Dualization and the access to occupational family ‐ friendly working ‐ time arrangements across Europe

This article examines outsiders' relative access to occupational level family ‐ friendly policies. I use data from the European Working Conditions Survey of 2015 across 30 European countries examining workers' access to two types of family ‐ friendly working ‐ time arrangements (WTAs): flexitime, and time off work for personal reasons. The article focuses on women with care responsibilities given that their demands for family ‐ friendly policies, as well as their outcomes, have been shown to be distinct from the rest of the working population. In addition to the outsider definition used in the labor market dualization and occupational segmentation literature, i.e., low ‐ skilled workers and those without a permanent contract, this article also defines outsiders as those perceiving their job as insecure. The results of the analysis show a segmentation between workers in their access to family ‐ friendly policies. Unlike statutory policies, occupational policies seem to be selectively provided mostly to workers where employers have a vested interest, i.e., insiders, resulting in a dualized system for most countries. However, rather than their contract status, the skill ‐ level of the job/workers, and their perceived insecurity were found to be important. The results further show that although Northern European and some continental European countries are those where family ‐ friendly WTAs are more readily available, it is in these countries where the division between insiders and outsiders is the greatest. The results of the article contribute to the literature by showing a need to move beyond the national level when examining family ‐ friendly policies, and to examine a more diverse definition of outsiders when examining dualization of working conditions.

& Haley-Lock, 2004). In addition, companies may provide various additional arrangements through occupational policies which are not set out in the national level agreements that are crucial in addressing reconciliation needs of workers (Davis & Kalleberg, 2006;Farnsworth, 2004;Kelly et al., 2014). Yet, studies examining occupational welfare have traditionally focused on occupational pensions and other types of arrangements as noted in the introduction to this issue (see also Seeleib-Kaiser & Fleckenstein, 2009).
The question raised in this article is who has access to occupational level family-friendly policies, and whether we find a segmentation in the labor force in the access. Unlike statutory policies, companies have more discretion over whether to provide occupational family-friendly policies, as well as to whom to provide them. Companies may provide additional arrangements to ensure the recruitment and maintenance of workers with additional family demands, as well as to maintain workers' loyalty and commitment (Dex & Scheibl, 2001;Wood, De Menezes, & Lasaosa, 2003).
The use of such occupational policies may be performance-driven (Ortega, 2009), resulting in the stratification within organizations in the access workers get to family-friendly policies (Chung, 2017a;Lambert & Haley-Lock, 2004). In other words, employers may only provide family-friendly policies to the select few that employers expect a return from (Swanberg, Pitt-Catsouphes, & Drescher-Burke, 2005) or those they are willing to keep/recruit (Wood et al., 2003). Using the dualization literature (Emmenegger, Häusermann, Palier, & Seeleib-Kaiser, 2012;Schwander & Häusermann, 2013), and organization segmentation literature (Lambert & Haley-Lock, 2004), this article examines the access to occupational family-friendly policies of insiders vs. outsiders. In addition to the more commonly used definition of outsiders, i.e., those without a permanent contract and low-skilled workers, the article also defines outsiders as those who feel insecure about their jobs, given the importance of subjective perceptions in detecting the complex condition the worker is actually placed in.
Of the various types of family-friendly policies, this article focuses on two working-time arrangements (WTAs), i.e., flexitime and time off work for personal reasons. These family-friendly WTAs are unique in that they are used both for performance-enhancing and work-family integration goals (Goodstein, 1994), are crucial for women in maintaining their labor market position after childbirth (Chung & Van der Horst, 2018), and are the most commonly used type of occupational level family-friendly arrangements across Europe (Chung, 2017b). The article examines how national level contexts shape workers' access to, and the division between workers in their access to, family-friendly WTAs.
Lastly, this article contributes to the literature by focusing on women with care responsibilities. Previous literature on occupational family-friendly policies has examined the working population as a whole. We know from the literature that women are much more likely to be responsible for household and care roles (Dotti Sani & Treas, 2016), and that the use of, and demand for, family-friendly arrangements is different for men and women (Singley & Hynes, 2005), and for those with and without care responsibilities (Wanrooy et al., 2013). Thus, it is important to examine women with care demands separately to understand how their work-family integration demands are met through occupational policies.
