A Comparison of the Interplay of Public and Occupational Work‒Family Policies in Austria, Denmark, Italy and the United Kingdom

ABSTRACT

This article analyses the interplay of public and occupational work‒family policies in institutionally different countries (Austria, Denmark, Italy and United Kingdom). Most of the existing studies do not analyse public and occupational family policy in conjunction, although this is necessary for a comprehensive understanding of family policy, and therefore the article adds knowledge on work‒family policy and the interplay of public and occupational based approaches. Based on a diverse case selection strategy and using comparative information from European statistics, surveys and reports, the crowding-out hypothesis is excluded, but no one consistent relationship is found for all countries. Instead, the article adds to existing knowledge that the country-specific public‒private mix depends on the institutional context (e.g. public family policy) and industrial relations.

Keywords:

• comparative family policy
• occupational family policy
• work‒family reconciliation
• working time
• parental leave
• secondary data analysis

Introduction

The boundaries between family and working life have been eroded over the last decades, due to economic and labour force changes such as higher women’s labour force participation and educational levels as well as the growth of the service sector (Appelbaum et al. 2006). Furthermore, often the salary of only one person in the household no longer guarantees a decent living standard. These socio-economic changes exert pressure not only on governments but also on employers to introduce policies that support the combination of work and family life.

Most of the studies related to work‒family policies analyse the role of the state and public family policies (e.g. Abrahamson et al. 2005; Lewis 2006; Morgan 2013; Adema and Ali 2014; Ferragina and Seeleib-Kaiser 2015; Sani 2018). Titmuss (1958) called for a greater consideration of occupational welfare activities in addition to public welfare benefits and services as they differ not only in their function or aims, but also in their administration. Therefore, they may duplicate and overlap public welfare. Most of the main contributions to the public‒private mix in social policy had focused on pensions (Rein and Rainwater 1986; Rein 1996; Rein and Wadensjö 1998; Rein and Turner 2001; Ebbinghaus 2011; Natali 2017) and also unemployment and health insurance (Hacker 2002; Seeleib-Kaiser et al. 2012). With regard to family policy, most studies so far have analysed public and occupational family policy¹ separately. Despite a large number of publications that focus on the description and explanation of occupational work‒family policy (e.g. Wood et al. 2003; Budd and Mumford 2004; Appelbaum et al. 2006; Heywood and Jirjahn 2009; Ravenswood and Markey 2011; Wiß 2017), only a few look at the overall public‒private mix. Adding to the literature, we are interested in what the interplay between the state and firm level looks like and whether we detect crowding-out or crowding-in effects (similar to Chung 2019).

According to Rein (1996), family policy is characterized by dual retreat (by the state and companies) on very low spending levels. In their overview, Appelbaum et al. (2006) mainly refer to occupational family policy in the US and only hint at the linkage between public and occupational family policy. In her qualitative case study on France (compared with the UK and US experience), Ollier-Malaterre (2009) found a negative relationship between public work‒family policies (high perceived state legitimacy and responsibility for family policies by human resource officers and employee representatives in companies) and employer activities (family-friendly working times and childcare) with regard to work‒family reconciliation (see also Evans 2002). However, her study does not cover Scandinavian and southern European countries, what is important to study and map variation due to the difference in the welfare state context. Furthermore, she does not refer to individual policies such as working time arrangements or leave programmes.

Den Dulk et al. (2012) could show that low levels of public social policy do not increase company activities with regard to work‒family reconciliation, based on data from 1999/2000 in 19 countries for large companies. Companies in conservative as well as social democratic welfare regimes adopted the most occupational work‒family policies. However, they took the respective welfare regime as a proxy for a generous welfare state and therefore did not measure the level of spending on public family policies. Furthermore, they focused on the mere number of employer-provided childcare schemes, leave and flexible work arrangements and not their quality, such as level and length of leave arrangements. A further study by Den Dulk et al. (2013), who explicitly analysed the public‒private mix in family policy, found that the national context (indicator consisting of parental leave, public childcare and flexible work regulations) matters for occupational work‒family policy. However, based on a company survey in 2004/2005, they could only analyse whether or not employers provide work‒family policies (flexible work, leave and childcare), but not how many employees have access or the details of these policies (e.g. level or length). Recently, Chung (2019) showed that generous public work‒family policies (measured as public family policy expenses, parental leave and formal childcare) crowd in schedule control arrangements (partial or total control over working hours) at the individual level. She looked at one specific form of occupational work‒family policy (schedule control) and in the model with several family policies only childcare crowds in access to schedule control.

