Remote work matters, but culture is the elephant in the room

Debates among managers and organisations about when and how to implement remote work are ongoing. This column provides empirical evidence on rates of remote work – across industries, occupations, and geographies over time – as well as the relative importance of remote work to job satisfaction and employee turnover. The findings suggest that the freedom to work remotely is less significant to employees than the corporate culture of an organisation. Feeling appreciated at work is a particularly strong predictor of job satisfaction and retention.

There is now a mountain of research on remote work – including an annual Remote Work Conference held at Stanford University – and at least one lesson is clear: there is vast heterogeneity in not only the rates of remote work but also the experiences of remote workers. Stanton et al. (2020) showed early on that small businesses were less likely to transition to remote work; Zarate et al. (2022) documented these patterns across 27 countries, which vary markedly; and Shah et al. (2024) demonstrate that remote work is here to stay, according to new managerial surveys, although firms vary in their stance on the degree and implementation.

Our recent paper (Makridis and Schloetzer, 2024a) uses proprietary data from PayScale, a crowdsourcing human capital-oriented company, to provide new empirical evidence on the rates of remote work across industries, occupations, and geographies over time, and the relative importance of remote work in explaining differences in employee job satisfaction and turnover rates.

PayScale operates on a ‘give to get’ model – that is, employees share information about their own employment situations to obtain information from the broader PayScale database about, for instance, a predicted market wage. Employees can use these numbers in salary negotiations or for making career decisions.

Crucially, our data – over 150,000 responses from May 2023 to December 2024 – allows us to control for employees’ perceptions about five dimensions of workplace practices, including perceptions about appreciation, communication, development and training opportunities, pay transparency, and managerial quality. Respondents rate these factors on a scale from one to five, which has its limitations but nonetheless provides useful variation.

We also see employees’ answers to a question about the frequency of remote work – consisting of “always”, “mostly”, “sometimes”, and “never” – more clearly than the traditional binary response.

Let us summarise a few of our main descriptive results:

• Time: remote work saw a sharp increase during the early COVID-19 period, with fully remote work rising from less than 10% in early 2020 to about 18% by early 2021. Hybrid arrangements also increased, with total remote work adoption reaching around 45% during the same period. These levels have remained stable since 2021, with nearly 20% of employees always working remotely, and 45–50% engaged in some form of hybrid work through late 2023.
• Industries: remote work adoption varies significantly by industry (Figure 1, top panel). Finance, insurance, information, and professional services maintain the highest rates, with over 60–70% of employees engaged in either fully remote or hybrid work. Between 2020 and 2023, industries like information and professional services experienced substantial, double-digit percentage point increases in remote work arrangements.

Figure 1 Changes in remote work, by industry and occupation

Notes: The figure plots the change in the share of respondents in different remote arrangements from 2020 to 2023 in percentage points: “always WFH,” “mostly WFH,” and “sometimes WFH,” by two-digit NAICS industry (top panel) and two-digit SOC occupation (bottom panel).
Sources: Payscale (May 2020-Dec 2023).

• Occupations: remote work patterns vary by occupation (Figure 1, bottom panel). Legal, healthcare support, and protective service roles showed increases in fully remote work, while occupations like sales, education, and construction experienced declines due to increased in-person interactions. There have been large, persistent gains in computer/math occupations, with significant growth in both fully remote and hybrid arrangements.
• Geography: state-level patterns reveal geographic variation in remote work adoption. While 13 states reported increases in remote work between 2022 and 2023, others saw declines. California, Texas, and New York consistently reported high rates of remote work, with state differences influenced by industry concentration and economic factors.

Next, we asked: does the value of remote work matter more during a time of crisis? We use statistical models to relate employee job satisfaction and intention to leave a company as measures of remote work, controlling for our measures of workplace practices, occupations, and employee demographics. While the relationship between remote work and job satisfaction is hard to causally estimate, our strategy controls directly for the types of factors that would typically go unobserved, such as workplace practices that are correlated with both wages and employee engagement. We realise it is imperfect, but we see our approach as a step in the right direction.

Overall, while there were positive correlations between remote work (particularly hybrid arrangements) and job satisfaction during 2020–2021, these associations become statistically insignificant as time passes, particularly after accounting for broader workplace practices (Figure 2). It appears these other practices, not remote work itself, drive employee satisfaction.

Figure 2 Association between remote work and job satisfaction over time

Notes: The figures report the coefficients associated with regressions of standardised job satisfaction on an indicator for always working remotely (Always Remote) and mostly/sometimes working remotely (Hybrid) under various specifications by month. All regressions control for logged total cash compensation (including bonuses, tips, and commission), worker demographics: age, male, race (White, Black, Hispanic, Missing), education (associates, bachelors, Ph.D., high school, masters), and years of work experience. Workplace characteristics include standardized indices of pay transparency, perceived fairness of pay, intra-firm communication, development/training opportunities, feeling appreciated at work, and managerial relationships. Standard errors are heteroskedasticity-robust.
Sources: Payscale (May 2020-Dec 2023).

If not remote work, then what is the most influential workplace attribute in our analyses of job satisfaction and intention to leave a company? We find that feeling appreciated at work is a strong predictor of both job satisfaction and retention. This factor consistently shows strong, positive associations with satisfaction and reduced turnover intentions across the 2020–2023 period. Remote work appears less influential in explaining these outcomes when compared to workplace attributes that contribute to the corporate culture of an organisation. These results are consistent with and part of our broader paper (Makridis and Schloetzer 2024b) on the relative importance of remote work in explaining job satisfaction and employee retention.

Our paper builds on studies of the incidence of remote work and its effects. The increasing consensus that hybrid work combines the best of both worlds – at least for certain jobs that can be done remotely – has been supported by several randomised controlled trials, such as Choudhury et al. (2024) and Bloom et al. (2024). Consistent with these, we show in Makridis and Schloetzer (2024b) that hybrid work – specifically ‘sometimes’ remote – is statistically associated with higher levels of job satisfaction, whereas ‘mostly’ and ‘always’ remote are not, after controlling for workplace attributes that contribute to the corporate culture of an organisation. In this sense, we view our results from observational data as complementing the narrower, albeit better-identified, results from randomised controlled trials.

The ongoing debate among managers and organisations is when and how to implement remote work. Our evidence suggests the implementation will need to vary by industry, occupation, and other aspects of the employee’s work environment that shape corporate culture.

Source : VOXeu