A process‐oriented, multilevel, multidimensional conceptual framework of work–life balance support: A multidisciplinary systematic literature review and future research agenda

Correspondence YuyangFan,Business School,University ofEdinburgh, 29BuccleuchPlace, EdinburghEH89JS,UK. Email: Yuyang.Fan@ed.ac.uk Abstract Work–life balance is shaped not only by how individuals manage their personal demands and resources, but also by stressors and work–life balance support mechanisms from external environment encompassing multilevel social systems. Our systematic literature review focuses particularly on the role of work– life balance support, drawing on 384 journal articles and book chapters published between 1960 and 2019 across five research disciplines, including management, applied psychology, industrial relations, family studies and sociology. We make four major contributions to the literature, including: (1) conceptualizing work– life balance support from a process-oriented perspective pertinent to a virtuous cycle of resource investment and return, drawing on the conservation of resources theory and the personal resource allocation framework; (2) adopting a multilevel approach that construes the interactions in terms of resource changes between individuals’ work–life experiences and their surrounding social systems nested at multiple levels, applying the socio-ecological systems theory; (3) proposing a multidimensional typology that differentiates the role of actual existence versus subjective perception of support mechanisms as inspired by social support literature; and (4) advocating a pluralist, multi-stakeholder approach to comprehending and reconciling multiple stakeholders’ shared and competing interests around provision/utilization of supportmechanisms based on insights frommultidisciplinary literature. Our process-oriented, multilevel and multidimensional framework conceptualizes the critical role of work–life balance support in iterative interactions between individuals and their multilevel social environment through resource changes and reality–perception transformation. This conceptual framework also underscores the importance of pluralist thinking, context specification and cost-effectiveness analysis for future research.


INTRODUCTION
Work-life balance is defined as an individual's perceived optimum allocation of personal resources that helps in coping with stressors and guarantees effective functioning of both work and non-work roles (Grawitch et al., 2010). Work-life balance research in the 1960s initially focused on the division of labour (between male breadwinners and female carers) and associated tensions around separating work and personal lives (Gatrell et al., 2013;Naithani, 2010). Increased female workforce participation from the 1980s (Snooks, 1996) exacerbated these tensions between work commitments and family responsibilities, resulting in both governmental and organizational family-friendly interventions (Lockwood, 2003). More recently, two additional work-life balance challenges have emerged. First, technological advancements are blurring work and nonwork roles (Kumar & Janakiram, 2017) because information technology enables free communication irrespective of time and space (Adisa et al., 2017). Second, ageing societies have resulted in heavier workloads and care burdens for the working population (Khallash & Kruse, 2012).
Many work-life challenges (e.g. motherhood penalties and delayed retirement) are embedded in societal problems (e.g. gender inequality and ageing societies) and the broader social environment and, therefore, cannot be solved through individual efforts alone; requiring extensive environmental support (Spinks, 2004) from family members (Halbesleben et al., 2012), organizations (Jang, 2009) and governments (Fagnani & Letablier, 2004). We combine the concepts of work-life balance (Grawitch et al., 2010) and social support (French et al., 2018;Muñoz-Laboy et al., 2014) to define work-life balance support as perceived or actual and tangible or intangible external resources invested by other people or social systems that expand individuals' personal resources or optimize personal resource allocation in fulfilling work and non-work roles.
We argue that work-life balance support mechanisms benefit not just individuals and their families but also organizations and society overall (Brough et al., 2008). Individual and family returns may include higher marital and family satisfaction (Ferguson et al., 2012), organizations may observe lower absenteeism (de Menezes & Kelliher, 2011) and societal returns may include increased labourforce participation (Brough et al., 2008). However, the work-life balance support literature has been criticized for focusing primarily on individuals and downplaying implications for families, organizations and societies (Gatrell et al., 2013;Ozbilgin et al., 2011).