Complementing these findings, our case study design allows us (a) to focus on both public and occupational work‒family policies by analysing the content of various public family policies and how additional occupational family policies are being provided, (b) to analyse beyond their sheer number the type of several working time arrangements and employees’ access as well as benefit levels of maternity, paternity and parental occupational leave arrangements including their coverage, and (c) to consider different countries accounting for differences in the institutional context within various approaches to family policy. Furthermore, we examine two possible factors that may influence the public‒private mix in family policy: institutional context (level of spending on public family policy as well as rules regarding leave) and industrial relations (collective bargaining coverage and trade unions).

We proceed by first presenting possible relationships between public and occupational family policy. After the selection of our countries, based on a diverse case selection strategy, we point to public work‒family policies. Then follows the analysis of firm-level work‒family policies and a discussion of the findings.

The Interplay of Public and Occupational Work‒Family Policy

Work‒Family Policy

By family policy, we mean those policies affecting, directly or indirectly, the lives of families (Saraceno 2013). They can thus consist of various instruments, for example direct economic support to families (Abrahamson et al. 2005), care for children, but also leave after birth and family-friendly working time arrangements. Although family policy may have different intentions over time and across countries, for example boosting fertility, supporting traditional family arrangements or preventing poverty (Saraceno et al. 2013), they can also reduce work‒family conflicts (Chung 2011). Furthermore, different types of public family policies have different outcomes for parents and children. Generous and gender-egalitarian policies result in more time for parents with their children, divided paid and unpaid work for mothers and fathers, and high child well-being (for details, see Gornick and Meyers 2004).

In general, occupational welfare – as public policies – can address and include work‒family policies such as financial benefits, leave, childcare and family-friendly working time arrangements (Evans 2002; Fleckenstein and Seeleib-Kaiser 2011; Greve 2012).

The Public‒Private Mix in Work‒Family Policies

From a rational choice perspective, employers adopt work‒family policies if the benefits exceed the costs (Appelbaum et al. 2006). In the business case, family-friendly policies reduce absenteeism, turnover and productivity thanks to higher satisfaction (Oswald et al. 2009; Den Dulk et al. 2012) and improve the attractiveness of a company and the retention of employees (Evans 2002; Wood et al. 2003; Budd and Mumford 2004; Fleckenstein and Seeleib-Kaiser 2011). The skill composition of the workforce is a further factor as employers provide work‒family arrangements more often to high-skilled than to low-skilled employees (Fleckenstein and Seeleib-Kaiser 2011; Wiß 2017). In terms of agency, employee representatives and trade unions can also have an impact on employer-provided work‒family policies (Wood et al. 2003; Budd and Mumford 2004; Fleckenstein and Seeleib-Kaiser 2011; Ravenswood and Markey 2011).

When analysing the mix of public and occupational family policies, the question arises whether there is, similar to other social policy fields, a relationship between the structure and level of public and occupational family policy. Similar to Künemund and Rein (1999), Motel-Klingebiel et al. (2005), Van Oorschot and Arts (2005), Den Dulk et al. (2012) and Chung (2019), we could expect three possible relationships.

(1) Crowding Out/Compensation

A strong welfare state may crowd out occupational policies or high levels of occupational policies compensate for a weak welfare state. In the former, high levels of public social policy are either a reaction to the failure of private arrangements to provide benefits or there is no need for firms to provide benefits if the state assumes responsibility (Etzioni 1995); and if employers already provide benefits to most employees, there is only a low need for additional public benefits in the latter (for a similar argument applied to pensions, see Rein and Turner 2001).

There is evidence from pensions and to lower extent also for health care. Countries with high public pensions (statutory health programmes) tend to have low coverage rates of occupational pensions (employer-provided health plans) and vice versa (Rein 1996; Ebbinghaus 2011). Similarly, public long-term care seems to crowd out private long-term care insurance (Costa-Font and Courbage 2015).

In terms of work‒family policy, companies make a business case and provide work‒family policies if public family policies are scarce (for France, see Ollier-Malaterre 2009), but they probably compensate the lack of public efforts only for certain employees and not the total workforce. According to Chung (2019), generous leave programmes crowd out employer-provided schedule control. However, we do not know the relationship with occupational leave programmes. Furthermore, crowding out takes place only at very high levels of public family policy, according to Evans (2002).

H1a:

High levels of public family policy crowd out occupational family policy.

H1b:

Low levels of public family policy (e.g. limited public leave) are compensated by high levels of occupational family policy (e.g. high occupational leave).

