I want to break free: the influence of the COVID-19 pandemic on work–life balance satisfaction

ABSTRACT

The revolution in the organization of work, induced by the COVID-19 pandemic, has had, and will have, some long-lasting consequences on how jobs will be performed, affecting adult life well-being. What we know so far about the determinant of satisfaction for work–life balance may not be valid anymore considering this sudden revolution in work organization. This paper presents an empirical analysis based on a cross-sectional survey of 803 workers in October 2020. Results suggest that the profile that benefits the most from the revolution in work organization is represented by a male, married, public sector worker working remotely and not living in the southern part of Italy.

KEYWORDS:

• remote working
• work–life balance satisfaction
• gender gap

JEL:

• I18
• J16

1. INTRODUCTION

The COVID-19 pandemic has undoubtedly provoked profound changes in society and the economy; it has harmed the economy, the well-being of people, work–life balance satisfaction and health (Eurofund, 2021). Work–life balance conditions have changed during the pandemic due to the increasing prevalence of remote working in telecommuting or work-from-home (WFH) practices (International Labour Organization (ILO), 2020b).

Italy has been one of the most severely affected countries worldwide and was the first Western democracy to be struck by the virus (Alfano & Ercolano, 2020; Bonacini et al., 2021; Mariotti et al., 2022b). In Italy, the restrictive measures adopted by the government at the beginning of March 2020 initially stopped the activity of entire production systems. They subsequently radically changed work organizations in both the public and private sectors. Remote working¹ proliferated, especially for workers in information and communication, financial and insurance activities, and business services (Barbieri et al., 2020; Sostero et al., 2020). The phenomenon also affected large parts of the Italian public administration. ISTAT (2020) estimated that about 7 million workers could work at a distance: among these, 4.1 million jobs require supervision and 2 million can operate with a high level of autonomy. Several large industries (the entire banking sector, for instance) kept workers in remote working or provided a varying mixture of work in the office and from home. In such scenarios, at least three issues affect companies, policymakers and workers and their satisfaction with work–life balance.

The literature exploring the effects of the COVID-19 pandemic on work–life balance satisfaction distinguishes between what happened before and during the pandemic. Specifically, the analyses carried out during the pandemic focus mainly on the effects of remote working on the gender gap (Beno, 2019; Bonacini et al., 2021; Marra, 2012, 2020), well-being (George et al., 2021; Sirgy & Lee, 2018; Teodorovicz et al., 2021) and work–life balance satisfaction (Allen et al., 2020; Eurofound, 2021; Fana et al., 2020; Greenhaus et al., 2012).

The gender gap worsened during the first wave of the pandemic, if compared with the situation before. Men and women took over caregiving tasks (Eurofund, 2021; ILO, 2020b), and studies show that women were more negatively impacted than men (Corsi & Ilkkaracan, 2022; Eurofound, 2022b; Marra, 2020), while parents with school-age children complained about distractions and a poor work–life balance (Craig & Churchill, 2021). Nevertheless, after the first wave of the pandemic, the results regarding the effects of remote working on work–life balance are not conclusive. Some studies find that remote working, which largely involves highly qualified and knowledge workers (Eurofound, 2021, 2022a; Florida et al., 2021), increases satisfaction and productivity and allows for a better reconciliation of work–family duties (i.e., shorter commuting times and a more comfortable working environment) (Eurofound, 2022a; Fana et al., 2020; ILO, 2020c). Others stress its effect on the gender gap, since women undertook more housework and childcare even when their partners were working remotely (Del Boca et al., 2021; Farré et al., 2020; Tavares et al., 2020).

In this context, this paper explores the factors that influence work–life balance satisfaction in light of the revolution induced by the COVID-19 pandemic in Italy. It aims to answer the following research question: Which factors might affect work–life balance satisfaction? Work–life balance satisfaction during the COVID-19 pandemic is regressed over: (1) gender; (2) education level; (3) family characteristics; (4) type of employment (sector, private versus public sector); (5) working mode (remote working); and (6) provincial context (average size of housing per resident and digital divide). The paper analyses a new database developed from an ad hoc survey of 803 workers in Italy carried out in October 2020 right before the second COVID-19 pandemic wave led to new restrictions being imposed in Italy. The focus on this specific period allows us to highlight the relationship between work–family balance satisfaction and the fast-growing tendency towards remote working, which in Italy has accelerated with COVID-19 restrictions. This article thus contributes to the broad-based gender gap literature and intends to situate the current debate on work–family friendly policies at the intersection of gendered disparities in the distribution of professional and caring work, remote work organization, and the spatial characteristics of conditions of social life in times of crisis.

The rest of the paper is organized as follows. Section 2 presents the literature review about the factors influencing work–life balance satisfaction. Section 3 focuses on the case of Italy. Data and methodology are presented in section 4. Results are given in section 5. Section 6 is dedicated to a discussion of the results, and conclusions.