The main questions addressed in this article are, therefore, which female workers with care demands are able to access family-friendly WTAs, whether there is an access gap between insiders and outsiders, and whether it varies across countries due to institutional contexts. These questions will be answered through the use of the European Working Conditions Survey (EWCS) of 2015 and a multilevel approach. Section 2 explains what is meant by family-friendly WTAs, and examines key literature and theories on the provision of these arrangements. Dualization and organizational segmentation theories are examined to help us understand the segmentation in the access to family-friendly arrangements provided at the occupational level. Section 3 examines the data and variables used, as well as the methodologies applied in the article. Section 4 presents the analysis results, and section 5 makes some final concluding remarks and suggestions for future studies.
2 | THEORY 2.1 | Defining family-friendly working-time arrangements In this article, I examine two types of flexible WTAs: 1 flexitime-the ability to alter the times workers start and end work, which can lead to the ability to change the number of hours worked; and time off work, i.e., the ability to take time off during working hours to meet personal demands. Work-family border theory (Clark, 2000) and boundary management theory (Kossek, Lautsch, & Eaton, 2005) argue that flexibility in your work can help facilitate the integration of work and home roles by allowing workers to adapt the borders of one domain around the demands of others-in this case, adapting the timing of work around family demands. In fact, several empirical studies show how flexible working is used, especially by women, to address various family demands (Maume, 2006;Singley & Hynes, 2005). Further, flexible working has been shown to relieve work-family conflict, i.e., the conflict workers feel due to the competing demands coming from work and family (e.g., Chung, 2011;Kelly et al., 2014), although others have argued that the effect is minimal (Allen, Johnson, Kiburz, & Shockley, 2013;Michel, Kotrba, Mitchelson, Clark, & Baltes, 2011). Flexible working-time has also been shown to reduce labor turnover or turnover intentions of workers (de Menezes & Kelliher, 2011) by increasing the fit between work and family life (McNall, Masuda, & Nicklin, 2009). In fact, family-friendly WTAs are frequently used by employers as a recruitment and maintenance tool to attract and keep workers-especially those with higher skills (Davis & Kalleberg, 2006). Flexible WTAs are used not only to increase family friendliness of the company, but also to enhance its performance (Ortega, 2009). Giving workers more control over their work can be used as a part of a high-involvement system (Wood & de Menezes, 2010) or high performance strategy which aims to increase performance by giving workers more discretion and influence over their work (Appelbaum, 2000;Davis & Kalleberg, 2006). Providing workers with more control over their work has been shown to make workers work harder and longer (Kelliher & Anderson, 2010;Lott & Chung, 2016) and have been linked to increased performance outcomes for companies (de Menezes & Kelliher, 2011).

| Dualization and division of working conditions between workers
The main idea behind dual labor market theory (e.g., Doeringer & Piore, 1975;Lindbeck & Snower, 1989) is that labor markets are divided into primary and secondary sectors, with limited mobility between the two (see Davidsson & Naczyk, 2009). The main purposes of this division is to keep a primary workforce that secures the core skills of the company, while relying on the secondary market to adjust to cyclical demands and fluctuations. Thus, workers in the primary workforce, the "insiders", enjoy high wages, good working conditions, prospects for career advancement, and job stability. Workers in the secondary market, the "outsiders", have jobs with low-pay, bad working conditions, few career advancement prospects, and work that is unstable with risks of frequent lay-offs (Doeringer & Piore, 1975, pp. 70-71;Rueda, 2014). It is not surprising that outsiders are less likely to get investment from employers, such as work-related training (Arulampalam & Booth, 1998;Lambert & Haley-Lock, 2004), most likely because employers do not expect high returns from this section of the workforce (Osterman, 1999).