(2) Crowding In (Growth in Tandem)

Strong welfare states stimulate high levels of occupational welfare. The institutional structure and position related to the need for dual-earner families implies both public and occupational welfare. In such a relationship, it is less a competition between the state and companies about who provides what and to what extent, but rather a reinforcing process (for a similar argument for social trust, see Brewer et al. 2014), in which firm-level policies conform to the institutional context, set by public policies (Den Dulk et al. 2012). Companies conform to national institutions and policies because of (informal) pressure by institutions, policies and normative beliefs (DiMaggio and Powell 1983), such as responsibility for work‒family reconciliation. Public institutions and policies may set standards resulting in public demand for companies to follow. Mimetic isomorphism may drive crowding in when (private sector) companies imitate the practices of other, especially public, competing companies (DiMaggio and Powell 1983). In this case, companies provide high levels of work‒family policies imitating successful practices of other companies (Davis and Kalleberg 2006).

Although not directly related to occupational family policy, Künemund and Rein (1999) propose the idea of crowding in for intergenerational solidarity. Generous welfare states increase the level of giving and receiving between generations, because in the case of higher public benefits, older people have more to give and, in exchange, they receive more. There is evidence that public work‒family policies (mainly childcare) crowd in employer-provided schedule control, but only for high-skilled employees (Chung 2019). Companies may top up public leave payments, provide them for a longer time or offer additional benefits not covered by statutory legislation (Den Dulk et al. 2013).

H2:

High levels of public family policy encourage high levels of occupational family policy.

(3) Dual Retreat or Supplement

A weak welfare state in conjunction with low levels of occupational welfare characterize the third relationship. In the case of lacking public provision of childcare or leave programmes, some employers may partially fill the gap if a business case can be made (Appelbaum et al. 2006). The majority of employers, however, will not provide any work‒life policy due to the collective action problem that arises when companies with family-friendly instruments experience higher costs than their competitors. In the case study of Farnsworth (2004) about the UK, employer-provided childcare and extra-statutory maternity leave were rather low in the 1990s, despite a lack of public provisions.

Instead of low public and firm activities, it could also be argued that public and occupational policies for work‒family reconciliation are very low individually, but by dealing with different family measures, together they may reach medium levels of family policy. The state and companies share the responsibility to provide work‒family policies, i.e. they can also work interchangeably. An example would be a situation in which the state provides only low leave replacement rates and firms top benefits up in order to attract employees in times of skill shortages (business case), or the state takes responsibility for childcare and companies for parental leave. Occupational welfare in principle might be more supplementary than substituting.

H3:

Low levels of both public (e.g. low public leave) and occupational family policy (e.g. low working time flexibility) complement each other, indicating no systematic relationship.

In terms of inequalities, we can expect a higher degree of dualization and divisions between insiders and outsiders in countries in which the public‒private mix is characterized by weak welfare states together with low levels of occupational welfare that is tailored only to certain groups of employees, although this is not the core focus of the article (but see e.g. Seeleib-Kaiser et al. 2012; Chung 2018). However, welfare dualism is not unique to private-only welfare or varying degrees of public‒private mixes, but is also often inherent to public welfare (Seeleib-Kaiser et al. 2012).

In general, the relationship between public and occupational work‒family policies is affected by numerous factors. However, the literature has examined two broad possible conditions that contribute to the specific design of the public‒private mix: the level of public family policy including the dominating family model and industrial relations (Evans 2002; Fleckenstein and Seeleib-Kaiser 2011; Den Dulk et al. 2012, 2013; Wiß 2017). As explained above, the mechanism behind the level of public family policy is compensation or complementarity. From a structuralist perspective, occupational welfare is not independent of public social policy and thus is shaped by the government (Rein 1996).

In terms of agency, occupational welfare as employer preferences depend not only on the institutional context (public social policy), but also on interactions with other actors (Rein and Turner 2001; Trampusch 2013). Collective bargaining and trade unions (industrial relations) condition occupational pensions (Øverbye 1996; Trampusch 2013), determining whether they are a fringe or general social benefit. The findings for the effect of trade unions on occupational work‒family policies are inconsistent (Appelbaum et al. 2006), with some studies confirming an effect of trade unions (e.g. Budd and Mumford 2004; Ravenswood and Markey 2011; Berg et al. 2014) and others that could not find a positive relationship (e.g. Osterman 1995; Deitch and Huffman 2001). One explanation is that the agency effect depends on factors such as union power and the country’s dominating family model (Wiß 2017). If the male breadwinner model dominates and women’s main role is to take care of (small) children, employers presumably behave in line with this. Furthermore, in such an environment there is only a limited demand for additional or compensating occupational work‒family policies.