2. THEORIES AND EVIDENCE UNDERLYING WORK–LIFE BALANCE SATISFACTION

As mentioned previously, existing studies on work–life balance satisfaction differentiate situations before and after the COVID-19 crisis and put great emphasis on: (1) the organization of work, with an increasing focus on remote working and work–life balance satisfaction; and (2) the gendered differences in the distribution of time spent on professional and caring responsibilities, as examined below.

2.1. Work–life balance satisfaction

According to Felstead et al. (2002, p. 56), work–life balance can be defined as ‘the relationship between the institutional and cultural times and spaces of work and non-work in societies where income is predominantly generated and distributed through labour markets’. This relationship can influence time availability for work, resources, household roles and social norms that affect work–life balance and satisfaction (Evertsson & Nermo, 2004; Geist & Ruppanner, 2018; Rios-Avila et al., 2021). Sirgy and Lee (2018, p. 232), on the other hand, define work–life balance as ‘a high level of engagement in work life and nonwork life with minimal conflict between social roles in work and nonwork life’. Indeed, work–life conflict generates much stress and reduces satisfaction in work and non-work life (e.g., Allen et al., 2020; Greenhaus et al., 2012).

Before the COVID-19 pandemic, the lack of accessible, affordable and good-quality care services, and the disproportionate amount of time spent on care activities, made it difficult for women to balance work and family life (EIGE, 2021). Several studies have demonstrated that a good work–life balance leads to higher organizational performance, job satisfaction and more substantial organizational commitment (Allen et al., 2020). It also plays an essential role in individual well-being (e.g., health satisfaction, family satisfaction and overall life satisfaction) (Sirgy & Lee, 2018).

According to the COVID-19 e-survey undertaken by Eurofound (2021), the work–life balance of European Union (EU) workers during the first wave of the pandemic deteriorated compared with the situation described in Eurofound’s 2015 European Working Conditions Survey (Corsi & Ilkkaracan, 2022). Across the EU-27, concentrating on work became more difficult for employed women than it did for employed men (8% of women and less than 5% of men), negatively impacting domestic work for women more than in the case of men. Nearly a third of the women (31%) reported feeling too tired after work to undertake some of the necessary household chores, compared with about a quarter (26%) of men. Furthermore, both found it difficult to spend enough time with their family. A fifth (21%) of women and men responded that their work prevented them from spending as much time as they wanted with their families, and this is particularly important as it represents an increase of 10 percentage points from 2015.

The adverse effects of the COVID-19 pandemic on work–life balance were attenuated after the first wave. Several studies found that flexible work options played a crucial role in work–life balance satisfaction because it gave employees greater control over managing work and multiple non-work activities (Allen, 2001; Fana et al., 2020; Thomas & Ganster, 1995). However, flexible work can only be conducted by knowledge workers specialized in sectors with a high teleworkability index (Barbieri et al., 2020; Sostero et al., 2020), and the situation thus results in increasing social and spatial inequalities. Around 37% of EU-27 workers are in occupations that can be carried out from home, a higher percentage if compared with pre-outbreak (15% regular or occasional telework in 2019) (Eurofound, 2021).

Thomas and Ganster (1995) found that only flexible scheduling significantly affected the psychological and physiological indicants of job strain outcomes. These findings may occur because all employees can profit from flexible work options. In contrast, childcare-related benefits are more likely to be of value to only a subset of the workforce.

Teodorovicz et al. (2021) collected data on 1192 US full-time employees in knowledge-intensive occupations in two waves: in August 2019 pre pandemic (615 participants) and in August 2020 during the pandemic (577 participants). Both the samples of workers, who were commuting before the pandemic, had to recollect the most important working day from the previous week and describe in a diary the main activities they had undertaken (type of activity, start time and end time). The survey also collected data on workers’ socio-economic characteristics and managerial responsibilities. The forced transition to WFH induced by the pandemic resulted in a drastic reduction in commuting time and increased time spent on work and/or personal activities. However, this change was heterogeneous across different workers and organizations: managers were more willing to reallocate the time gained from commuting into more time for meetings, and WFH did not impact self-reported measures of well-being (Teodorovicz et al., 2021).

Bu et al. (2021), in their study of time use during the COVID-19 lockdown measures in the UK, found that changes in time spent on a range of activities (e.g., working, volunteering, doing housework, gardening, exercising, reading, engaging in hobbies, etc.) were associated with increased life satisfaction. Ciolfi and Lockley (2018) state that flexibility in setting and removing boundaries can help workers improve their work–life balance, thus reducing the risk of becoming a workaholic. By contrast, increased time devoted to following COVID-19-related news was associated with decreased life satisfaction (Bu et al., 2021).² Interestingly, Teodorovicz et al. (2021) also found that for some WFH workers, work and personal life collide.

2.2. Gender gap and remote working

The literature shows that ‘paid and unpaid work are both heavily gender-segregated, reflecting deep-rooted social norms about gender roles of women as primary caregivers. These divides translate into gaps in the labour market, pay and well-being, including in poverty and work–life conflict’ (Eurofound, 2022b, p. 1). Before the pandemic, women working from home reported slightly better work–life balance scores than men, and they valued flexible work schedules and limited commuting time more than men (Eurofound, 2017). Women could dedicate that time to carrying out caregiving and domestic responsibilities (EIGE, 2021), and thus may be more positively affected by the opportunity to work from home.

