Our follow-up analyses among the Finnish private sector service workers, the essential workers during the COVID-19 pandemic, demonstrated that the effects of the first pandemic year on their well-being indicators were bidirectional; both improvements and deterioration were seen. Overall, self-perceived health deteriorated from the pre-COVID year 2019 to the end of 2020 among service workers, whereas food security improved. Slight improvements were also observed in heavy episodic drinking, smoking, and self-perceived adequacy of income. The sociodemographic characteristics, the well-being profile, and the impact of COVID-19 on well-being varied by industry. The socioeconomic profile (education and income) was the lowest in the property maintenance sector. Also, food security remained at worst level in property maintenance. Weakened self-perceived adequacy of income was reported only in the hospitality sector, which was most affected by the first pandemic wave. In contrast to our prior assumption of the solely negative impact of the COVID-19 crisis on service workers' well-being, our data showed also some positive changes that varied in magnitude in different industry sectors.
The overall well-being profile of service workers is worse than the population average. Compared with the nationally representative FinHealth 2017 survey (31), a lower proportion of respondents perceived their income as sufficient (53% among service workers vs. men 65%/women 59% in the FinHealth study). In addition, heavy episodic drinking (8% vs. men 6%/women 3%) and smoking (23% vs. men 16%/women 11%) were more common among service workers than in the adult Finnish population. The median BMI exceeded the normal weight limit (BMI≥25), as in the entire Finnish population (32). The proportion of those who perceived their health as average or worse corresponded to that of the general Finnish adult population during COVID-19 in the National FinSote survey 2020 (32).
Our results on changes in health-related indicators are in line with studies in other Western employees (5–7) and in the general Finnish adult population (19), among whom the differential impact of the COVID-19 pandemic on self-perceived health (5–7) and health behaviour (19) was observed. Large disparities in COVID-19 outcomes may arise partly from disparate working conditions and from socioeconomic differences between employees. A French study revealed that an unexpected rise in self-perceived health and well-being during the COVID-19 lockdown was not consistent across French society, as the level of self-reported health of blue-collar workers declined over the lockdown period (7). Similar changes were reported in a Swiss and German study (5); mandatory short-term workers and those who lost their job felt the negative COVID-19 impact the most. The adverse COVID-19 outcomes are partly related to higher risk exposure and higher COVID-19 incidence among blue-collar workers.
Finnish service sector workers rarely worked remotely during the pandemic. Overall, in Finland, the remote workers rated their health as better during the pandemic than those who continued working on site (10). Self-perceived health has been demonstrated to be a comprehensive measure of health status, also reflecting the condition of people beyond clinical diagnoses (33). Negative disposition and the poorer expected development of own health, recognized components in the evaluation of self-perceived health, could have been highlighted during COVID-19 and the lockdown measures. No significant changes in BMI were observed in our participants, whereas COVID-19-related weight gain has been observed in younger Finnish women and men with low education, groups known to be vulnerable for weight gain (34). As weight gain may develop over a long period of time, a longer follow-up is needed to address the long-term effects of COVID-19.
Overall, food insecurity improved from 2019 to 2020 among the service workers surveyed. This contrasts with reports from NGOs in Finland of increased need for food aid. Furthermore, studies in high-income countries have shown an increase in food insecurity since the COVID-19 pandemic (12–14). However, Lamarche et al. (15) described the prevalence of food insecurity to decrease from 3.8–1.0% during the early lockdown in Quebec, Canada. They also reported a slight improvement in diet quality, while Carroll et al. (35) noted that parents spent more time cooking meals from scratch during early lockdown. Lockdown, social distancing, and staying at home may have led to reductions in the frequency of eating out, which may have led to a better dietary quality (15). However, the prevalence of food insecurity was still alarming in 2020, with one-quarter of respondents being severely food insecure.
Multiple studies have shown that COVID-19-related job or income disruption was associated with increased severe food insecurity (e.g. 12,14,36). In 2020, the greatest proportion of employees in the hospitality sector was affected by the reduction in wages and salaries during the pandemic as layoffs and redundancies hit the sector globally (37). This was reflected in our data, where the deterioration of self-perceived adequacy of income was reported mostly in the hospitality sector, while in other industries income adequacy (relative to expenses) had improved slightly. Hence, they did not experience income loss or, if they did, their expenses diminished in parallel. If people were able to maintain their jobs and income, but possibilities to eat out, travel, and spend on shopping and hobbies were limited, it may have left more money and time to spend on food and other necessities, as well as food preparation. As speculated by Lamarche et al. (15) and Carroll et al. (36), this could have improved food insecurity. Furthermore, there are multiple reports of hospitality workers transferring to retail work, which has been associated with a smaller likelihood of food insecurity (11).
Welfare state and trade union support measures for workers might have mitigated some of the adverse effects of the COVID-19 crisis. Most unemployed and laid-off PAM union members have been on earnings-related daily allowances rather than on basic social security. According to PAM (A. Veirto 12/2022, personal communication), about 12% of PAM’s members received unemployment benefits from the PAM Unemployment Fund in 2020. It is important to note that the most vulnerable sections of the population are less likely to participate in surveys (38) or even be members of a trade union (20, 21). Furthermore, the timing of the follow-up survey in autumn 2020 may mean we were unable to capture the initial drastic effects of the COVID-19 pandemic on income sufficiency and food insecurity.
Among the service sector workers, most substance users reported no changes in their alcohol and tobacco use. However, significant likelihoods for decreasing alcohol consumption and quitting smoking were observed. This may be linked to limited possibilities to drink and eat outside the home during lockdown measures. Our observation is in line with an observed decreasing trend of alcohol and tobacco use in Finland and Europe (19, 39). It is, however, noteworthy that the share of individuals with weekly heavy episodic drinking exceeds the Finnish population average and this relatively high prevalence persisted. Despite COVID-19, a disease that particularly strains the lungs, the prevalence of smoking remained the same; about one-fifth of employees were still smokers. The results highlight the growing health inequality gap when factoring in the health consequences of alcohol and tobacco use.
Strengths and limitations
The novelty of our research lies in the seldomly studied target group of essential workers facing the COVID-19 pandemic. The follow-up data supplemented with register data enabled assessment of the effects of the first COVID-19 outbreak year on a vulnerable employee group characterized by low salaries, part-time and temporary work contracts, and physically demanding shift work. However, the low response rate raises questions about the representativeness of our sample. Based on statistics provided by PAM (21), at the end of 2019, 76% of PAM members were women, compared with 80% among the respondents in our study. Both ends of the age distribution, the youngest (less than 30 years) and the oldest (over 60 years), were under-represented in our study. Among respondents for whom employment industry category was available, the shares of those working in retail and hospitality were very similar to PAM statistics (21), but those in property maintenance were under-represented. Therefore, our sample cannot be considered fully representative of all PAM members at the time, but it does provide a means to study the well-being and coping ability of workers with limited resources during an uncertain period. Trade union membership is lower among young people, men, migrants, the unemployed, and those in part-time or fixed-term contracts (20, 21), indicating that some of most vulnerable groups were not reached since the recruitment was based on union membership. The COVID-19 pandemic has highlighted the precarious position of migrant workers who endure worse working conditions or lower pay than those protected by trade unions (40). Another validity issue concerns the extent to which people adapt to their circumstances, and the implications of adaptation for interpreting subjective well-being measures. Extreme events can result in “recalibration” of the scale (41), rather than true adaptation.