Selection of Countries and Work‒Family Policies

Austria, Denmark, Italy and the UK are chosen according to the diverse case selection strategy (Seawright and Gerring 2008). Large variation in the variables of interest and contrasting institutional settings increase the representativeness of the study. The countries differ in terms of their welfare regime, industrial relations, market economy and – most importantly – levels of public family policy. Each country represents a typical case for one cluster of the de-familization typology.

A familialistic system is “one in which public policy assumes – indeed insists – that households must carry the principal responsibility for their members’ welfare. A defamilising regime is one which seeks to unburden the household and diminish individuals’ welfare dependent on kinship” (Esping-Andersen 1999, p. 51). From this perspective (for earlier concepts of de-familization, see Orloff 1993; Lister 1994), the possibility to combine work and family life contributes to the independence of women and freedom of choices regarding lifestyle.

The cluster analysis of Cho (2014) suggests four groups based on the economic independence of women through employment (gender-employment gap, gender-wage gap) and freedom from care responsibilities (father-specific leave, spending on child- and eldercare): countries in the cluster of strong de-familization such as Denmark show high levels of women’s economic independence through employment and high levels of freedom from child- and eldercare due to public spending. Countries belonging to the cluster of moderate de-familization such as the UK are characterized by medium levels on the economic independence of women and on their care responsibilities. The economic independence of women in the cluster of limited de-familization (Austria) is at medium level and freedom from care responsibilities is low. Countries in the cluster of weak de-familization (Italy) follow traditional family models. The economic independence of women as well as the provision of public childcare is lowest.

According to other typologies that cluster countries with regard to family policies (Leitner 2003; Mischke 2011; Lohmann and Zagel 2016), our selected countries always fall into different groups confirming their variety with regard to the degree of de-familization. Although de-familization is not necessarily achieved only through the state (Saraceno and Keck 2010), almost all typologies only point to public family policy, i.e. there is a lack of integration of occupational welfare in order to understand overall work‒family policies.

One reason for a lack of this kind of research is the paucity of data. Therefore, our selection of work‒family policies is partially driven by the availability of comparative reports and data. Our focus is on policies facilitating the combination of work and family life and less on poverty prevention measures. We examine public spending for family policies based on data from Eurostat. Further dimensions are childcare, leave programmes and working time arrangements, because they are – according to the literature (e.g. Gornick and Meyers 2004) – important instruments facilitating work and family life and – from the parents’ perspective – they are the most mentioned measures to balance work and family life according to the Eurobarometer 2009 survey (Saxonberg 2014). Leave programmes grant parents time for childcare and substitute income from work, usually combined with the possibility to return to work after leave. Formal childcare enables parents, especially women, to work. It is true that flexible working time arrangements are a multi-faceted policy field where actors, especially employees and employers, pursue different interests. Therefore, we only refer to employee-friendly working time arrangements such as working time sovereignty that can facilitate work‒life balance, as well as the possibility to take time off during working hours that is important for work‒family reconciliation in the case of unforeseeable family matters. We do not mean employer-friendly working time arrangements such as shift work, acknowledging that there might be overlaps (for details, see Gregory and Milner 2009; Chung and Tijdens 2013).

Public Spending and Formal Childcare

We refer to the purchasing power approach² divided by the population below the age of 15 for the comparison of public spending on family policy (Table 1). Overall spending is highest in Denmark and Austria, followed by the UK and Italy. With regard to the distribution of expenses, Denmark spent 61 per cent on benefits-in-kind – reflecting its strong focus on public care policies – much more than Austria (27 per cent) and the UK (21 per cent). In Italy, a country with limited de-familization and low dual-earner support, cash benefits dominate and only 15 per cent of all public expenses on family policy are spent on benefits-in-kind.

Table 1. Public spending on family policy in Purchasing Power Standard (PPS) per person below the age of 15 in 2014

	All Expenses	Cash Benefits	Benefits-in-kind
Austria	6,897	5,014	1,883
Denmark	6,959	2,706	4,253
Italy	2,940	2,503	437
UK	4,598	3,629	969

Sources: Eurostat, spr_exp_ffa and demo_pjanbroad (accessed 26 October 2017).

Notes: Age group below 15 is used as children often can get benefits up to that age.

