Teleworking is not going anywhere - here's why

COVID-19 triggered a surge in teleworking, with about 40–50% of the workforce in advanced economies working from home at the height of the pandemic (Taneja et al. 2021, Barrero et al. 2021a, Davis et al. 2021). In two recent and complementary studies, we source new data to answer three questions related to the surge in telework. First, will teleworking persist when pandemic-related mobility restrictions recede? Second, what is the impact on productivity and wellbeing as perceived by managers and workers? Third, how can policymakers and managers maximise the benefits and minimise the risks from more widespread teleworking?

In the first paper (Adrjan et al. 2021), we analyse developments in online job postings that advertise telework on the global job site Indeed in 20 countries over the past two years.¹ In the second paper (Criscuolo et al. 2021), we report the results from an OECD survey on teleworking of both workers and managers that we implemented in 25 countries.

Indeed retrieves job postings from a variety of sources, including direct postings on its proprietary platform as well as thousands of online job boards, career sites, and recruiter listings. In the 20 countries covered by our study, Indeed job postings represent the near-universe of online job offers. Using state-of-the-art text analysis algorithms, we identify postings that mention the possibility of teleworking in the title, description or location and construct a variable measuring the share of job postings advertising telework.

Job postings that advertise teleworking differ substantially from measures of realised teleworking because they relate to the firms’ future hires rather than their existing workforces. Hence, measuring teleworking through online job postings gives us the best available measure of its medium-term adoption, beyond ad-hoc arrangements adopted during COVID-19-related lockdowns. Another advantage is that data on job postings are available in near real-time, currently up to the end of 2021.

Taking a first look at the data, we find that the average share of remote postings across the countries in the study more than tripled over the pandemic, from just 2.5% of job postings in January 2020 to 8.5% in December 2021 (Figure 1). The increase was initially associated with the introduction of pandemic-related mobility restrictions: across countries, restrictions to mobility during the pandemic were positively correlated with changes to remote postings. In countries with high restrictions, such as Ireland, Italy, Spain and the UK, the share of remote working increased significantly more than in countries with low restrictions, such as Japan and New Zealand. However, despite the easing of restrictions during the first half of 2021, the average share of remote postings has remained near the peak it first reached in April 2021.

Figure 1 Share of job postings advertising telework and government restrictions index

As one might expect, the effects of pandemic-related mobility restrictions are largely driven by knowledge-intensive occupations where teleworking is more feasible. These include occupations such as software development and marketing, where advertised teleworking increased by 12–13 percentage points. By contrast, in occupations such as food preparation or driving, which typically require physical presence at the workplace, the share of advertised telework in online job ads increased very little, confirming evidence from other sources (OECD 2021).

Will teleworking revert back to its pre-pandemic level once the pandemic is under control? To answer this question, we separately analyse the effects of a tightening and an easing of pandemic severity on the share of job postings advertising teleworking. For the estimation, we exploit the two stylised facts we just established – that teleworking increased more in countries with more severe restrictions and in occupations where it is more feasible – to develop an econometric triple-differences strategy. Specifically, we compare the evolution of advertised telework (i) over time, (ii) across countries, and (iii) across occupations with a higher or lower potential for teleworking, which we measure using the classification of jobs into teleworkable and non-teleworkable from Dingel and Neiman (2020).

We first look at the effect of a tightening of mobility restrictions. We find that, on average, a one standard deviation increase in the stringency index of the Oxford COVID-19 Government Response Tracker (corresponding to about 10 points on a scale of 100) raises advertised teleworking by about 0.6 percentage points in high-potential occupations relative to low-potential ones over a six-month window (Panel A of Figure 2). Since the average increase in teleworking in low-potential occupations is negligible, this estimate roughly corresponds to the absolute effect of a tightening of pandemic severity on teleworking in the occupations where it is feasible. We estimate even larger effects when we measure the severity of the pandemic using COVID-19 fatalities or reductions in the Google Mobility Index.

Figure 2 Effect of tightening and easing of pandemic restrictions on advertised telework at different time horizons

Next, we turn to the effect of an easing of pandemic severity. Regardless of the measure that we use to measure pandemic severity, our analysis suggests that an easing does not have any effect on advertised teleworking. In other words, the pandemic appears to be durably driving up advertised teleworking. In complementary analysis, we find that this is particularly true in countries with good digital infrastructure as measured by broadband internet penetration. This path dependence could reflect large investments into digital tools and infrastructure as well as the acquisition of teleworking skills on the part of both employers and employees.