It is important to note the different definitions of outsiders in the literature. Some of the earlier works on dual labor markets distinguish between those in and out of employment (e.g., Blanchard & Summers, 1987;Lindbeck & Snower, 1986). More recently, the focus has been on the relative vulnerability of workers, outsiders defined as those with atypical contracts, i.e., temporary and involuntary part time, and non-unionized workers (e.g., Emmenegger, 2009;Rueda, 2005). Some scholars (Biegert, 2014;Schwander & Häusermann, 2013) argue that definitions of outsiders based on employment status are too static, and that the fluidity of positions and mobility between the two segments should to be taken into account. Schwander and Häusermann (2013) thus propose to use occupational categories to indicate employment biographies. Lastly, Chung (2016) argues that rather than objective insecurity statuses, subjectively perceived insecurity may be a better indicator of the actual situation the worker is placed in. Although objective insecurity is closely related to subjective insecurity, the latter relates to the psychological reactions to the worker's job insecurity, which can be affected by personal, organizational, and institutional contexts (Chung & Mau, 2014;Greenhalgh & Rosenblatt, 1984). In other words, the same permanent contract status may actually entail very different levels of security and, in many cases, negotiative power of the worker across countries (Chung, 2016), across organizations (de Cuyper, Sora, de Witte, Caballer, & Peiró, 2009), and even across different workers within the same organization. For the purpose of this study, I define outsiders as workers without permanent contracts, those in lower (and specific) skill occupational groups, and those who feel that their job is insecure. I exclude part-time work from the definition because, in many cases, in Europe, part-time work is taken up for voluntary reasons, and due to the limitations in the data, it is not possible to distinguish between voluntary and involuntary part-time workers.

| Outsiders and access to family-friendly working-time arrangements
Who gets access to family-friendly WTAs? Unlike statutory policies, where workers' access to family-friendly policies is guided by law, and limiting access may come with legal consequences, provision of occupational level family-friendly policies will largely depend on employers. Previous studies (Lambert & Haley-Lock, 2004;Swanberg et al., 2005) discuss three principles employers use to decide who gets access to family-friendly WTAs; namely, need, equity, and equality. When employers are genuinely interested in addressing the work-family needs of workers, those with the most family demands or most in need of family-friendly WTAs are likely to get access. On the contrary, when employers are more interested in the enhanced performance/outcomes gained from introducing the arrangements (the principle of equity), the workers who companies believe they will reap more benefit out of by providing familyfriendly WTAs will have access. Lastly, when the equality principle takes precedence, access to WTAs will be provided to all workers equally regardless of their care needs or potential performance outcome.
When companies mainly provide flexible WTAs based on the principle of equity, we can assume that outsiders will be less likely to have access (Swanberg et al., 2005). Employers do not usually invest in the outsider segment of their labor force. Furthermore, when providing family-friendly WTAs as an incentive for employees as a recruitment/maintenance tool, employers are less likely to use such arrangements for the outsider segment of their workforce (Wood et al., 2003) since they are more easily replaceable. Thus I expect that outsiders are less likely to have access to flexible WTAs (H1).
Empirically, numerous studies show that high-skilled and higher-educated workers are most likely to have access to flexible WTAs (Brescoll, Glass, & Sedlovskaya, 2013;Golden, 2009;Ortega, 2009;Seeleib-Kaiser & Fleckenstein, 2009). Workers in disadvantaged positions-e.g., low wage, low-skilled, lower educated-are least likely to have such access (e.g., Chung, 2017a;Golden, 2009;Swanberg et al., 2005;Wiß, 2017). Similarly, those with fixed-term contracts have also been shown to have less access to flexible WTAs (Präg & Mills, 2014), although other studies say there are no significant differences (Chung, 2017a). There is no research to the author's knowledge regarding how subjective insecurity relates to access to family-friendly WTAs. What is more, most of the studies on the division in the access to family-friendly arrangements have been single country case studies, leaving the question whether this division is different across different regimes.

| Cross-national variation in the division between insiders and outsiders
Dualization theorists argue that certain institutional configurations make it easier for dual labor market patterns to emerge (Emmenegger et al., 2012;Palier & Thelen, 2010;Rueda, 2005). Accordingly, the degree of segmentation within the labor market, and the insider/outsider divide, varies across welfare regimes where different protection mechanisms are in place (Biegert, 2014;Chung, 2016;Schwander & Häusermann, 2013). Scholars (Schwander & Häusermann, 2013;Yoon & Chung, 2016) also argue that due to these institutional and other social-economic contexts the groups that are most likely to be categorized as "outsiders" are different across countries and welfare regimes, as does the degree to which they are outsiders and insiders. Similarly, I expect a cross-national variation in the gap between insiders and outsiders in their access to flexible WTAs (H2).
