Workplace and lifestyle heterogeneity in subjective wellbeing: Latent Class Analysis of UK Time Use Survey before and during COVID-19

Data source

UKTUS (2015) is a large-scale household survey that provides data on how people spend their time. Time diaries record activity sequences, corresponding locations, as well as the respondent’s level of SWB throughout the day (Gershuny & Sullivan, 2016).

A novel longitudinal SWB data set was compiled by combining the pre-pandemic UKTUS 2015 with four additional waves of population-representative (quota sample) time use diary data collected during the pandemic. These waves were collected online in 1) 2016; 2) May-June 2020 at the peak of first UK COVID-19 lockdown; 3) August 2020 following the relaxation of social restrictions 4) November 2020 during the second lockdown (Gershuny & Sullivan, 2021). The latter three waves collected throughout 2020 were classified in our analysis as ‘during COVID’, while the ‘pre-COVID’ data consisted of UKTUS 2015 and the first online wave collected in 2016. All data was obtained from the UK Data Service.

The study sample comprised all those aged 16-64, in full employment, with sufficient data on activity-level episodic SWB, and a non-zero survey weight. Our analysis included 3855 individuals, each contributing 24 hours of time use schedule during a weekday considered typical to them. A total of 369,394 activity-location-SWB bundles were analysed (see sample distribution in Appendix 3).

Measures

To understand the impact of the COVID-19 pandemic on workers’ time use, we first separated all samples by the reported location of paid work into three distinct working modes: 1) homeworkers; 2) commuters; 3) hybrid. Homeworkers engaged in paid work exclusively at home during the diary day; commuters engaged in paid work exclusively at the workplace; whilst hybrid workers reported paid work activities both at home and in the workplace. Such information had been collected consistently across survey periods. Other sociodemographic characteristics considered were income, educational attainment, occupation classification, gender, age, and marital status.

The breakdown of activity recorded in UKTUS was highly detailed, with 144 distinct activities. To reduce the complexity, we followed literature (Ines & Tipoe, 2021) and aggregated them into five broad activity types: 1) personal; 2) paid work; 3) non-paid work; 4) leisure; 5) transport (see breakdown in Appendix 2). Intraday time use pattern was also aggregated from 10-minute intervals as in the original data into 24 hourly categorical variables, based on the broad activity type the respondent was engaged in for the majority of the given hour.

SWB was measured with an episodic enjoyment variable in the original data, reporting the instantaneous utility of the respondent on a seven-point ordinal Linkert scale. Our analysis found that the distribution of ordinal responses was heavily right skewed, where the proportion of the top two scores accounted for roughly half of overall responses (see distribution histogram in Appendix 4). As such, it was transformed into a binary variable with enjoyment six and above defined as high SWB, five and below defined as average-to-low SWB. The use of binary SWB measurement had two advantages. Firstly, the original enjoyment measurement was purely ordinal, hence hardly comparable across years. Using a binary variable could, to some extent, enhance the comparability of enjoyment measurement between survey periods. It enabled us to focus on identifying determinants that may lead to higher-than-average enjoyment, as opposed to factors that may lead to a general increase of enjoyment. Secondly, it reduced the computing load for model estimation.

Statistical analysis

We estimated latent class models using robust maximum likelihood estimation with Newton-Raphson stepping mechanism, repeated for one to six latent classes. The 24 hourly categorical variables indicating activity type were used as explanatory variables. To help with model convergence, we used a two-step integration process: Laplacian approximation was used in the first round of integration, the results matrix was then used as starting value for the more accurate mean-variance adaptive Gauss-Hermite quadrature integration method. The final class size within each working mode was determined using the Akaike’s information criterion (AIC) and Bayesian information criterion (BIC) values (see model fit estimations in Appendix 5), average latent class probabilities, and substantive interpretation of the classes identified. Note that the latent class analysis (LCA) was conducted using pooled samples from both pre- and during COVID. After LCA, the samples were then separated into pre- and during COVID groups according to the survey date.

Outcomes

A total of ten lifestyles were identified across the three working modes: three within homeworkers; four within commuters; and three within hybrid workers. Figure 1 shows the distinct intraday time use patterns of different lifestyles and the corresponding intraday SWB patterns. Figure 2 provides the pen portrait for each lifestyle, which features a combination of generic lifestyles and reported working modes. We define lifestyles primarily based on paid work schedules because of its dominant role in lifestyle choice and identity (Antilla et al., 2015). Sociodemographic characteristics of each lifestyle are summarised in Appendix 1.

• Download figure
• Open in new tab

Figure 1: Intraday patterns of time use (top two rows) and SWB (bottom row) *

* The x-axis represents 24 hours of the day for all charts. For time use patterns, the y-axis is the percentage proportion of the samples engaged in each of the five time use categories summing to 100%. For SWB patterns, the y-axis is the percentage proportion of the samples reporting high SWB. Usual working hours (6am-6pm) are highlighted for SWB charts.