Childcare has been one of the most prominent reform topics in many countries in recent years (Morgan 2013). In 2016, 70 per cent of children below the age of three were covered by formal childcare in Denmark, followed by Italy (34 per cent) (Figure 1). In contrast, childcare was lowest in Austria (21 per cent). Almost all places in Denmark cared for children for 30 hours or more – supporting the dual full-time earner model – whereas in the UK and Austria institutions offering less than 30 hours dominated, supporting the full-time/part-time earner model.

Figure 1. Formal childcare for children in 2016 (% of the age group)

Sources: Own calculations based on Eurostat EU-SILC.

Notes: Formal childcare includes education at pre-school or compulsory education, centre-based services or day-care centres organized by a public or private structure. Formal childcare is only a proxy, because there are no comparable data only for public childcare as a clear separation of public and private often is not possible. Some countries finance childcare, some subsidize employer or private formal childcare and some have tax measures (see also Blome 2017). National sources for Italy, for example, show lower numbers than our data. However, the country differences are rather robust when using different sources corroborating our arguments.

Coverage rates for children from three years to minimum compulsory school age were much higher than for children below three years. Formal childcare was near-universal in Denmark and very high also in Italy (93 per cent) – providing mainly full-time care – compared with only 73 per cent in the UK. In terms of coverage and duration of public formal childcare, Denmark clearly supports the dual full-time earner model independent of children’s age, and Italy for families with children older than three. The low UK coverage rate for children from three years on in conjunction with mainly part-time care only allows for the one full-time and one part-time earner model.³

Public Leave and Working Time Regulations

Austria provides fully paid maternity leave for 16 weeks and, in terms of parental leave, parents can choose between different models (Table 2). The models with long leave periods and low payments support the traditional female caregiver and the model with a short leave period and income related benefit is appropriate for the dual-earner model. In terms of family-friendly working times, parents with children are entitled to change from full- to part-time work (and back), until the child turns seven or until the child enters school. Parents also have the right to take two weeks fully paid leave per year to care for sick children (Rille-Pfeiffer et al. 2017).

Table 2. Statutory leave, 2017

	Maternity leave	Paternity leave	Parental leave
Austria	Yes, 16 weeks, earnings related, 100%	No	Yes, (a) €15,449 if mother and father take leave (granted for 15 to 35 months) or (b) up to €12,366 if only one takes leave (granted for 12 to 35 months) or (c) earnings related, 80% with a ceiling, for 12/14 months
Denmark	Yes, 18 weeks, earnings related, 90% with a ceiling	Yes, 2 weeks, earnings related, 90% with a ceiling	Yes, 32 weeks, earnings related, 90% with a ceiling
Italy	Yes, 20 weeks, 80% or more	2 days, 100%	Yes, maximum 10 months, unless father takes 3 months, then 11 months, earnings related, 30% for 6 months (part-time possible)
UK	Yes, 52 weeks, 6 weeks earnings related, 90% without ceiling, 33 weeks up to 90% or flat rate (whichever is lower), 13 weeks unpaid	Yes, 1 or 2 weeks, flat rate of £141 or 90% average weekly earnings if less; up to 50 weeks of maternity leave can be transferred to the father (Shared Parental Leave)	Yes, 18 weeks per parent (not continuously), no payment

Sources: www.leavenetwork.org.

In Denmark, all three public leave programmes substitute 90 per cent of the former earnings (with a ceiling) and are paid for 18 weeks for the mother, 2 weeks for the father and a further 32 weeks that parents can split. No state regulation exists with regard to family-friendly working time arrangements, but parents benefit from collectively agreed regulations (see the next section).

Parents in Italy can stay at home for up to 11 months (parental leave), but the replacement rate of 30 per cent is rather low. Parents are entitled to apply to work flexibly as long as their children are below six years old, but the employer can refuse it. Furthermore, similar to Austria, they have the possibility to take time off for childcare – but unpaid – unlimited for children below three years old and five days for children aged three to eight (Addabbo et al. 2017).

In the UK, maternity leave is paid for 39 weeks, open also for men. Similar to Italy, working parents of children under six have had the right to request flexible working time since 2003, remarkable for a country with low public dual-earner support. This regulation was extended in 2014 to all employees, but the employers can refuse the request for clear business reasons (O’Brien and Koslowski 2017). Furthermore, employees can take a reasonable amount of time off to deal with emergencies of dependants.

