Lisanne Bergefurt ran a path analysis on Leesman data from 57,286 employees collected between March 2020 and May 2021 by a workplace consultancy for its own purposes and not for research. How well a setting supports the specific task relates to where an employee decides to work. Support for concentrated work and support for collaborative work both relate to perceived productivity, and they do so at home and at the office.
Zhihong Pang put 81 questions to people working from home during the pandemic and collected 1,460 responses from 35 countries, of which 1,137 were complete enough to analyse. The spatial layout of the home had a significant effect on wellbeing. Distractions and noise arising from having no personal workspace prevented productive work. A lack of scenic view, inadequate daylighting and poor acoustics were each reported as damaging to productivity and to the general experience of working at home.
Eunhwa Yang surveyed US knowledge workers through Amazon Mechanical Turk, collecting 1,651 responses and analysing 648. Perceived work-life balance improved during the pandemic against the period before it. The physical boundary between workplace and home went the other way and weakened. Workplace flexibility, the environmental conditions of the home office and organisational support were all positively associated with productivity, with satisfaction with working from home and with work-life balance.
Fifty-nine hybrid employees filled in an experience sampling survey on five consecutive workdays, giving Grace Elizabeth Catello 236 data points on physical activity, eating, relaxation and sleep. Each person appears on both home days and office days, so the confounding between how often someone works at home and what kind of person they are drops out. Vigorous physical activity and relaxation were both higher on home days. Eating behaviour and sleep did not differ by location at all.
Sonia Schifano tracked wellbeing across France, Italy, Germany, Spain and Sweden over four waves between May and November 2020, with over 8,000 people in the first wave and a little over 9,700 observations in the analysis. Bergefurt’s 57,286 is the larger sample; Schifano’s is the only panel. Working from home went with lower wellbeing than working elsewhere, and not working at all was worse than either. The penalty fell hardest on older workers, better-educated workers, parents of young children and people whose housing was more crowded, which is the one place in this literature where the dwelling itself enters the result.
Four of these five studies report associations and no single figure, which is why this sheet carries fewer numbers than the others. Bergefurt’s path analysis, Pang’s regression and Yang’s correlations all give coefficients rather than percentages, and only the abstracts of Pang and Yang are available here. Bergefurt worked with 57,286 responses, and a workplace consultancy collected them for its own purposes and not for research. Pang analysed 1,137 responses from 35 countries. Yang analysed 648 from one. Schifano ran a panel across five countries and four waves. Catello followed 59 people for five days, and that is the only study here that measured the same person in both places.