Also, limited research addresses the interplay between actual existence (i.e. structural dimensions) and subjective perception (i.e. functional dimensions) of work-life balance support, despite recognition that both dimensions make a difference to individuals' work-life experiences (Lim & Lee, 2011;Yuile et al., 2012). Consequently, extant literature fails to unpack the process through which individuals make sense of, and effectively utilize, work-life balance support mechanisms to balance competing work and life demands, achieve positive work and life outcomes, and contribute to their families, organizations and wider society.
Drawing on multidisciplinary research, we address three gaps by conceptualizing work-life balance support in a (1) process-oriented, (2) multilevel and (3) multidimensional framework, and advance a more holistic research agenda by exploring the iterative interactions between individuals and social environments at different levels and distinguishing between perception versus reality.
Our systematic review makes four major theoretical contributions. First, drawing on the conservation of resources (COR) theory (Hobfoll, 1989) and the personal resource allocation (PRA) framework (Grawitch et al., 2010), we provide new insights into work-life interface from a processoriented perspective (see Figures 1 and 4) of resource changes given interactions between individuals and social environments. We conceptualize work-life balance support as resource investments by people and social environments that promote individuals' expansion or effective allocation of personal resources to fulfil work and life demands. Consequently, individuals' enhanced work-life balance brings other positive returns and enriches future resources for both individuals and their environments. Iterative worklife balance support investments and returns form a virtuous cycle of resource accumulation; essential for the effective functioning of families, organizations and societies.
Second, we undertake a multilevel and systematic approach (see Figure 1 and Table 1) by drawing on the socioecological systems theory (Bronfenbrenner, 1979;Pocock et al., 2012). We challenge the traditional understanding of work-life balance as an individual-level phenomenon and instead, emphasize how the interactions between individuals' work-life experiences and their environment shape work-life balance. Theoretically, environmental stressors necessitate individual acquisition of work-life balance support at different levels (i.e. micro, workplace and societal levels) to achieve positive outcomes for both individuals and multilevel social systems. Current work-life challenges cannot be solved through individual effort alone, given their embeddedness in broader social environments. Therefore, we claim that minimizing work-life conflict and enhancing individuals' quality of life necessitates systematic changes that combine individual effort with external support from families, organizations and governments.
Third, by considering multidisciplinary literature (i.e. management, applied psychology, industrial relations, family studies and sociology), we go beyond work-life F I G U R E 1 A process-oriented, multilevel conceptual framework of work-life balance support balance as an individual-level phenomenon and posit it as a societal concern involving multiple stakeholders (e.g. individuals, employers and governments) situated in multilevel social environments. Practically, employers' and policymakers' design and implementation of work-life balance support mechanisms necessitates a systematic and pluralist consideration of different stakeholder interests. Therefore, we recommend scholars bridge this researchpractice gap by aligning their investigation of work-life balance support with key stakeholders' goals.
Finally, we adapt Cohen and Wills' (1985) classification of social support into a novel multidimensional typology (see Table 2) that distinguishes between the reality versus perception of work-life balance support. We propose that structural dimensions depict the actual existence and quantity of work-life balance support, while functional dimensions offer a more perceptual and qualitative evaluation. We explore this reality-perception transformation by considering spiral interactions between physical environments and people's subjective perceptions (i.e. interplay between structural and functional dimensions) of work-life balance support via three interconnected processes: decision-making, realization and subjective feedback (see Figure 4). Next, we propose our conceptual framework, highlighting the need for a process-oriented, multilevel and multidimensional conceptualization and investigation of work-life balance support. Following an overview of our methodological approach, our findings illustrate how extant literature has conceptualized and investigated work-life balance support as well as its antecedents and outcomes. We conclude with a future research agenda.

THEORETICAL FOUNDATION AND CONCEPTUAL FRAMEWORK
Building on resource theories, the socio-ecological systems theory and social support literature as our theoretical foundation, we advance a process-oriented, multilevel, multidimensional conceptual framework of work-life balance support.

Resource theories: Process of resource investment and return in support of work-life balance
We draw on the COR theory (Hobfoll, 1989;Hobfoll et al., 2018) and the PRA framework (Grawitch et al., 2010) since these theories enable our conceptualization of interactions between individuals' work-life experiences and their social surroundings via the circular process of resource investment and return.
