Teleworking: Does it make employees healthier and more productive? A cross-sectional study in a Southern European population | BMC Public Health

Key Findings

Most teleworkers have a positive perspective on this work situation, which is emphasized by the very high percentage of people willing to maintain this work regime. Nevertheless, TW conditions are not beneficial for everyone, with a substantial part of our sample reporting an inadequate TW environment. Also, about one sixth of participants reported a deterioration in health since TW, with a noticeable worsening of sleep and concentration problems, anxiety, stress, loneliness and sadness. Spending more time in TW was significantly related to a higher risk of health deterioration, but to a lower risk of productivity loss (although non-significant). A significant relationship was observed between inadequate TW conditions, health deterioration and lower productivity, suggesting that further support is needed for teleworkers to protect their health and safety at home, so that TW reaches its maximum benefit. Finally, TW conditions did not moderate the relationship between TW frequency, health and productivity.

Interpretation

Interestingly, the risk of health deterioration increased with more time spent in TW, compared to one or two days. The inverse relationship has already been observed between mental and physical health and the percentage of time spent working from home (9). Following the framework of Gajendran and Harrison, we postulated that a small proportion of TW (one or two days) increased autonomy and flexibility, while maintaining social interactions; but that this may not be the case when TW becomes the most common work practice, i.e. when isolation and poor communication become serious concerns (5). This first hypothesis was not fully confirmed by our data. If, on the one hand, it was observed that three or more days of TW were detrimental to health, on the other hand, we could not compare the effect of one or two days of TW with the absence of TW, since most of our sample consisted of teleworkers. The negative health consequences of more intensive TW could be explained by isolation and poor communication, as described by Gajendran and Harrison (5). Other explanations could be considered, such as the observed increase in unhealthy behavior in our sample, with easier access to unhealthy food or alcohol outside the workplace, confirmed in other studies that also showed more sedentary behavior (29, 30). This aspect might be related to the well-known impact of context on lifestyle, which is not shaped by peer pressure (workmates) and employers’ norms in teleworking.

The beneficial effects on productivity perception have already been observed in the literature (22). Bloom and colleagues suggest that productivity is not increased by TW from the perspective of managers, but from the perspective of employees, who produce the same work in fewer hours because of the savings in travel time (2).

To our knowledge, the role of physical TW conditions at home has not yet been investigated to explain the relationship between TW and health. Studies in this area have focused more on the role of living arrangements (31). However, an extensive literature shows that the indoor environment at work affects employees’ well-being and productivity (32). For example, the role of some working conditions and productivity has already been established (33); in turn, the relationship between poor health and lower productivity has also been demonstrated (34). Our findings confirm our second hypothesis of a relationship between TW conditions and health and productivity outcomes. Based on this relationship, we can hypothesize that inadequate working conditions at home reduce employees’ performance through their negative effect on mental health.

Although exploratory, our descriptive analysis provides insights on mental health that we consider relevant and that deserve further specific research. Our findings mainly confirm the worsening of mental health symptoms observed in another article that showed increased stress (35); however, other studies observed decreased stress (36,37,38), alcohol abuse and depression (39), which contradicts our results. Also, other studies did not observe significant changes in mental health symptoms (40), and a global review confirmed mixed results (17). Note that the results of our study are not fully comparable with previous contributions, as our sample was mainly composed of teleworkers, which hinders a comparison with other workers, including more detailed statistical analyses. Moreover, our study was conducted during the post-pandemic period, making our findings hardly comparable with those obtained in the pre-pandemic or the pandemic period, where TW was in its first steps or justified by public health measures, respectively. In contrast, our findings likely focus more on the effective impact of TW on health outcomes, as these are likely less contaminated by the effects of the pandemic, particularly on mental health.

Limits

A first possible limitation is that our sample is younger than the Portuguese teleworking population, 52% of whom are over 45, while only 20.6% in our sample are over 50 (data from the National Institute of Statistics, specific module on “working from home”, from 2022, accessed at www.ine.pt on November 2, 2023).

Nevertheless, the educational level was quite similar (74.9% of participants with a higher education in our sample, for 72.3% in the Portuguese population), as was the type of occupation (87.7% employed in the service sector in our sample, for 85.5% in the Portuguese population. As such, our sample is quite representative of Portuguese teleworkers, grouping mainly people with a higher education working in the service sector. Nevertheless, our sample mainly includes teleworkers, which hampers a comparison with people who work entirely at the workplace, including the identification of statistical differences in health outcomes, productivity and well-being. Second, although the study was conducted in the post-pandemic period, we cannot completely rule out a “pandemic effect”. Unfortunately, no information was collected on the time since TW, which would have helped to better disentangle these influences. Third, in order to increase the sample size, we chose to complement our sample with people who regularly participate in Gfk surveys. The combination of paid and unpaid participants may represent a bias in response behavior, although we did not observe a major difference in the results for each sample considered separately (table available upon request). Fourth, TW may adopt different models, in terms of location, seasonality, frequency of online contacts, supervision, etc. All these aspects contribute to the TW intensity level. In the absence of better information, we limited our analysis to the number of weekly days in TW. This may not fully cover all aspects of TW intensity, but is aligned with the most current discussion at the company level, namely on full versus hybrid TW, or about the optimal number of TW days in the case of a hybrid model.

Finally, relying on self-reported data introduces potential sources of bias, including recall bias, justification bias, and desirability bias. For example, respondents may unintentionally recall or present information in a way that aligns with their perceptions or preferences, thereby affecting the accuracy of the collected data. Respondents may overestimate the impact of TW on productivity, stemming from a desire to maintain the perceived benefits of their current way of working. These biases raise concerns about the validity of causal inference analyses and the extent to which definitive conclusions can be drawn from our study results. Given these limitations, while our findings provide valuable insights, it is essential to interpret them with caution and consider the potential influence of the aforementioned biases on the overall outcomes of the study.

Implications

This article highlights the relationship between TW conditions, well-being, health and productivity. That is, even if our sample included people who mainly enjoyed favourable TW conditions, there is a potential harm to the health and well-being of those who do not enjoy such conditions. This finding highlights the need to monitor TW conditions and potentially provide support to those who are not expected to benefit from adequate housing conditions.

To our knowledge, Portugal was the first European country to legally regulate TW in the private sector, in 2003. One of the main objectives of the Portuguese law was to ensure equal treatment of workers in this regime. In practice, the law defines a list of limited compensation values ​​to be paid to the teleworker (an unallocated daily payment). At the organizational level, teleworkers are subject to the same limits of daily and weekly working hours as other workers, while remote meetings and joint tasks must take place within working hours. Nevertheless, in practice, not all of these legal concerns are applied and workers are rarely compensated.

Based on our results, we propose that regulations should include medical surveillance of workers, starting with a self-report questionnaire and supplemented by a physical examination by an occupational health professional (doctor or nurse). Preventive and corrective measures at the worker’s home may be necessary. All procedures referred to should be framed by laws, to ensure that occupational health prevention is effective.