Which countries are more likely to have a larger access gap of family-friendly WTAs between insiders and outsiders? There are no established theories on the variation in the segmentation patterns in the access to family-friendly arrangements across countries. Thus I borrow existing theories that examine the broader issue of variation in the labor markets' segmentation patterns. One key argument maintained by dualization scholars is that corporatist countries with stronger and centralized unions are those where the division between insiders and outsiders is largest. According to the power resource theory, the power that is mobilized by wage earners can influence welfare state development (Korpi, 1989). Similarly, the power of unions may influence employers in providing family-friendly benefits at the company level. Several studies have shown that collective bargaining coverage rates and union density are positively correlated to the provision of family-friendly WTAs at the company level (Berg, Appelbaum, Bailey, & Kalleberg, 2004;Lyness, Gornick, Stone, & Grotto, 2012;Präg & Mills, 2014). Further, Palier and Thelen (2010) argue that traditionally strong coordinated unions contributed to the diffusion, generalization, and institutionalization of good working conditions to the wider population, reducing inequalities between different groups of workers. Based on this, we would expect a smaller access gap between insiders and outsiders in countries with strong and centralized unions. However, Palier and Thelen (2010) also note that, in recent years, in the midst of liberalization and external economic pressures, dualism has been especially prevalent in these corporatist countries. On the one hand, in corporatist countries unions were successful in protecting insiders from the pressures of labor shedding strategies through negotiations with employers.
However, they also allowed employers to increase flexibility on the secondary market, exposing outsiders to increased insecurity in a so-called "dual reform" (Davidsson & Emmenegger, 2013;Ebbinghaus & Eichhorst, 2007;Palier & Thelen, 2010). Empirically, Biegert (2014) shows how compared to liberal countries such as the United Kingdom, corporatist countries such as Germany have stronger structural barriers between the insider and outsider markets, with less mobility between the two markets. Similarly, we could expect that countries with stronger, more centralized unions, i.e., corporatist countries, to be where the access gap of family-friendly WTAs between insiders and outsiders is the largest. Family policies at the national level may also be relevant in explaining who gets access to family-friendly WTAs. The "crowding out" theory argues that national-level policies will crowd out lower level welfare engagements (Etzioni, 1995), i.e., companies will not provide company-level family-friendly policies when there are generous policies at the national level. The counter argument to this comes from the "crowding in" theory (e.g., Künemund & Rein, 1999;Van Oorschot & Arts, 2005), which argues a positive, rather than negative, relationship between generous national-level policy and occupational welfare. This can be because institutions, laws, and policies may put pressure on organizations to become similar to national institutions (DiMaggio & Powell, 1983). The pressure may be in the form of normative isomorphic pressure, i.e., national-level policies changing the norm and subsequent public demand for companies to be more family-friendly (den Dulk, Groeneveld, Ollier-Malaterre, & Valcour, 2013), or mimetic pressure, i.e., where companies imitate or mimic the practices of other (successful) organizations (Been et al., 2017;Davis & Kalleberg, 2006). Based on this theory, we can expect that when there are generous national-level family policies, this will raise the benchmark and change the culture companies operate in, making them more likely to provide company-level family-friendly policies. Been et al. (2017) provide evidence to show that when generous national-level family policies exist, company-level family-friendly arrangements are also considered as more of a general term of employment, and managers are more likely to provide the arrangements across the board equally to all workers. Following from this, we could expect not only better access, but also a smaller division between insiders and outsiders in the access to family-friendly WTAs in countries with generous family policies. On the other hand, Chung (2017a) provides evidence to show that generous national-level family policies, especially work-facilitating policies (see also, Korpi et al., 2013;Misra, Budig, & Boeckmann, 2011) such as childcare provision, "crowd-in" occupational welfare, but not equally for all workers. Rather, companies seem to target the more profitable workers, given their added incentive of keeping and recruiting these workers, resulting in a larger division between insiders and outsiders in countries with generous family policies.
Lastly, it is important to note that these contexts, family policies, and corporatism/union strength are correlated. It is thus rather difficult to disentangle which of the variables are the driving factors. In fact, examining the correlation matrix of the context variables (see the Supporting Information), we can see that all four context variables are highly correlated. In other words, countries with strong centralized unions, which are more likely to be corporatist countries, are also those where family policies are generous and childcare coverage extensive. Thus, I also take a more case-focused, rather than a variable-focused, approach when examining the variance in the gaps between insiders and outsiders, looking at the (groups of) countries and their patterns.

