1. Questions

The Covid-19 pandemic has impacted all aspects of life, most noticeably that of transitioning to work from home (WFH) for hundreds of millions of workers worldwide (Soares, Bonnet, and Berg 2021). By eliminating commute trips, WFH can contribute to transportation efficiency and environmental quality. Understanding the impact of the current mandated and prolonged WFH situation on work productivity (Kazekami 2020; Nakrošienė, Bučiūnienė, and Goštautaitė 2019; Pigini and Staffolani 2019) might inform commute trip reduction policies after the pandemic subsides. Complementing previous research (Shi et al. 2020), this study explores the impact of WFH on teleworkers’ productivity over time. We ask:

1. Did perceived work productivity change as workers continued to WFH over time?

2. Did the relationships between socioeconomic status (SES), lifestyle, and perceived teleworking productivity differ between spring and fall 2020?

2. Methods

Study and Survey Design

Two waves of a cross-sectional survey were conducted in the spring (April-June) and fall (October-November) of 2020. The data were pooled for analysis to identify differences over time, but the results do not infer causal relations. Consistent questions asked respondents about changes in their work productivity, daily routines, and mental well-being since transitioning to WFH per state restrictions following Covid-19. Sampling targeted a population of adults 18 years old or older, living in the Seattle region (King, Kitsap, Pierce, and Snohomish Counties). The survey provided no financial incentive for participation. Both surveys were distributed online through professional email lists, including public agencies, non-governmental organizations, community groups, universities and colleges. Protocols were approved by the University of Washington Internal Review Board.

Survey respondents came from 87% of the region ZIP codes. They totaled 4,151 with 3,402 responding to the spring survey and 749 to the fall. The present study included the 2,678 (65%) who had shifted from working away to WFH since March 2020. Respondents from the two survey waves shared similar characteristics (Table 1). Compared to the region’s population (US Census Bureau 2019), more respondents identified as female (69% in spring and 64% in fall vs. 50% in the region), had a graduate degree (52% and 54% vs. 17% region), and belonged to households with incomes above $150,000 (34% and 38% vs. 27% region). There were fewer 18-to-29-year-old adults (18% and 13% vs. 22% region).

Variables

Analyses were conducted at the respondent level. The outcome variable was self-reported change in productivity since respondents started WFH. Three response options were provided: no change, a decrease, or an increase in productivity. Productivity was treated as an ordinal variable with decrease in productivity as the reference.

To explain changes in teleworking productivity over time, we used factors identified as significantly associated with teleworking productivity in our previous study (Shi et al. 2020). Factors were organized into four domains: socioeconomic status (SES), pre-pandemic commute trip mode and duration, lifestyle changes, and mental wellbeing (Table 1). All respondents reported commute duration for each mode (walking, biking, transit, SOV, and HOV), with four ordinal levels (0-do not use, 1- <30min, 2- 31-60min, 3- >1 hour). Depression was evaluated using two questions from the Patient Health Questionnaire (Kroenke, Spitzer, and Williams 2003): how often respondents felt depressed or had little interest in doing things on a 4-point Likert scale (0 = Not at all; 1 = Several days; 2 = More than half the days; 3 = Nearly every day). The depression measure was indexed by summing all scaled question results, with higher scores indicating higher levels of perceived depression (range = 0 to 6). The measures have been tested for validity in previous studies (Cohen-Cline, Turkheimer, and Duncan 2015).

Change over the time of the pandemic was captured by adding a dummy variable that identified respondents from the spring and fall survey waves.

Statistical Analysis

Ordered logit models served to assess factors associated with self-reported changes in productivity. We first tested univariate models for all hypothesized predictors, with survey wave dummy added to examine the effect of time. A base model was estimated with all variables significant in the univariate models as well as the survey wave dummy. We then tested interaction terms between each factor and the survey wave dummy and only reported those that were significant at p<0.05 level. We conducted brant tests on the proportional odds assumption and results showed the assumptions were not violated (Brant 1990).

3. Findings

In our sample, productivity appeared to increase over the course of the pandemic. More respondents reported experiencing no change in productivity in the fall than in the spring (44% vs. 38%) or being more productive (26% vs. 24%) (Table 1). After controlling for SES, pre-pandemic commute trip duration, lifestyle changes, and mental wellbeing status, fall respondents were 48% more likely than spring respondents to report experiencing no change or an increase in productivity (Table 2 Model 1). This suggested that people have adapted to teleworking gradually and positively as the pandemic subsisted.

Income and living arrangement had a moderating effect on the capacity to adapt to teleworking. Change in self-reported productivity was different for people coming from different household income brackets (Table 2 Model 2). A significantly higher proportion of people with household incomes above $150,000 reported being more productive compared to spring respondents (Table 3, Figure 1). Therefore, people who were more financially advantaged appeared to be better able to adapt to teleworking than those with lower incomes. A likely explanation is that higher income translates into greater access to resources and the ability to make timely adjustments, such as enhancing internet infrastructure, having a spare room for work, having satisfactory child care, etc. (Cox et al. 2020; Prime, Wade, and Browne 2020), all conditions that may lead to greater productivity.

Figure 1.Modeled differences in the effect of income between the spring and fall survey.

Living arrangement had a moderating effect on productivity change. Compared to other living arrangements, people living with roommates and friends experienced a decrease in productivity (Table 2 Model 3). However, the effect was no longer significant after considering that of income (Appendix Table A1 Model 4), likely because household income and living arrangement were correlated, where people living with friends and relatives tended to also have household incomes below $150,000 (Appendix Table A2).

Other factors found to influence teleworking productivity in the spring survey (Shi et al. 2020) (age, pre-pandemic commuting patterns, lifestyle changes, and mental wellbeing status) remained significant after considering the temporal effect. The pooled data structure used did not infer causal relations, however, findings strongly suggested that many higher income workers could continue to WFH past the pandemic while remaining productive, and could potentially reduce the number of future commute trips. Extra effort should be made to support those with fewer resources to telework productively if they choose to.

Acknowledgements

This work has been funded in part by the US Department of Transportation’s University Transportation Center program through the Pacific Northwest Regional University Transportation Center (PacTrans) and by the Puget Sound Regional Council (PSRC). Views expressed in the paper do not represent those of the sponsors and the authors are responsible for all errors that may exist.

Appendix