The gap between the end of well-paid leave⁴ and the start of entitlement to formal childcare hints at the need and expectation for additional occupational family policy in order to give both parents the opportunity to equally share paid work and unpaid care. In 2017, well-paid leave ended after 14 months in Austria and parents of children aged five are entitled to part-time childcare resulting in a 36-month gap between the end of leave and start of childcare entitlement, and even a 46-month gap for well-paid leave (for details for all countries, see Blum et al. 2017). The respective gaps were 22.1 months for leave and 34.6 months for well-paid leave in the UK. There was no entitlement to formal childcare in Italy in 2017. Therefore, parents are forced to find care solutions by themselves after 14.8 months of leave and 3.8 months of well-paid leave (until school age). In contrast, no gap exists in Denmark with the end of well-paid leave after 11.2 months and an entitlement to full-time childcare for children aged six month.

Occupational Family Policy

Family-Friendly Working Times

Flexible working hours are one of the most important measures for the balance of work and family life according to a recent Eurobarometer survey (Saxonberg 2014). There are only minimum state regulations in Austria, Italy and the UK, and no state regulation at all in Denmark, leaving room to manoeuvre for employers. In the UK, 69 per cent of all firms offered their employees the possibility to adapt the start and finish time of the working day according to their personal needs or wishes (flexible working time) (Table 3). The high percentages in the UK may partly result from the right to request flexible working time, whereas the 87 per cent of firms offering flexible working times in Denmark are based on voluntary actions or collective agreements. If firms offered flexible working times, 45 per cent of all firms in Denmark covered 80–100 per cent of all employees whereas only 32 per cent of all firms did so in the UK. Apart from working time flexibility, firms in Austria and Denmark offered the most generous arrangements with regard to the possibility to accumulate hours to take days off (95 and 88 per cent) – for example for unforeseen or planned family activities. The Danish social partners negotiated that most employees are entitled to take one day off to care for a sick child and even two days if they work in the public sector (Bloksgaard and Rostgaard 2017). Family-friendly working time arrangements are included in many Danish collective agreements across different sectors (for an overview, see Eurofound 2006). Flexible working time arrangements are also a prominent issue in the chemical industry, the financial sector and trading in Italy (European Commission 2005; Eurofound 2006). Due to the statutory entitlement to request flexible working times in the UK, additional family-friendly working time arrangements exist only in some firms and, if so, are agreed at company level and not for the entire sector (Eurofound 2006). The incidence of collective agreements containing work‒family arrangements in Austria is also below the European average (Eurofound 2006).

Table 3. Availability of family-friendly working time arrangements, 2013 (% of all firms)

	Flexible working time	80–100% of employees entitled to flexible working time^a	Accumulation of overtime for half/full days off^a
AT	80.8	40.8	94.9
DK	86.7	45.0	87.5
IT	66.3	40.8	51.4
UK	68.5	32.1	57.9

Sources: European Company Survey 2013 (Eurofound 2015).

Notes: Only firms with at least ten employees, weighted results. a = % of all firms who provide flexible working time.

The results of the company survey may overestimate occupational family policy, because only firms with at least ten employees were asked and because firms aim to be labelled “family-friendly”; therefore they may be answering that they provide family-friendly arrangements although they do not or only to few employees. Based on an employee survey – and similar to the employer survey – Denmark is the most family-friendly country and Italy the least (Table 4). In Italy, for more than two-thirds of employees the working times were set by the employer without any possibility for changes. In Denmark, 12 per cent of the employees determined their working hours entirely by themselves and 51 per cent could choose between fixed working schedules or adapted their working hours within certain limits, enabling a better work‒life balance. With regard to taking an hour or two off during working hours to take care of family matters – important for unforeseeable family issues – the UK ranked second (73 per cent), after Denmark (74 per cent). In addition to generous public family policies, many Danish firms provide family-friendly working time arrangements, whereas firms in Italy do not seem to substitute low levels of public family policies.

Table 4. Family-friendly working time arrangements: the employees’ perspective, 2015 (% of employees)

	Working time arrangement			Flexible starting/finishing working time	Take off hours for family^a
	Self-determined working time	Other working time settings	Working time set by employer
AT	10.1	35.5	54.2	45.5	62.7
DK	12.2	51.0	36.4	43.6	74.4
IT	3.8	23.1	72.4	27.2	58.2
UK	5.3	34.6	59.7	36.7	73.0

Source: European Working Conditions Survey 2015 (Eurofound 2017).

Notes: Only employees, weighted results. (a) For you arranging to take an hour or two off during working hours to take care of family matters is very or fairly easy.