The COR theory rationalizes human instinct to obtain and preserve resources (i.e. valuable things needed for survival; Hobfoll et al., 2018). Resource loss creates stress, necessitating individuals' and social systems' investment in existing resources to accumulate future resources as well as protect against/recover from resource loss (Hobfoll et al., 2018). Individuals need to invest personal resources such as time and money to fulfil work and non-work demands and achieve good work-life outcomes. Other higher-level social systems, such as families, organizations, communities and the state, can also invest resources (i.e. work-life balance support) that enhance individuals' work-life balance, and generate positive family, organizational and societal resource returns.
Similar to the Matthew effect (Merton, 1968) of (dis)advantage accumulation, whereby wealthy people get wealthier while the impoverished become poorer (Bask & Bask, 2015), the COR theory (Hobfoll et al., 2018) suggests that individuals with more personal resources and external work-life balance support are more capable of achieving positive work and life outcomes and gaining future resources, forming a virtuous cycle of resource accumulation. Furthermore, the concept of caravan passageways highlights environmental conditions that facilitate or hinder resource accumulation. Supportive caravan passageways, such as family-friendly corporate cultures or more inclusive societies that promote gender equality, facilitate individual access to obtain personal resources needed for work-life balance. This resource caravan passageways principle reinforces the importance of studying a work-life balance supportive social system that creates synergies between individuals' personal resources and related stakeholders' work-life balance support and generates positive returns for all stakeholders. (1) Provision ♦Definition: The existence/frequency/number/amount/quantity of a social system or an individual offering support/aid/help to facilitate work-life balance. ♦Example: -'Today, I willingly gave my time to help colleagues who had work-related problems.' (1 = never to 5 = always) (Lin et al., 2017) -'Whether an organization offered the following five practices: workplace childcare, childcare allowances, career break schemes, maternity leave, and/or paternity leave.' 'Whether the organization provided these practices above and beyond the existing statutory requirements.' (1 = yes; 0 = no) (respondents were human resource managers) (Giardini & Kabst, 2008) (2) Utilization ♦Definition: The existence/frequency/number/amount/quantity of using/utilizing/receiving/adopting a support mechanism to achieve work-life balance. ♦Example: -'Respondents were asked to choose the work-life balance practices [from a list of practices] that they were currently using.' (1 = yes; 0 = no) (Thakur & Bhatnagar, 2017) (3) Intervention ♦Definition: The intervention of work-life balance support mechanisms in a laboratory, field or natural experiment; the process of implementing work-life balance support mechanisms. ♦Example: -Vignette experiment: 'HR benefits provision was manipulated in the recruitment advertisement presented to the respondents. The company described in the Control Condition provided standard pay benefits (which is common to all versions of the survey); the company in the Condition 1 survey offered health care and insurance benefits in addition to standard pay benefits; and the company in Condition 2 offered WLBs in addition to standard pay benefits.' (Firfiray & Mayo, 2017) -Field experiment: 'exposure to the STAR intervention [STAR (Support. Transform. Achieve. Results) a group-randomized field trial of an organizational intervention designed to promote control over work time and supervisor support for employees' personal and family life].' (1 = experimental group; 0 = control group) (Moen et al., 2016) -Natural experiment: 'Family-friendly law (Act 39/99) approved in Spain in 1999.' (de la Rica & Gorjon, 2016) (4) Timing ♦Definition: The time point or (life/career/historical) stage at which support is offered/received/used; the length of time it takes for the individual to use/offer a support mechanism or the length of time an organization or government takes to implement a support mechanism; the length of time for a support mechanism to take effect or cease to be effective. ♦Example: -'We collected a more detailed set of information about the benefits we labelled as family-supportive, including when they had been implemented.' 'The way that we do this is to make a distinction between "new" and "old" benefits. "Old" benefits are defined as those that have been offered for five or more years as of the survey date.' (Baughman et al., 2003) Functionality (i.e. the perceptual and qualitive evaluation of work-life balance support) Parents were asked to indicate their level of satisfaction with aspects of childcare arrangements including cost, location, and quality.' (1 = very dissatisfied to 5 = very satisfied) (Roberts et al., 2004) (9) Perceived consequence ♦Definition: The perceived negative outcomes of using/providing the support mechanism. ♦Example: -'Using family-friendly programs would harm my status at work.' 