The situation changed during the pandemic because men and women took over caregiving tasks depending on their job typologies (e.g., essential jobs versus knowledge-intensive jobs) (ILO, 2020b), and family composition (Eurofound, 2022b; Fana et al., 2020). Nevertheless, the growth of remote working led to an increase in unpaid work for women, especially during the lockdowns when schools were closed down and teaching was online. The Eurofound (2022b) survey, carried out during the pandemic, suggests that women undertook a disproportionate amount of housework and childcare, even when both parents worked remotely. Similar results have been found by Del Boca et al. (2021) and Bonacini et al. (2021) in Italy, Farré et al. (2020) and Seiz (2021) in Spain, Germany, Tavares et al. (2020), in Portugal, United Nations Development Programme (UNDP) (2020) in Turkey, and Fana et al. (2020) in France, Spain and Italy. Bonacini et al. (2021), in their study of WFH among Italian employees, find a positive shift in WFH feasibility that would be associated with an increase in average labour income. However, this average labour income would favour male, older, highly educated and highly paid employees. In addition, employees living in provinces that were more affected by the COVID-19 pandemic would benefit more from WFH, thus increasing the pre-existing inequalities in the labour market. Fana et al. (2020) find that the effects of remote working on work–life balance in France, Spain and Italy depend on workers’ occupation and family composition. Especially during the pandemic, some workers suffered from isolation, and those with school-age children complained about distractions and a poor work–life balance (Craig & Churchill, 2021). Similarly, in their study of Portugal, Tavares et al. (2020) point out the difficulties in reconciling teleworking with family life, household chores, dedication to children and time/schedule management. Recently, a UNDP time–use survey conducted in Turkey during pandemic lockdown conditions in May 2020 has revealed a range of gender impacts caused by the crisis, both with paid and unpaid work (UNDP, 2020). Several changes induced by the pandemic (i.e., school closures, and higher demand for household consumption items and domestic and care services) negatively affected household production and care labour. Women provided supply in the face of this demand shock by increasing their paid and unpaid work hours. More employed women than men left their jobs and had less access to leave with pay. The survey results also highlight a substantial increase in men’s unpaid work time, specifically among men who continue working from home, thus underlining that, under certain conditions, a more gender-equal sharing of household and care work should be advocated and promoted. By contrast, Arntz et al. (2020), in their study of Germany, found that WFH reduced the gender gap in work hours and monthly earnings primarily because women with children who teleworked could increase their work hours more than those who did not have the option to telework.

Another aspect that previous literature suggests might have negatively affected women is the care of the elderly and people with disabilities and mobility restrictions since, during the pandemic, daycare services were closed down. In the first lockdown period, women spent 4.5 h per week, on average, caring for elderly or disabled family members, while men spent 2.8 h doing the same (Eurofound, 2021). In addition, women are more willing to fill the care infrastructure gap (Corsi & Ilkkaracan, 2022; Folbre & Bittman, 2004; Henz, 2009; Luppi & Nazio, 2019; Saraceno, 2008). In 2018 more than 10 million workers in the EU-28 (6 million of them women) cared for family members (15 years and older), representing 6% of employed women and 4% of employed men (EIGE, 2021).

It can be concluded that the overall effect of remote working on the gender gap and gender wage gap is ambiguous: it depends, among other things, not only on potential changes for women but also on potential changes for men. The unequal distribution of household responsibilities may contribute to the gender wage gap by reducing women’s participation in paid work or pushing them towards part-time jobs and the informal sector (Marra, 2014, 2020). On the other hand, home-based telecommuting arrangements are related to increases in the gender pay gap in Italy (Bonacini et al., 2021) and Austria (Beno, 2019).

3. WORK–LIFE BALANCE SATISFACTION IN ITALY: STYLIZED FACTS

The latest data from the Labour Force Survey (LFS) by Eurostat (2019) relate to the pre-COVID-19 period and provide interesting information regarding the distribution of remote working practices among professionals. In Italy, the percentage of workers engaged in remote working compared with all workers in the 15–64 age group was 1.1% for salaried employees, 3.6% for casual workers and 12.9% for self-employed workers (Table 1). France, Spain and Germany were significantly higher at 4.7%, 2.5% and 3.2%, respectively, for salaried employees, 7.0%, 4.8% and 5.2% for casual workers, and 24.4%, 17.4% and 25.9% for casual self-employed workers. The situation is very different in the countries that historically have been most sensitive to new forms of work, such as Finland and the Netherlands, where the percentages are 10% for salaried employees, 14% for casual workers and over 40% for self-employed. There was thus very little use for remote working in Italy before the pandemic. This lack was also due to persistent problems characterizing the Italian labour market, particularly those concerning productivity, precarious contracts, low wage levels, polarization between salaried employees and the self-employed, the gender gap, and the absence of lifelong learning measures aimed at the reskilling and upskilling of workers.