• Download figure
• Open in new tab

Figure 2:

Lifestyle pen portraits

Time use and SWB patterns outside usual working hours (before 6am or after 6pm)

SWB declined uniformly across lifestyles outside usual working hours when most people engaged in leisure or sleeping. The uniform SWB decline corroborated the deteriorated mental health in the UK during COVID-19, however, time patterns during these hours remained virtually unchanged pre- and during COVID, which suggested that the SWB decline outside usual working hours seemed uncorrelated with time use and may be attributed to general anxiety and potentially blurred boundary between work and life associated with flexible working. Before COVID-19, a rapid rebound of SWB was observed outside of usual working hours consistently across lifestyles. During COVID-19, however, the rate of SWB rebound in the early evening flattened, particularly for homeworkers and hybrid workers who tend to be more susceptible to the effect of blurred boundary between work and life.

Time use and SWB patterns within usual working hours (6am-6pm)

The negative SWB impact of paid work activities was reflected in the majority of lifestyles. Enjoyment tended to plummet as usual working hour started, with slight uptick in mid-day which coincided with usual lunch breaks, and returned to pre-work levels after work. Lifestyles with more concentrated paid work periods such as EH, CC, and CH would experience greater intraday SWB fluctuations.

The direction of COVID-19 impact differed across working modes: average SWB of homeworkers decreased significantly; commuters remained relatively stable; by contrast, some hybrid lifestyles experienced a positive impact. SWB changes for each working mode are further discussed below.

Homeworker, in particularly Highflying Homeworker (HH) and Domestic Homeworker (DH), had higher SWB pre-pandemic, which was attributed to the less negative SWB impact from paid work during usual working hours compared to commuting and hybrid lifestyles. During the pandemic, hours of concentrated paid work became significantly more detrimental to SWB for EH lifestyle. In addition, the uptick of SWB around mid-day, which was prevalent pre-pandemic, disappeared for HH and DH during the pandemic.

Commuter lifestyles maintained similar levels of SWB during usual working hours before and during the pandemic. Workers who were able to commute to work during the pandemic were likely essential workers. Nightshift Commuter (NC) was the only commuting lifestyle experiencing a notable downward shift in usual-working-hour SWB during COVID-19. Increased workload and anxiety linked to the pandemic and job security concerns might explain the SWB decline.

Hybrid worker lifestyles, CH and Long-hour Hybrid (LH), experienced an increase of SWB within usual working hours during COVID-19. In contrast to the general adverse SWB impact observed for other working modes, this finding suggested the optionality of having dedicated workplace and the ability to continue to travel during the pandemic seemed to bring SWB betterment. We also found that the SWB increase may stem from enjoyment boost associated with some activity types but not all, which will be investigated in the next section. Other plausible factors that may contribute to the SWB increase may include social interaction opportunities associated with the workplace, the ability to travel outside during the pandemic per se, and increased desirability of the hybrid working mode from lived experience of mandated homeworking.

Heterogeneity of SWB impact across latent lifestyles

To better understand the heterogeneity of SWB impact across lifestyles, Figure 3 shows average SWB changes from pre-to during COVID-19 within usual working hours (6am-6pm) by activity type.

• Download figure
• Open in new tab

Figure 3:

Difference in mean SWB within usual working hours (6am-6pm) by activity type

In terms of the direction of COVID-19 impact on SWB, paid work average decreased for all lifestyles except Long-distance Commuters (LC) and hybrid lifestyles CH and LH. SWB decline from leisure activities was observed for almost all lifestyles but was particularly pronounced for LC. Enjoyment associated with travel activities (mainly as derived demand) has increased for most lifestyle groups except the Domestic Homeworker (DH), who spent more time on transport than other homeworkers and on par with some commuters. Overall, paid work appeared to be the main driver of directional changes, and non-paid work further contributes as its directional changes are often consistent with paid work.

In terms of the magnitude of COVID-19 impact, five out of six large SWB changes (>= 0.5 or <=-0.5) occurred in late and dispersed lifestyles: HH, LC, and LH. This is contrasted by the relatively low magnitude of impact for the non-work dominated lifestyles: DH, HC, NC, and MH. These lifestyles tended to be less constrained by established work-time regime, enabling a higher degree of flexibility for adjusting their time use to mitigate the pandemic impact.

Findings statement

Our analysis shows COVID-19 impact on SWB is heterogeneous in direction and magnitude for different working modes and lifestyles. The identification of latent lifestyles and the comparison of SWB impact across lifestyles shed a new light on the conceptualisation of ‘flexibility’ from a time use perspective. Specifically, flexibility could be conceptualised over two dimensions, spatial flexibility and temporal flexibility (Antilla et al., 2015).