In sum, the analysis from our first paper (Adrjan et al. 2021) suggests that teleworking is here to stay, especially in knowledge-intensive occupations and in countries with good digital infrastructure. This is consistent with results from our complementary survey-based paper (Criscuolo et al. 2021), which further shows that, across countries, managers and workers agree on 2–3 working days per week as the ideal intensity of telework – similarly to previous findings by Bloom et al. (2021) for the UK. We now turn to perceptions of teleworking and expectations among managers and workers in more detail.

The impact of teleworking on firms’ productivity and workers’ wellbeing is a priori ambiguous. In a previous policy note (OECD 2020), we postulated that this relationship might be hump-shaped: low levels of telework adoption – and the consequent lack of flexibility – might reduce workers’ satisfaction, hamper their efficiency, and hence reduce the company’s performance. However, high levels of telework adoption might also be counterproductive: some workers start feeling isolated, and communication and knowledge flows within the company could become more difficult.

To shed light on the subjective opinion of managers and workers towards teleworking in a systematic manner, we undertook an online survey among a network of employer and employee associations in the first half of 2021, which resulted in more than 5,000 responses from 25 countries.

Figure 3 shows the three most important perceived advantages (Panel A) and disadvantages (Panel B) of teleworking emerging from the survey for both managers and workers. Around 60% of managers think that workers work more and more productively in a teleworking environment – a result well aligned with findings from Barrero et al. (2021b) and Taneja et al. (2021). A similarly high share of managers also values the possibility of hiring workers living far away from the office premises.

Figure 3 Advantages and disadvantages for managers and workers

Panel A) Advantages for firm performance (managers) and individual well-being (workers)

Panel B) Disadvantages for firm performance (managers) and individual well-being (workers)

Workers value the advantages of teleworking even more prominently. Almost 90% of employees cite less commuting as the most important upside of this working arrangement. As in Barrero et al. (2020), the time saved on the commute allows workers not only to work more on their primary activity but also to have more free time (according to 84% of them). Finally, 85% of employees indicate that they can work better on tasks that require more concentration as an additional benefit of teleworking.

Still, between 70% and 80% of managers seem to fear reduced collaboration, corporate identity, and knowledge sharing that may result from an excessive adoption of teleworking. On the workers’ side, more than 80% of them consider lack of social interactions and the fusing of private and professional life as the major downsides of teleworking. Working from uncomfortable office spaces represents a third crucial perceived drawback for almost 70% of workers. Overall, workers’ opinion about the pros and cons of teleworking confirms our conjectured trade-off regarding worker satisfaction.

If teleworking is here to stay, firms and governments will have to implement additional measures to maximise its benefits and minimise its drawbacks. Our survey allows us to investigate the measures companies foresee (and workers desire) to accommodate and facilitate the use of this working arrangement. Figure 4 shows that there is agreement among a significant share of managers and workers that the schedules of workers should be coordinated, so that at least on some office days colleagues have the possibility to meet in person, exploiting the advantages of both planned and informal interactions.

Figure 4 Planned and desired measures to be implemented in the future to maximise the benefits of telework

The second area where actions are required is company investment in information and communication technologies (ICT). More than 20% of managers and 40% of workers also stress the importance of providing ICT training to workers (‘hard skills’), complemented with training for managers on how to manage remote and hybrid teams and for workers on how to work independently from home (‘soft skills’). Our survey also finds that more productive firms plan to introduce these adaptive measures more intensively. This asymmetric response to the challenges of more intensive teleworking risks further widening performance gaps across businesses (Andrews et al. 2017), especially given the differences in the skills of the workforce of the most productive firms – in particular digital and managerial skills. (Criscuolo et al. 2021).

In this context, public policies play a key role in ensuring that teleworking has a positive impact on productivity and wellbeing, and that its advantages are not reaped solely by the most advanced firms and highest-skilled workers (Adams-Prassl et al. 2020). First, policies should enable access to teleworking from anywhere by promoting investments in broadband access and childcare facilities – both in urban and rural areas. Second, they should empower workers and managers by supporting upskilling and training on both hard (notably ICT) and soft skills. Finally, policies should protect workers from excessive teleworking by adapting the legal environment – with a special focus on health insurance coverage for remote working and the right to disconnect. Importantly, regulations should ensure this working arrangement remains a choice made jointly by employers and employees. Dialogue among social partners will be crucial to achieving these goals.