Occupational Leave

Employer-financed leave is the second examined occupational family policy, because extending leave periods and increasing their benefit levels is for almost half of all respondents of the Eurobarometer survey 2009 a measure with priority for work‒family reconciliation (Saxonberg 2014). In general, employers in Austria do not provide any extra-statutory leave benefits, only some firms substitute the lack of public paternity leave. In the public sector, the social partners agreed upon four weeks of leave, but unpaid (Rille-Pfeiffer et al. 2017).

In line with the crowding-in hypothesis, generous public leave benefits and widespread childcare provisions go hand in hand with extensive firm-level activities in Denmark. Occupational family policies are an important issue in collective bargaining (European Commission 2005; Eurofound 2006). In addition to the public leave programme, 84 per cent of all employees profit from extra-statutory leave based on collective agreements. The employer usually pays the difference between the public benefit and the former income for persons on maternity and paternity leave, thereby achieving a 100 per cent replacement rate up to 26 weeks (Bloksgaard and Rostgaard 2017). Due to their collective status, occupational work‒family policies in Denmark are more a general social benefit than a fringe benefit.

Although only a few collective agreements cover parental leave and childcare in Italy (Eurofound 2006), employers fill the gap of on average 20 per cent between the public maternity leave benefit and the former salary in the public sector, and for some occupational groups based on collective agreements (for information about extra-statutory leave, see Addabbo et al. 2017; European Commission 2005). This is quite generous, although paid for only 20 weeks. With regard to parental leave, only public employers pay the full salary for 30 days. Social partners in northern Italy – with much higher employment rates of mothers compared with the South – agreed to implement (employer-provided) initiatives addressing working parents (e.g. flexible working hours and training for mothers after return to work) (Eurofound 2006).

Only a few firms in the UK complement the statutory maternity leave. According to the Work‒Life Balance Employer Survey 2013 (Department for Business, Innovation and Skills 2014) – covering firms with five or more employees – 13 per cent of all firms covering 36 per cent of all employees pay at least some of their employees during occupational maternity leave for on average 21 weeks. In terms of paternity leave, 27 per cent of all firms covering 33 per cent of all employees offer paid or unpaid extra-statutory paternity leave for on average 4.8 weeks. In addition, 73 per cent of firms not offering any occupational paternity leave nevertheless pay the full salary for the statutory two weeks (Department for Business, Innovation and Skills 2014). The vast majority of parents in the UK relies only on the public maternity leave scheme that partly covers parental leave due its length of 52 weeks, of which only the first six weeks really substitute the former salary. The UK is the only country where occupational childcare plays a role.⁵ Around 8 per cent of all firms covering a quarter of all employees provide some form of childcare (European Commission 2005).

Discussion

With few exceptions, occupational family policy in Austria, Italy and the UK is provided by individual employers and not – as in Denmark – on widespread collective agreement, which makes work‒family policies in these countries more an instrument to attract (skilled) employees rather than a general social benefit. Strong trade unions and high collective bargaining coverage in Denmark and rather weak unions in the UK may provide an explanation (see Table 5).

Table 5. Industrial relations 2013

	Collective bargaining coverage (%)	Trade union density (%)
AT	98	27
DK	84	67
IT	80^a	37
UK	30	26

Source: Visser (2016).

Notes: a = 2010.

However, occupational work‒family policies exist only for a few employees in Austria and Italy – despite high coverage rates of collective bargaining and only medium levels of public family policy. Work‒family issues are of higher priority for unions in Denmark with a large public and service sector compared to unions in Austria with a strong male-dominated industrial sector (Wiß 2017).

Although Den Dulk et al. (2012) also confirm the crowding-in effect, our findings contrast some of their results, because Denmark, belonging to the social democratic regime in their study, clearly has the highest level of occupational family policy across our countries and not conservative welfare states (as in their study). Similar to their findings, the social democratic regime (Denmark in our study) is associated with the largest number of flexible work arrangements. Confirming what Den Dulk et al. (2013) found for the number of occupational work‒family policies provided, we could show that high levels of public family policy are also positively related to employee access and quality of these policies (level of benefits). The results of Chung (2019) are challenged only by our Danish case. We could not confirm the crowding-out effect of generous leave entitlements on schedule control. Instead, we even found a crowding-in effect of public leave arrangements on occupational leave policies. It also disproves Evans (2002), who argued that very high levels of public family policies crowd out occupational welfare.