Table 1. Employed in smart working by type of work (15–64 years old) as a percentage of total employed, 2019.

Country	Permanent workers (%)	Temporary workers (%)	Self-employed (%)
Italy	1.1%	3.6%	12.9%
France	4.7%	7.0%	24.4%
Spain	2.5%	4.8%	17.4%
Germany	3.2%	5.2%	25.9%
Sweden	3.8%	5.9%	27.9%
Finland	10.0%	14.1%	44.8%
The Netherlands	9.0%	14.1%	43.0%

Source: Eurostat (2019).

Following the ISTAT LFS (2019), out of a national workforce of about 23 million people, only a little more than 1.3 million individuals (5.7%) declared that they used their home as a principal, secondary or occasional place of work. The remainder – about 22 million people (94.3%) – stated that they only used traditional workplaces. Those over 15 years old who claimed to work from home or in conventional workplaces are broken down by occupational status, sector of economic activity and profession. As regards the professional level, the dependent contractors (self-employed with a single client) and the self-employed were those most interested in remote work, with about 18% and 17% declaring they worked from home. Next were employers (10.8%), family workers (5.8%), permanent employees (3.1%) and temporary workers (2.3%). As expected, remote working was more developed in the service sector (7%) than in agriculture (3.5%) and industry (2.9%). And finally, the professions where skilled workers were more often employed at home (around 13%), followed by executives in office work (2.5%), workers and executives in commerce (1.9%) and unskilled occupations (0.9%).

According to the Osservatorio Smart Working of the Politecnico di Milano, the Italian remote workers who answered the survey in 2020 did not define ‘home’ as the best workplace. Remote workers complained about inadequate technology (29%), a sense of loneliness (29%), difficulties in guaranteeing work–life balance (27%) and the sensation of being constantly connected (26%) (Osservatorio Smart Working, 2020).

The pandemic crisis heavily impacted the labour market, especially among women (Villa, 2021). ISTAT (2020) estimated that the pandemic shock impacted around 3 million households, namely those with at least one child under 14 and one or both parents employed and concluded that pandemic stress varies according to the type of work (skilled or unskilled) and manner of working (smart working or not). What were the consequences for women?

ISTAT (2021) reports that between November 2019 and November 2020, 98% of the roughly 100,000 employees fired in Italy were women. The reasons for this quite astonishing figure are to be found in the lack of employment policies that make it possible to reconcile work and family, in the scarcity of early childhood services (especially in southern Italy), and in the distribution of family workloads, which is still highly unbalanced, to the disadvantage of women. Regarding the latter, it is helpful to report an ISTAT survey (2018) stating that 42.6% of women with children aged 0–5 declared they had adaptation costs related to the difficulties of work–family reconciliation, compared with only 12.6% of men.

The effect on work organization and work–life balance arising from the forced closure of schools and the use of online teaching, which affected all Italian regions during the first COVID-19 wave (March–May 2020), and only some of these in the period under analysis (October 2020), is of particular interest. Without dwelling on the asymmetric effects produced on the accumulation of human capital among children divided by income, education and parental work, and geographical origin,³ this measure likely altered the balance between family workloads and work–life balance. There are good reasons to believe that these changes further exacerbated the country’s chronic imbalances, to the detriment of women (Alfano et al., 2021; Naldini & Saraceno, 2022).

3.1. Work–life balance in Italy during the pandemic: public versus private sector

Another aspect affecting work–life balance satisfaction is the in which sector workers are employed. We might distinguish, for example, between jobs in the private and public sectors. According to Osservatorio Smartworking (2020), public administration workers have a better work–life balance than those working in the private sector (Figure 1).

Figure 1. Criticalities related to remote working.

Source: Osservatorio Smartworking (2020).

3.2. Work–life balance and the Italian north–south divide

Finally, another aspect that can play a role in work–life balance is where people live and work. For instance, the home-care workload falls more on women than on men in the southern Italian regions, which are also characterized by poor early childhood services and an unbalanced family load, which remains highly disadvantaged for women.

In accordance with the literature review, we frame the following four research hypotheses:

Hypothesis 1: Work–life balance satisfaction is related to gender and family composition.

Hypothesis 2: Work–life balance satisfaction increases for public sector workers.

Hypothesis 3: Work–life balance satisfaction increases if workers can work remotely.

Hypothesis 4: Work–life balance satisfaction increases if workers work in the north of Italy.