'Using family-friendly programs would hurt my career progress.' (1 = strongly disagree to 5 = strongly agree) (Butler et al., 2004) -'Supervisors/co-workers make negative comments if someone benefits from tools aimed at supporting work-family balance.' (1 = disagree to 4 = agree) (Ghislieri et al., 2017) Applying the COR theory (Hobfoll, 1989) specifically to work-life balance research, the PRA framework (Grawitch et al., 2010) highlights the virtuous cycle whereby individuals appraise and allocate personal resources for handling work-life demands, achieve positive personal outcomes and acquire future resources for achieving work-life balance. Personal resources for achieving a good work-life balance comprise time, finances, physical and mental attributes (e.g. energy), knowledge and skills and social status/networks (Fletcher & Fincham, 1991;Grawitch et al., 2010). Demands refer to tasks or responsibilities that require consumption of an individual's personal resources (Grawitch et al., 2010). Work-life balance support is distinct from personal resources because it encompasses external resources invested by other people and surrounding social systems (e.g. families, organizations, communities and the state) aimed at promoting individuals' effective allocation of personal resources to fulfil work and non-work demands.
In summary, drawing on a process-oriented perspective from the COR theory (Hobfoll et al., 2018) and the PRA framework (Grawitch et al., 2010), we frame work-life balance support as a virtuous cycle of resource investment and return. Individuals invest personal resources to fulfil demands, achieve positive work-life outcomes and replenish personal resources. Meanwhile, higher-level social systems invest in work-life balance support, generating positive family, organizational and societal outcomes. In other words, higher-level social systems themselves may also benefit from individuals' work-life balance. We agree with Kossek's (2015, p. 372) claim that work-life balance as an 'individual-level phenomenon should be bracketed and understood across multiple levels'. The socio-ecological systems theory enables us to envision this process of resource investment and return in terms of work-life balance support mechanisms in the context of social systems at different levels.

Socio-ecological systems theory: A multilevel support system for work-life balance
We draw on a multilevel and systematic approach to addressing individuals' dynamic interactions with broader social environments (Golden & Earp, 2012) from the socio-ecological systems theory (Bronfenbrenner, 1979) and its theoretical extensions to work-life balance research (Pocock et al., 2012;Voydanoff, 2008) because it fits our research purpose of understanding work-life balance as a societal concern rather than an individual-level phenomenon (Brough et al., 2008). The socio-ecological systems theory compares human society to natural ecology and underlines the dynamic interplay between individuals and external social systems across multiple levels (Richard et al., 2011).
A social system refers to a collective entity of interdependent and interactive individuals and surroundings that share common goals, values and beliefs (Rogers, 2003). Typical examples of social systems include families, organizations, communities and countries. Crucially, different social systems are related to each other. For example, a family intersects with an organization when a family member is also an employee; this family is also a subset of a community.
In this paper, we draw on the socio-ecological systems theory to conceptualize work-life balance support as a multilevel construct. However, instead of using their classification of social systems, we adopt conceptual levels (i.e. micro, workplace and societal levels) more commonly used in work-life balance support literature. This aligns with Abendroth and den Dulk's (2011) contention that families, organizations and countries are the most frequently studied units of social systems that provide work-life balance support mechanisms to individuals. Therefore, we categorize social systems into three broad conceptual levels: (1) the micro level, comprising family members and personal social networks, such as friends and neighbours; (2) the workplace level, including the working environment and professional contacts, such as supervisors and team members; and (3) the societal level, defined by geographical and/or psychological boundaries, such as communities and countries. Additionally, we adapt the chronosystem proposed by the socio-ecological systems theory (Bronfenbrenner, 1986) into a timing dimension that depicts the temporal or historical aspect of work-life balance support in our multidimensional typology.