Spatial flexibility denotes the optionality on place of work and the suitability of the space chosen. When work-from-home became mandatory, many commuters (CC lifestyle) would, perceivably, switch hastily into the dominant EH lifestyle. The domestic space of those new homeworkers, however, may not be suitable for homeworking. Compared with commuters and homeworkers, hybrid working was the only working mode with prior experience of working effectively both at home and the usual workplace, and as such retained spatial flexibility during the pandemic. The spatial flexibility may explain the increase of SWB of hybrid workers within usual working hours during COVID-19.

Temporal flexibility refers to the flexibility in deciding when to work, for how long and the time split between paid work and other activities. Temporal flexibility is often overlooked as spatial flexibility dominates discussion around flexible working practices. Empirically, more than half of the samples in each working mode adopted an early-concentrated working lifestyle (EH, CC, and CH), and a majority of the perceived “flexible workers” were still constrained to the 8-hour workday model, which revealed a highly inflexible temporal structure in the labour market. Our study found that workers of HH, LC, and LH lifestyles tended to spread out their paid work activities leading to less time for other activity types, particularly non-paid work, which is in line with the findings from the Australian TUS (Craig & Powell, 2011). Nonetheless, temporal flexibility seemed to come at a price – average income of non-work dominated lifestyles, particularly DH, HC, and NC, tended to be lower than that of other lifestyles in the same working mode.

COVID-19 brought about considerable changes to the spatial flexibility of homeworkers, and their decrease in SWB was associated with diminishing spatial flexibility during lockdown; hybrid workers maintained their spatial flexibility and experienced in increase in SWB during usual working hours. Higher degree of temporal flexibility was associated with the higher SWB resilience of non-work dominated lifestyles; those who worked long and dispersed hours were more sensitive to COVID-19 SWB impacts in both positive and negative directions.

Implication for key stakeholders

The argument for flexibilisation of work relies heavily on its implicit benefits for workers’ wellbeing. Spatial focus on flexible working (i.e., home and hybrid working) became highly pertinent during COVID-19, but the temporal dimension of work flexibility had been overlooked. The 4-day work week proposal is gaining popularity in certain knowledge intensive sectors, and in regions looking to attract highly skilled workers (Whillans & Lockhart, 2021). However, the unintended consequence might be intensification of work on the intraday basis. The resulting lifestyle change will likely be a shift from early-concentrated to long-dispersed working schedule, regardless of working mode. Preparing for the future of work, giving workers more choice of working modes, providing wider support for lifestyle adaptation and transitions, and maintaining the important social function of work should be prioritised.

Strengths and weaknesses of the study

The main strength of the study is our novel time use approach, allowing us to identify latent but distinct lifestyles within working modes. Our research is timely and provides insights from the SWB perspective when the future of work is rapidly changing. The richness of the multi-wave time use data allows us to visualise intraday patterns in an interpretable way, and reveal distinguishable differences between working modes, between lifestyles, before and during the pandemic. The proposed method for identifying latent lifestyles provides a viable strategy for tackling the high dimensionality issue associated with time use data.

A key limitation of the study is the measurement issue of SWB. As per the original data, each activity episode had a single SWB variable. The measurement weakness is twofold: firstly, we had to transform the ordinal results into a binary SWB outcome, increasing interpretability but trading off some sensitivity; secondly, the relationship between instantaneous utility and the long-term SWB is not yet clear. A more comprehensive set of SWB questions are required to better understand the impact mechanisms of time use on wellbeing. A supplemental questionnaire on long-term mental wellbeing of the respondents will allow future studies to understand the long-term mental health implications based on lifestyle differences. Furthermore, the data from the UKTUS are population-representative but nevertheless cross-sectional. Higher quality longitudinal data, such as a cohort study may be required to better quantify the relationship between lifestyles and SWB.

Future directions

To better quantify the relationship between spatio-temporal flexibility and wellbeing, a dedicated wellbeing module will be necessary for future TUS collections in the UK and beyond. Standardised TUS data is collected in more than 30 countries, currently only the US and UK incorporate activity-episode level SWB questions. If more countries incorporate an ordinal SWB variable into their national time use surveys, a large-scale international comparison could be used as benchmark for understanding of the changing future of work from an occupational health perspective.

Funding and competing interest

This study received no external funding. The authors have declared that no competing interests exist.

Research ethics approval

Ethics approval for this study was not required. All data involved is openly available to the public before the initiation of the study. The public data is hosted by UK Data Service, and can be accessed via the DOIs below: http://doi.org/10.5255/UKDA-SN-8128-1

http://doi.org/10.5255/UKDA-SN-8741-3