4. DATA AND METHODOLOGY

The empirical analysis is based on a cross-sectional dataset. Data are taken from a survey of 803 workers from 1 to 6 October2020. This period belongs to the so-called ‘second phase’ of the pandemic (June–October 2020), when there was a gradual easing of the previous restrictions and containment measures and economic activities reopened (Mariotti et al., 2022b).⁴

The survey was conducted⁵ through a computer-assisted web interviewing (CAWI), and is part of a more comprehensive study about the behaviour of companies and workers during the COVID-19 pandemic, founded by the Associazione per lo Sviluppo dell’Industria nel Mezzogiorno (SVIMEZ). The survey consisted of the following parts: (1) personal information (age, gender, education, residence, family); (2) job information (sector, contractual position, organizational methodologies of work); and (3) work–life balance satisfaction. To estimate the impact on the work–life balance satisfaction of the first wave of the COVID-19 pandemic, we modelled the individual theoretical satisfaction for the work–life balance as a function of several variables through the following equation: (1) $y_i=\alpha +{\text{β}}_1\mathrm{Jo}{b}_i+{\text{β}}_2\mathrm{Sout}{h}_i+{\text{β}}_3\mathrm{Fa}{m}_i+{\text{β}}_4\mathrm{Pro}{v}_i+{ϵ}_i$(1) where y represents the work–life balance of individual i; Job is a matrix of variables discriminating for the various kinds of jobs of individual i; South is a dichotomous dummy variable discriminating for individuals residing in the southern part of Italy; Fam is a matrix controlling for the family situation of individual i; Prov is a matrix controlling for the provincial context; and finally $ϵ$, as usual, is the error term.

The literature has identified various factors that may play a role in personal satisfaction with the work–life balance, which may be summed up in these three groups of variables. More specifically, the dependent variable y is proxied by answering the question about the respondents’ satisfaction with the ‘Balance between work and private life’.⁶ From ‘A lot’ to ‘None’, four possible answers were transformed into an ordinal variable distributed on a 1–4 scale, where 4 represents the most satisfied respondents. We built our dependent variable y accordingly. Figure 2 shows the mean values of this variable in each Italian NUTS-3 province.

Figure 2. Average values per year (work–life balance satisfaction).

Source: Authors’ elaboration from Associazione per lo Sviluppo dell’Industria nel Mezzogiorno (SVIMEZ) survey data.

The matrix Job is composed of the variables that belong to the domain of job characteristics that may play a role in work–life balance satisfaction. These are:

• DGraduated : a dichotomous dummy variable that assumes the value of 1 if the respondent declares him- or herself to be a graduate, and 0 otherwise. It is crucial to control for this discontinuity since the literature has typically found higher satisfaction among more educated people (Glenn Dutcher, 2012).

• DRemoteworking : a dichotomous dummy variable that assumes the value of 1 if the respondent started working remotely when the COVID-19 crisis erupted (it includes the answer ‘I kept working in the same region but in smart working’ and ‘I kept working in smart working in a different region from my working place’,⁷ to question VD9 of the survey). It is crucial to control for this discontinuity since people forced to work at home may significantly differ in their satisfaction with the work–life balance if compared with the rest of the population (Barbieri et al., 2020).

• DPublicSector : a dichotomous dummy variable that assumes the value of 1 if the respondent declares that he or she is employed in the public sector. It is essential to control for this characteristic since people working in the public sector may have different utility from being forced to work from home rather than people in the private sector.

• DServicesSector : a dichotomous dummy variable that assumes the value of 1 if the respondent declares that he or she is employed in the service sector. It is crucial to control for this characteristic since, as we have seen, these people are more affected by this change in the Italian job market (Barbieri et al., 2020).

The matrix Fam concerns the variables regarding gender, marital status, and having or not having children, which may play a role in work–life balance satisfaction:

• DFemale: a dichotomous dummy variable that assumes the value of 1 if the respondent is female. It is crucial to control for gender since remote working may have affected women more than men because of the gender gap in taking care of family chores and women’s higher share of job losses during the pandemic.

• DMarried: a dichotomous dummy variable that discriminates between married and unmarried (singles, divorced, and widows) respondents. It is crucial to control for this discontinuity since the literature has identified that being married is an essential determinant of work–life balance.

• DChildren: a dichotomous dummy variable distinguishing between respondents with co-living offspring and the rest. Remote working is expected to be significantly different from working from home depending on whether one lives with or without children.

The matrix Prov concerns the variables regarding the average number of square meters per resident in the province, and the availability of high-speed internet:

• Av.Sq.Meters : a variable measuring the average number of square meters available per resident in each Italian province. Data is taken from an ISTAT survey in 2011 (the last year in which it is available).

• H.FTTH per capita: a variable measuring the availability of high-speed internet in Italian provinces. Following Mariotti et al. (2022b), these data are gathered by AGCOM (autorità per le garanzie nelle comunicazioni) and refer to 2022. It is the sum of households in the province served by FTTH technology (calculated with approximated methodology), divided by the population to obtain a per capita value.

• DSouth: a dichotomous dummy variable differentiating between people living in a southern region or on an island and people from the rest of Italy. Italy is a highly heterogeneous country characterized by a north–south divide due to its productive structure and distribution of social capital (SVIMEZ, 2021). Table 2 presents descriptive statistics for all the variables included in the model.

Table 2. Descriptive statistics.

