Key research papers on remote work and working from home

How the literature is organised

The pre-pandemic literature is small and well synthesised. Bailey and Kurland (2002) reviewed telework research across transportation, law, sociology and organisational studies and concluded that little clear evidence existed that telework increases job satisfaction and productivity, as it was often asserted to do. Gajendran and Harrison (2007) meta-analysed 46 studies covering 12,883 employees and found small but mainly beneficial effects on autonomy and work-family conflict, and on job satisfaction, performance, turnover intent and role stress, largely mediated by perceived autonomy; only high-intensity telecommuting, more than 2.5 days a week, harmed relationships with co-workers. Allen, Golden and Shockley (2015) assessed the state of the science and made the definitional problem central to it. Bloom et al. (2015) supplied the experiment. Around these sit studies of mechanisms: boundary management (Kossek, Lautsch and Eaton, 2006), professional isolation (Golden, Veiga and Dino, 2008, who found it lowered job performance among 261 teleworkers, with the effect growing with time spent teleworking and shrinking with face-to-face contact) and work intensification (Felstead and Henseke, 2017, who found remote working associated with higher commitment, satisfaction and job-related wellbeing, at the cost of work intensification and a greater inability to switch off).

Then came 2020. The snapshot behind this page counts 156 titles for 2019 and 1,891 for 2021, and the most-cited recent papers are studies of employees sent home during lockdowns.

Main debates

The productivity question has moved from "does it work?" to "for whom, and for which tasks?". Anakpo, Nqwayibana and Mishi (2023) reviewed 26 studies and found the answer depends on the nature of the work, the employer, the industry and the home setting, with most reporting a positive effect and a few none or a negative one. Individually measurable work holds up; collaboration is where the evidence turns. Yang et al. (2021) used emails, calendars, instant messages, calls and workweek hours for 61,182 US Microsoft employees over the first six months of 2020, treating the firm-wide shift as a natural experiment, and found that it made the collaboration network more static and siloed with fewer bridges between disparate parts, and shifted communication from synchronous to asynchronous — which, they argue, may make it harder to acquire and share new information across the network.

Careers and health are the other live issues. Bloom et al. (2015) found home workers' promotion rate conditional on performance fell, even as performance rose. On health, the honest summary is that we do not know: Lunde et al. (2022) found only 14 relevant studies covering 28 outcomes and graded the whole body of evidence low or very low, concluding that the absence of studies on many important outcomes is itself the finding.

Where recent work is heading

Hybrid arrangements are the present focus. Barrero, Bloom and Davis (2023) report that full days worked at home accounted for 28% of paid workdays among Americans aged 20 to 64 as of mid-2023, about four times the 2019 rate, and set out why the shift has endured rather than reverting. Leonardi, Parker and Shen (2024) refract remote work into four kinds of distance — psychological, temporal, technological and structural — which is a useful framework if your study is about something other than location. McPhail et al. (2024) scope the effects on people, productivity and the planet, and note that most of what has been published since 2020 is data-driven rather than theory-driven. That gap is the opening for a thesis: a comparison of fully remote, hybrid and on-site employees in one organisation or sector, grounded in an existing theory rather than in the data you happen to have.