| Data
For the purposes of this article, I use the EWCS of 2015. This data is gathered by the European Foundation, and aims to provide information on a number of dimensions of working conditions for workers across Europe. Individuals across European Union (EU28) and five candidate countries were included. For comparability issues, I use the EU28 countries plus Norway and Switzerland. A random stratified sampling procedure was used to gather a representative sample of those aged 15 or over and in employment (a minimum of one hour a week) at the time of the survey, and was conducted through face-to-face interviews. Approximately 1,000 cases are included per country, with varying response rates. Of the total sample, I restrict the analysis to those in dependent employment, and further exclude those in the armed forces, and in agriculture/fishery due to the specific nature of these jobs. Flexible WTAs are used by men and women, as well as those with and without care responsibilities for different purposes (Clawson & Gerstel, 2014), and can also lead to different outcomes (Lott, 2015). Further, the access gap for flexible WTAs may be different for men and women (Chung, 2018). Thus I only focus on female workers with care responsibilities. Here, care responsibilities were measured through the following variable: "In general, how often are you involved in any of the following activities outside work?": "Caring for and/or educating your children, grandchildren", and "Caring for elderly/disabled relatives". Those individuals who are caring for children at least several times a week, and those who are caring for elderly/disabled relative at least several times a month are considered those with care responsibilities. The analysis further excludes workers over the retirement age of 65 and excludes all cases with a missing value in any one of the variables in the model results in 7,845 cases across 30 countries. 2

| Dependent variable
The article examines workers' access to family-friendly WTAs. The provision of flexitime was measured through the following question, "How are your working time arrangements set?", where the workers could answer: (1) "They are set by the company/organization with no possibility for changes"; (2) "You can choose between several fixed working schedules determined by the company/organization"; (3) "You can adapt your working hours within certain limits (e.g. flexitime)"; or (4) "Your working hours are entirely determined by yourself". Those who answered (3) or (4) to this question were considered to have flexitime. Second, time off work for personal reasons was measured through the question, "Would you say that for you arranging to take an hour or two off during working hours to take care of personal or family matters is...", where respondents could answer: "very easy", "fairly easy", "fairly difficult", and "very difficult". Those who answered the first two were considered to have the ability to take time off work.

| Independent variables
The key independent variables in the analysis are workers' contract status, job insecurity perception, and skill-level "Contract of limited duration/fixed-term"; (3) "A temporary employment agency contract"; (4) "An apprenticeship or other training scheme"; or (5) "No contract". Those who answered (1) to the question are considered as those with a permanent contract, and all others as those who do not. Lastly, to measure workers' perceived job insecurity, I used the following question, "To what extent do you agree or disagree with the following statement about your job? ... I might lose my job in the next 6 months", where respondents could choose from "strongly agree", "tend to agree", "neither agree nor disagree", "tend to disagree",or"strongly disagree". Those who chose the first two were considered to perceive subjective job insecurity. 3 Based on previous studies (e.g., Chung, 2017a;Wiß, 2017), I included the following variables as controls: age; whether the respondent lives with a partner; parental status, i.e., whether the respondent lives with a preschool child under six years of age; whether the respondent lives with a primary school aged child (6-12); respondent's working hours; whether the worker holds a supervisory role; existence of an employee representative at the company; management support; gender of the direct manager; gender dominance of the job post; and, lastly, the size and sector (public vs. private, as well as the line of business, reference group: Commerce and hospitality) of the company the respondent works in.
At the national level, to measure union bargaining power and structure, union density and collective bargaining coverage rate are used, both represented as a percentage of wage earners. These variables indicate bargaining power and to a certain degree corporatism. These variables are from the ICTWSS data set 5.1 and are for 2013 or the closest year available due to lack of data. Family policies are multi-dimensional, with very different labor market outcomes for women (Korpi et al., 2013;O'Connor, Orloff, & Shaver, 1999). Thus this article focuses on two different aspects of family policies. First, general generosity of family policies is measured through public expenditure on family policies as a percentage of gross domestic product (GDP). Second, it focuses on "work-facilitating" measures, which have been shown to be most important in explaining mothers' employment and gender gap (Misra et al., 2011) and in determining workers' access to flexible WTAs (see also Chung, 2017a). This is measured through the proportion of children using formal childcare for the age group zero to three years.