Variable	Variable	Observations	Mean	SD	Minimum	Maximum
Work–Priv Balance Satisfaction	Ordinal variable ranking on a 1–4 scale: satisfaction with work–life balance	796	2.839196	0.7176281	1	4
DGraduated	Dichotomous dummy variable equal to 1 if the respondent has a degree	796	0.4032663	0.4908618	0	1
DRemoteWorking	Dichotomous dummy variable equal to 1 if the respondent is in remote working after the COVID-19 crisis	796	0.3844221	0.4867642	0	1
DServices Sector	Dichotomous dummy variable equal to 1 if the respondent works in the services sector	796	0.1796482	0.3841356	0	1
DPublic Sector	Dichotomous dummy variable equal to 1 if the respondent works in the public sector	796	0.3178392	0.4659294	0	1
DSouth	Dichotomous dummy variable equal to 1 if the respondent is resident in southern Italy or the islands	796	0.1306533	0.3372327	0	1
Av.Sq.Meters	Average m^2 of inhabitation per resident in the province	796	41.48677	3.322745	30.73	48.92
H.FTTH per capita	Number of households served by FTTH technology in the province divided by the population	796	0.2587974	0.0874112	0.0646788	0.4523697
DFemale	Dichotomous dummy variable equal to 1 if the respondent is female	796	0.451005	0.4979066	0	1
DMarried	Dichotomous dummy variable equal to 1 if the respondent is married	796	0.5615578	0.4965081	0	1
DChildren	Dichotomous dummy variable equal to 1 if the respondent has co-living children	796	0.491206	0.500237	0	1

While the cross-sectoral nature of the data is not the best setting in which to derive causality, data constraints suggest that an order probit estimator is the best choice to empirically estimate equation (1(1) $y_i=\alpha +{\text{β}}_1\mathrm{Jo}{b}_i+{\text{β}}_2\mathrm{Sout}{h}_i+{\text{β}}_3\mathrm{Fa}{m}_i+{\text{β}}_4\mathrm{Pro}{v}_i+{ϵ}_i$(1) ), because the dependent variable y is an ordinal variable ranking qualitative answers to a question regarding satisfaction (Daykin & Moffatt, 2002). The ordered probit model deals consistently with the various features of such data and is therefore the best statistical device when working with these data. A robustness check may be performed through an ordered logit estimator that assumes a different non-linear distribution from the inverse of the cumulative distribution function of the standard normal distribution (Williams, 2016).

5. RESULTS

Table 3 presents the results of the ordered probit estimation. The matrixes are included one at a time, in three different specifications of the model, to better control multicollinearity and endogeneity issues due to the simultaneous presence of different variables in the regression.

Table 3. Ordered probit model.

	(3.1)	(3.2)	(3.3)
	Work–Priv Balance Satisfaction	Work–Priv Balance Satisfaction	Work–Priv Balance Satisfaction
DGraduated	−0.0834	−0.0786	−0.0558
	(−1.00)	(−0.94)	(−0.66)
DRemoteworking	0.168**	0.160*	0.150*
	(1.97)	(1.87)	(1.75)
DServices Sector	0.0357	0.0420	0.0582
	(0.33)	(0.38)	(0.53)
DPublicSector	0.125	0.140	0.153*
	(1.37)	(1.53)	(1.65)
DSouth		−0.187	−0.193
		(−1.54)	(−1.58)
H.FTTH per capita		−0.00930	−0.0692
		(−0.02)	(−0.15)
Av.Sq.Meters		0.00808	0.00765
		(0.64)	(0.60)
DFemale			−0.125
			(−1.56)
DMarried			0.250***
			(2.64)
DChildren			−0.0787
			(−0.84)
cut1	−1.670***	−1.362**	−1.354**
	(−17.89)	(−2.32)	(−2.30)
cut2	−0.546***	−0.234	−0.215
	(−7.93)	(−0.40)	(−0.37)
cut3	1.126***	1.442**	1.473**
	(14.97)	(2.46)	(2.50)
Observations	796	796	796
Pseudo-R^2	0.004	0.006	0.012

Note: t-statistics are shown in parentheses. *p < 0.1; **p < 0.05; ***p < 0.01.

As can be seen, in relation to the variables of the first matrix, ‘not being a graduate or working in the service sector’ has a statistically significant effect on work–life satisfaction. The difference between this result and the main findings reported in the literature may be due to the particular time the survey was carried out. Indeed, in October 2020 the service sector was experiencing a deep crisis, along with many not strictly necessary sectors. Degree holders employed in this sector may therefore have had lower work–life satisfaction because their stability of work was in danger (Alfano et al., 2021). The dummy discriminating for people working in the public sector is statistically significant (at the 10% level)⁸ only in the complete specification (3.3). This shows a positive sign, suggesting that working in the public sector has some impact on work–life satisfaction, possibly because workers from this sector are among those whose work is most secure, and whose future appears most stable. Finally, the DRemoteworking variable has a positive coefficient and is statistically significant at 5% (3.1) and 10% (3.2 and 3.3). These results confirm the literature review’s main findings, and suggest that respondents who engage in remote working are more satisfied with their work–life balance than others.

