuting, large gatherings), and a generally slower pace of life plausibly improved well-being. These mechanisms align with prior evidence that dense environments reduce comfort and with theories of social acceleration and its discontents. Heterogeneity analyses show broadly similar positive effects across demographic and professional subgroups, with notable attenuation among parents (especially for work-life balance) and in certain fields (e.g., geology, mathematics, economics). The absence of an additional gender-specific decline among female scientists may reflect sample-specific contexts (highly educated dual-earner households with greater feasibility of home-office and more equitable childcare distribution). Comparisons with UK panel data showing declines in life satisfaction underscore contextual differences: the present study focuses on a professional group with stable employment and continued activity during lockdown in countries less severely affected in early 2020, whereas general-population samples include individuals experiencing job loss and greater economic hardship. Thus, the results are interpreted as evidence on the effects of transitioning to home-office within relatively secure knowledge-work contexts, with potential implications for similar occupations beyond academia.

This study leverages a quasi-experimental before–after design from a large survey of scientists in Austria, Germany, and Switzerland to show that life satisfaction—general and in domains of health, work, work-life balance, and leisure—slightly increased during the initial COVID-19 lockdown. Results are consistent across most subgroups, with attenuated benefits for parents’ work-life balance and for specific scientific fields. A plausible mechanism is the shift to home-office, enabling greater flexibility, reduced crowding, and a slower pace of life. These insights suggest that policies supporting flexible work and home-office arrangements may enhance well-being in occupations where remote work is feasible and tasks require sustained, concentrated effort. Future research should isolate the causal effects of home-office more precisely (e.g., within-person designs), assess long-term outcomes and sustainability, and evaluate productivity and coordination trade-offs relative to traditional office work. Exploring generalizability to other professional groups and contexts, and examining subgroup-specific challenges (e.g., parents’ work-life balance), are important next steps.

- Generalizability: The sample comprises scientists at universities in Austria, Germany, and Switzerland; findings may not extend to broader or more economically vulnerable populations or to occupations where home-office is infeasible. - Design: Quasi-experimental before–after comparisons with different respondents pre vs post; no within-person longitudinal tracking, leaving potential unobserved compositional differences despite controls. - Sample composition: Overrepresentation of professors and certain disciplines; underrepresentation of medical fields; addressed via controls but residual bias is possible. - Concurrent changes: Lockdown altered multiple life domains simultaneously; isolating the specific effect of home-office is challenging. - Short-term window: Results reflect the initial lockdown period; long-term satisfaction and productivity effects remain uncertain. - Measurement: Self-reported satisfaction and workload shares may be subject to reporting biases; the workload question may capture only gradual changes.