All data is for 2015 or the closest year available. All context variables have been centered and standardized for the models. For more details of the operationalization and descriptive analysis of all variables, please see the Supporting Information.

| Method
Random-slope multilevel regression models are used for the purposes of this article. Multilevel modelling assumes that the lower level sample-here, individuals-is subject to the influences of groupings (Rasbash, Steele, Browne, & Goldstein, 2009)-here, countries. In this article, I examine the empty model, before moving on to the multivariate analysis to see the influence of individual-(and company-) level characteristics that can explain workers' access to family-friendly WTAs to test H1. Next, random slopes models are used to test the varying impact of skill, insecurity and contract status across different countries (H2). A significant variance in the random slope entails that there are countries where the access gap for insiders and outsiders is significantly different from the average gap found across Europe. Lastly, I include cross-level interaction terms with the national-level variables and the random slope variables which allow us to see which country level factor can explain the varying access gap across countries. STATA 14.2 meqrlogit is used for the analysis.

| Descriptive
First, we examine the percentage of working female carers with access to family-friendly WTAs across different countries. As Figure 1 shows, more than a quarter of all female carers have access to flexitime. The Northern European countries-namely, Denmark, Sweden, the Netherlands, and Norway-are champions in their provision of flexitime, with half or more of all female workers with care responsibility having access to flexitime. Finland, Belgium, the United Kingdom, and other continental European countries are not far behind, with one-third of all female carers having access to flexitime. On the other hand, many southern-eastern European countries-such as Bulgaria, Greece, Lithuania, Portugal, Croatia, Cyprus, Romania, and Slovakia-stand out as those where on average less than 10% of all working female carers have access to flexitime. Examining access to time off work for personal reasons, almost two-thirds of all workers across Europe have access (see Figure 2). Again, the Netherlands, Norway, Sweden, Denmark and, to a certain extent, Finland rank high, but now countries such as Latvia, Bulgaria, Ireland, and the United Kingdom also come out as countries where female carers feel it is fairly easy to take a couple of hours off of work to tend to personal issues. Again Greece, Slovakia, and Cyprus, but now with the Czech Republic, are where female carers do not have access to time off during working hours.

| Outsiders' access to family-friendly working-time arrangements
In our next step, we examine whether outsiders' relative access to family-friendly WTAs depends on workers' labor market positions. Table 1 shows the multivariate multilevel regression results. High-skilled female carers are twice as likely as those with low and specific skills to have access to flexitime, and slightly more likely to have access to time off work, having controlled for a wide range of factors. On the other hand, there is no difference between permanent and non-permanent workers in their access to flexitime or time off work, although for the latter, permanent workers were slightly more likely to have it (p<0.010). On the other hand, those who feel their job is insecure were significantly less likely to have access to both types of family-friendly WTAs, especially for time off work, although the gap was smaller than that for workers in different skill-level jobs in the case of flexitime.
In addition, older workers, workers with a preschool child, in supervisory roles, and with management support, working in financial services, public administration, and other service sectors were more likely to have access to flexitime. On the other hand, those working in workplaces which are mostly female, working in small to medium sized public companies, and working in education, health and social services sectors-typical female-dominated sectors-were less likely to have access to flexitime. Similarly, those in supervisory roles, with management support, working in,  financial services, public administration, and other service sectors, now alongside those working in industry sectors were more likely to have access to time off work. In addition, those with a partner, working in micro-companies, and those in male-dominated job posts are more likely to have access to time off work. On the other hand, again those working in female-dominated jobs and, interestingly, with female direct bosses, those working longer hours, and somewhat those with primary school age children, were less likely to have access to time off work. Another thing to note in Table 1 is that having taken the composition effects into account, the variance across countries in the access to family-friendly WTAs is larger.