Finally, when analysing the Fam matrix, we may see from specification 3.3 that not being female or having co-living offspring has a statistically significant effect on satisfaction with the balance between work and private life. While surprising in light of the gender gap literature, this is a finding that is further confirmed by the histogram in Figure 3, which shows that the distribution of the answers from respondents of both genders is very similar. Thus, neither living with offspring nor being female plays a role in this satisfaction.

Figure 3. Histogram of dependent variable, female, and male.

Source: Authors’ elaboration from Associazione per lo Sviluppo dell’Industria nel Mezzogiorno (SVIMEZ) survey data.

The most statistically significant variable is instead DMarried (at the 1% level), whose coefficient has a positive sign. This suggests that married respondents are more satisfied than unmarried ones. Being married is thus the most important factor favouring the balance between work and personal life.⁹

To check the robustness of our results we replicated the analysis using a different estimator, namely an Ordered Logit model. These results, presented in Table 4, are largely consistent, suggesting some stability in our estimates. The main difference is that the dummy discriminating for residence in Italy’s southern regions and islands has a negative and statistically significant (at the 10% level) coefficient, suggesting that with this model, people from these areas are less satisfied with their work–life balance than those in the northern and central parts of Italy.

Table 4. Ordered logit model.

	(4.1)	(4.2)	(4.3)
	Work–Priv Balance Satisfaction	Work–Priv Balance Satisfaction	Work–Priv Balance Satisfaction
DGraduated	−0.167	−0.158	−0.113
	(−1.13)	(−1.07)	(−0.75)
DRemoteworking	0.305**	0.295*	0.279*
	(2.02)	(1.94)	(1.83)
DServices Sector	0.0629	0.0676	0.0895
	(0.32)	(0.35)	(0.45)
DPublicSector	0.214	0.234	0.257
	(1.32)	(1.44)	(1.56)
DSouth		−0.338	−0.356*
		(−1.58)	(−1.66)
H.FTTH per capita		0.0497	−0.0647
		(0.06)	(−0.08)
Av.Sq.Meters		0.0119	0.0109
		(0.52)	(0.48)
DFemale			−0.202
			(−1.42)
DMarried			0.473***
			(2.81)
DChildren			−0.163
			(−0.98)
cut1	−3.026***	−2.565**	−2.544**
	(−15.36)	(−2.42)	(−2.39)
cut2	−0.881***	−0.415	−0.379
	(−7.47)	(−0.40)	(−0.36)
cut3	1.894***	2.369**	2.433**
	(13.87)	(2.25)	(2.30)
Observations	796	796	796
Pseudo-R^2	0.004	0.006	0.012

Note: t-statistics are shown in parentheses. *p < 0.1; **p < 0.05; ***p < 0.01.

6. DISCUSSION AND CONCLUSIONS

Before going more deeply into the discussion, we must highlight that our analysis is based on a cross-sectional dataset, studying a picture of the situation in early October 2020 and not comparing it with the previous cases. Empirical analysis allows us to disentangle the winners from the losers as regards work–life balance satisfaction. According to our research, the profile that benefits the most from the revolution in the organization of work is a male, married, public sector worker, working remotely, and not living in the southern part of Italy. This tests the four research hypotheses framed in section 3.

This paper contributes to filling the gap in the literature about work–life balance satisfaction in the case of Italy, the country which was most severely affected by the first wave of the COVID-19 pandemic. The results confirm the main findings of the literature about the gender gap, and could be read, somewhat provocatively, as implying that married men are more satisfied because they can (more) easily offload their (family) burden on their partner.

These findings and the literature on the gender gap call for policy interventions aimed at reducing and redistributing care work. The rise of remote working during the COVID-19 pandemic has improved work–life balance conditions but widened gender inequalities in the labour market. In addition, the pandemic has increased awareness of the importance of care labour and care workers (UNDP, 2020), thus establishing a solid basis for advocacy of gender-equal care policies (Corsi & Ilkkaracan, 2022). An exogenous shock such as the pandemic has triggered the urgent need for policies to solve the unequal division of unpaid care work. Without these policy measures and responses (e.g., family-friendly benefits), the existing gender division of labour may be further intensified, along with gender job segregation and the gender earnings gap (UNDP, 2020). This would negatively affect the resilience of households against future health and economic crises. Interestingly, the Work–Life Balance Directive 2022 sets minimum standards for family leave and flexible working arrangements, and promotes equal sharing of caring tasks between parents. It also highlights the importance of providing pension credits for care-related career breaks (Eurofund, 2022b).

In Italy specifically, while workers were protected from dismissal, traditional measures such as parental leave were also strengthened during the emergency period. With the Cura Italia Decree, parental leave was temporarily extended to 15 days from 5 March to 3 April 2020, with the provision of 50% of normal salaries for families with children younger than 12. Likewise, the Rilancio Decree extended the support measures for several more months (Art. 72), granting a maximum of 30 days of special parental leave, including remote workers (Rymkevich, 2021). Despite their temporary nature, these emergency provisions for leave are meant to offset the lack of public and private educational and care services. Furthermore, it is worth emphasizing that Legislative Decree No. 105/2022, the so-called ‘Work–life conciliation decree’, has introduced several innovations regarding work–life balance for parents and carers, extending (among other things) the duration of parental leave, as well as the scope of application of priority access to smart working.