Conclusion

The aim of the article was to analyse the interplay of public and occupational work‒family policies. Our contribution is the analysis of the relation between different public family policies (spending, childcare, leave and working time regulation) and various occupational family policies (working time arrangements and leave) in four diverse countries. In contrast to mainly large-n studies on that topic, our comparison of four countries allows for a detailed investigation of the content of public as well as occupational family policy considering the national context. Our results – with a limited number of countries – lend support to the crowding-in hypothesis. In Denmark, high levels of public family policy encourage high levels of occupational family policy, because employers are faced with strong trade unions and widespread collective agreements in an environment that clearly supports the dual-earner model. We could not find evidence for the crowding-out and compensation hypotheses that high levels of public family policy suppress employer activities and that low levels of public family policy excessively incentivize employers to step in.

The third hypothesis that low levels of both public and occupational family complement each other is only partially supported by the Austrian and UK cases hinting at the lack of a systematic relationship. Medium levels of public family policy in Austria encourage employer activities only to a limited extent. Although trade unions are rather powerful and collective bargaining is universal, the still dominant male breadwinner family model in Austria – although currently changing – prevents unions calling for more widespread work‒family policies. In such an environment, employers mainly provide policies only to certain (skilled) employees. Similarly, public and occupational activities sum up to only medium levels of family policy in the UK, although for different reasons. The job of the state has traditionally been poverty prevention for families, and employers and the market have been responsible for additional measures facilitating work‒family reconciliation. Trade unions are not powerful enough to push employers for more comprehensive measures and – in an environment that is more progressive in terms of gender roles than Austria and Italy, but less than Denmark – employers implement such policies voluntarily only to certain employees.

Italy suggests a retreat by both the state and companies. Low levels of occupational family policy do not compensate for low levels of public family policies. In a context where the male breadwinner model is prevailing in the policy culture, employers seem to conform to it, increasing the barrier for mothers to work. However, adverse economic conditions probably are a further factor that conditions rather low firm (and public) activities. In sum, there is not one consistent relationship between public and occupational work‒family policy across the four countries. Hence, we cannot really confirm or reject only one hypothesis.

The article does not provide a causal analysis, but a deeper understanding of the interplay of public and occupational work‒family policies including possible mechanisms. Based on two important factors that determine the country-specific public‒private mix, the relationship depends on the institutional context (e.g. public family policy) and industrial relations.

Future typologies of family policy are well advised to focus not only on public family policy, but to integrate occupational policies in order to get a more complete picture of a country’s family policy. High levels of public family policy may mask an even higher level of overall family policy when considering employer activities, and inequalities in access to work‒family arrangements become visible if only some employers provide family-related benefits or services, or only to some employees. Future studies need to test our findings for a larger country sample, extend the dimensions of occupational family policy beyond time issues and leave programmes and analyse in more detail the mechanisms that explain the varying levels of the public‒private mix (e.g. agency).

In terms of public policy, the state cannot rely on and refer only to (voluntary) employer activities if its aim is to increase the combination of work and family life. Instead, it is more likely that companies will assume responsibility if the state sets a good example and increases institutional pressure.

Additional information

Funding

This work was supported by the Austrian Science Fund (FWF [grant M 1565-G16]).

Notes on contributors

Tobias Wiß

Tobias Wiß is Assistant Professor at the Johannes Kepler University Linz, Austria. Prior to joining JKU, he held postdoctoral appointments at the Hertie School of Governance Berlin, the Johannes Kepler University Linz and at the Mannheim Centre for European Social Research (MZES).

Bent Greve

Bent Greve is Professor in Social Science at the University of Roskilde, Denmark. His research interest focuses on the welfare state, and social and labour market policy, often from a comparative perspective. He has published extensively on social and labour market policy, social security, tax expenditures, public sector expenditures and financing of the welfare state. He is editor of Social Policy & Administration.

Notes

1. By occupational family policy, we mean voluntary employer-provided extra-statutory policies on top of public policies that facilitate work‒family reconciliation. We use “occupational family policy” and “firm-level policy” interchangeably as the employer is the provider of benefits and/or services. We acknowledge that they can be negotiated at different structural levels (firm, industry and country) and may differ by industry, but the available comparable data did not allow considering this.

2. The purchasing power standard is an artificial currency unit that considers price differences across countries.

3. Dual-earner couples are eligible for 30 hours free childcare for three- to four-year-olds in England since September 2017 (O’Brien and Koslowski 2017).

4. According to Blum et al. (2017), well-paid leave is defined as half or more of leave period paid at high flat-rate (≥ 1.000€/month) or ≥ 66 per cent of earnings.

5. Occupational childcare is not analysed for all four countries due to missing data and information. It hardly plays a role in Austria, Denmark and Italy.