| Variance across countries in outsiders' access to family-friendly working-time arrangements
Examining the random slopes models, there was no cross-country variance in the effect of contract type as well as job insecurity on the access to both flexitime and time off work (p>0.100 in all cases). In other words, the access gap in family-friendly WTAs for those who feel their job is insecure vs. those who do not is relatively stable across all European countries, as is the lack of division between permanent and non-permanent workers. The effect of being a high-skilled worker on access to time off work is also stable across countries (p>0.100). However, the gap between workers with different skill-levels in their access to flexitime varied significantly (p=0.009, with significant reduction in log-likelihood). Taking a closer look at the cross-national variance (Figure 3), in the United Kingdom, the Nordic (Finland, Norway, Sweden, and Denmark), and some Continental European countries (France, Luxemburg, and Belgium), the gap between high-and low-skilled workers is large and significantly different from the European average. On the other hand, in Austria and Malta the gap is significantly smaller, and there is almost no gap between high-and low-skilled female workers with care responsibilities. Given that the Nordic and Continental European countries have been noted as the countries with the most prevalent dualization patterns (Palier & Thelen, 2010;Rueda, 2005), the results seem to reflect previous literature which focuses more on labor market insecurity patterns. In  Table 1 Source. Author's calculations. This is confirmed in the next step of the analysis, where I examine the cross-level interaction terms with skilllevels and context variables. Looking at the main country context variable effect in Table 2, we can see that countries with strong centralized unions-i.e., high collective-bargaining coverage and high union density, and those with generous family policies (large family policy spending, and extensive childcare coverage for young children)-are those where, in general, workers have more access to flexitime. The cross-level interaction terms show that these countries are also those where the gap between high-skilled and lower-skilled workers is the largest. Of the context variables, family policy expenditure seems to be the most influential, explaining up to 52% of the cross-national variance in the flexitime access gap between high-and low-skilled workers. It remains significant even when other context variables are included in the model (Model 2-5 and other models available upon request). As shown in Figure 4, countries with generous family policies at the national level are where both low-and high-skilled female workers with care responsibilities have more access to flexitime. However, this is especially the case for high-skilled female carers making these countries those where the largest gaps between workers in different skill-levels are found. Similar patterns are found for other context variables (available upon request). Further, we see a segmentation pattern between workers in the access to these arrangements. Unlike statutory policies, occupational policies seem to be only provided to workers where employers have a vested interest and in better bargaining positions, resulting in a dualized system. Confirming the organizational segmentation and other previous studies (Lambert & Haley-Lock, 2004;Seeleib-Kaiser & Fleckenstein, 2009), I found a division based on the skilllevel of the worker. However, unlike previous studies, I found that this gap varied across countries. Similar to previous studies on dualization of labor market insecurity (e.g., Chung, 2016;Palier & Thelen, 2010;Rueda, 2005), it is the Nordic and continental European countries, with stronger more centralized unions, extensive childcare coverage for younger children, and high levels of family policy spending, where the division between high-vs. low-skilled workers is the largest. However, it should be noted that the level of dualization is high in these countries only because of the very high provision made to insiders. The institutional contexts examined in this article, mainly strong centralized unions and generous national level family policies, helped both low-and high-skilled workers to gain access to family-friendly WTAs, yet this was especially true for the latter. Another major contribution of this article is to show that the most commonly used definitions of outsiders may not fully capture the actual divisions taking place in the labour market (see also Yoon & Chung, 2016). This may especially be the case when looking at a wider range of labor market segmentation issues and working conditions. This article provides evidence to show that it is the subjective insecurity perceptions rather than the objective contract status that matters in determining access to family-friendly WTAs especially for time off work. This may be because since subjective insecurity perceptions are influenced by a wider range of factors, the subjective state may be a better indication of the (bargaining) position of the individual within the workplace, the (perceived economic) situation of the company, as well as the general labor market condition particular to the worker. The objective contractual status, on the other hand, may entail different things depending on the context (Chung, 2016). Future scholars examining labor market segmentation should thus consider a wider range of definition of segmentation, such as subjective job insecurity, when examining patterns of divisions in the labor market. Several recent studies (TUC, 2017;Working Families, 2017) indicate that the fear of negative career consequences and the fear of job loss prohibit workers from taking up family-friendly arrangements. This study provides further evidence that those who fear that they will lose their job are less likely to feel that they have access to family-friendly WTAs. Policymakers should thus find solutions to address such fears, and ensure that especially the most vulnerable workers can have a genuine right to flexible working regardless of their bargaining positions. The recent European Commission proposal for the right to request flexible working for parents 4 comes with a mechanism to protect workers from job losses and discrimination when requesting the right. Such policy structures are crucial in ensuring that all workers get genuine access to flexible working.