More broadly, the social policy literature has shown that although family-friendly benefits (e.g., flexible work schedules, child-care referrals, and leaves of absence) can help employees improve work–life balance, the availability of these benefits alone does not allow one to successfully balance career and family (Addabbo et al., 2018; Allen, 2001). Family-friendly benefits should be accompanied by a change in organizational norms and values regarding the appropriate interaction between work and family (Lobel & Kossek, 1996; Naldini & Saraceno, 2022).

Furthermore, telecommuting, flexible family-friendly work schedules and hours, and the opportunity to work in third places (i.e., coworking and hybrid spaces) (Mariotti et al., 2022a) should be promoted when WFH is not the best solution for workers. In addition, it is crucial that access be increased to quality social care services (e.g., childcare, elderly, ill, and disabled care, long-term care, education and health services). This issue has been explored and discussed within the 15-minute city concept, which focuses on the importance of urban rhythms and the quality of life in cities, thus promoting proximity-based services to inhabitants, including care services (Moreno et al., 2021). An exogenous shock such as the COVID-19 pandemic has underlined how the policies mentioned above should be promoted to increase awareness of the crucial role of care work in ensuring recovery.

A fundamental question related to our findings is the sustainability of lockdown-related effects. In more precise terms, as highlighted by Spurk and Straub (2020), does a lockdown imply a short or long-term impact on the labour force? Does the intensity of the COVID-19 experience influence these effects? Is it an exogenous shock with a (more or less rapid) return toward a steady state, or are we heading toward a ‘new normal’? While answering all these questions is beyond the scope of this short paper, the important point is that these experiences are likely to be affected by personal characteristics, and hence the use of micro-level data can be helpful in deriving insights. While some individuals may enjoy remote working (avoiding toxic workplaces), others could have a negative experience (overburdened by domestic work), or even experience a career shock (Akkermans et al., 2018).

In studying the pandemic’s long-term effects, we should also look for more significant shifts in career development and the impact of COVID-19 on individual career satisfaction and job security. At the same time, it may be too early to address these research questions, at least from an empirical perspective. A longitudinal research design must be implemented, with numerous observations (e.g., assessments) of individual data registered over several months or even several years (Spurk & Straub, 2020).

Although it sheds some light on the relationship between remote working and work–life balance satisfaction, the present analysis has some limitations. First, in regard to the adverse effects of COVID-19, Italy is a very peculiar case. Second, among advanced economies, Italy has a lower women’s employment rate, which is associated with family welfare provision, which in turn shifts more of a family caring burden onto women (Naldini & Saraceno, 2022; Riva, 2016). Third, the data we used were gathered at the end of the first wave of the pandemic, and right before the second, and this may affect the interviewees’ responses, and therefore our results too. Finally, the data did not allow us to account for the work–family diversity that exists across and within genders, which may be analysed through a more fine-grained intersectional approach. Hence, future studies may explore spatial and social variety across regions, countries, and timeframes.

ACKNOWLEDGEMENTS

The authors acknowledge with thanks the cooperation of the Associazione per lo Sviluppo dell’Industria nel Mezzogiorno (SVIMEZ) for making the data available for this study.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1 ‘Remote working’ is a general umbrella term that includes other flexible ways of working, such as teleworking, smart and agile working, and working from home (ILO, 2020a).

2 The analytical sample consisted of 55,204 adults living in the UK who were monitored for the strict 11-week lockdown period from 21 March to 31 May 2020 (Bu et al., 2021).

3 On these territorial differences, see the excellent empirical paper by Riccardi et al. (2020) based on the results of the INVALSI survey of 480,474 primary school students across the country.

4 The progressive closure of non-essential economic and institutional activities in Italy started with the Prime Minister’s Decree of 4 March 2020 (Gazzetta Ufficiale Serie Generale n. 55/2020). The COVID-19 pandemic in Italy was characterized by the following waves: (1) March–May 2020: first strict lockdown; (2) June–October 2020: ‘second phase’; and (3) November 2020–March 2021: second lockdown (Mariotti et al., 2022b).

5 The survey was conducted by Datamining on behalf of the Associazione per lo Sviluppo dell’Industria nel Mezzogiorno (SVIMEZ). The sample of 803 interviews with members of the Italian working population is stratified by gender, age and geographical area. The sample error is 3.53%.

6 In the original version of the questionnaire in Italian, it is ‘Equilibrio tra lavoro e vita privata’.

7 In the original, in Italian, this is ‘Ho continuato a lavorare regolarmente nella stessa regione della mia sede lavorativa, ma in smart working’ and ‘Ho continuato a lavorare regolarmente in smart working, ma in una regione diversa della mia sede lavorativa’.

8 This is nonetheless an interesting threshold because the population involves 803 observations.

9 To check the existence of an interactive effect we also ran different models adding an interaction term between the control variables. No result of interest was found.