Proposed model: A process-oriented and multilevel conceptual framework
Drawing on resource theories and the socio-ecological systems theory, we propose a process-oriented, multilevel conceptual framework of work-life balance support, whereby an individual's work-life experiences interact with the external environment via the process of (1) drawing worklife balance support from external social systems to adjust, expand and/or reallocate personal resources to (2) cope with stressors from social systems nested at multiple levels and, subsequently, (3) bring about positive outcomes to multilevel social systems. These positive outcomes may further (4) replenish/expand individuals' resources to (5) cope with stressors associated with work-life balance, forming a virtuous cycle.
We define stressors as objects, events or environmental conditions that pose actual or potential survival challenges or opportunities for an individual (Deckers, 2016). External stressors are internalized as individuals' personal demands that necessitate their consumption of personal resources. Outcomes include work-related (e.g. job satisfaction and performance), life-related (e.g. family and leisure satisfaction) and cross-domain (e.g. physical/mental health and work-life conflict/enrichment/balance) achievements. For example, utilization of teleworking arrangements (work-life balance support) provided by the organization (workplace-level social system) allows parents (individual) to spend more time (personal resource) taking care (personal demand) of their children (stressor) in their family (micro-level social system), subsequently enhancing the parent-child relationship (outcome and personal resource).
We propose a process-oriented, multilevel conceptual framework (see Figure 1), which highlights several paths for further investigation: 1. An individual recognizes personal demands in response to stressors from external social systems. 2. An individual draws on work-life balance support mechanisms from external social systems to adjust, expand and/or reallocate personal resources. 3. Personal demands deplete an individual's personal resources. 4. Stressors and/or personal demands pose challenges for survival and hence negatively influence an individual's work-related, life-related and cross-domain personal outcomes.

An individual consumes personal resources to fulfil
personal demands and respond to external stressors. Hence, personal resources may moderate the influence of stressors and/or personal demands on an individual's personal outcomes. 6. In line with the resource accumulation assumption of the COR theory (Hobfoll et al., 2018), we propose that an individual's increased personal resources can generate positive personal outcomes. 7. The aggregation of personal resources, personal demands and personal outcomes of individuals nested within a social system (e.g. family/organization/ country) may influence this social system and generate higher-level (e.g. family/organizational/societal) outcomes. 8. Similar to path (6), work-life balance support investments by an external social system may bring positive outcomes to that system. 9. Similar to path (5), work-life balance support may moderate the relationship between stressors and/or personal demands and an individual's personal outcomes.
Work-life balance support may also moderate the influence of stressors from a higher-level social system on higher-level outcomes. 10. Positive higher-level family, organizational and social outcomes may become new sources of work-life balance support, whereas negative higher-level outcomes may become new stressors, forming a work-life balance resource cycle.
11. Some antecedents from multilevel social systems may influence an individual's perception, utilization and provision of work-life balance support, while other antecedents may influence the provision of work-life balance support by a certain social system.
Since this review focuses on work-life balance support, we systematically synthesized and organized the findings according to paths (2), (8), (9), (10) and (11) highlighted above. Specifically, we documented the measures and conceptual levels of work-life balance support mechanisms, their antecedents and outcomes and the relationships between these measures. Other paths in this conceptual framework have been well theorized in extant work and are not this review's focus. For instance, relationships between stressors, personal demands, personal resources and personal outcomes, that is, paths (1), (3), (4), (5), (6) and (7), have been widely theorized and empirically examined utilizing different occupational stress and coping theories (e.g. Bakker & Demerouti, 2007;ten Brummelhuis & Bakker, 2012).
Furthermore, extant literature suggests that key stakeholders, for example, the individual, the employer and the government, may sometimes pursue complementary yet conflicting goals (Leitner & Wroblewski, 2006) whereby: (1) individuals utilize support mechanisms to meet personal work and life demands; (2) employers provide support mechanisms to maximize profitability and fulfil social responsibility (Been et al., 2017;Faria & Machado, 2018); and (3) governments offer support mechanisms to achieve efficiency (e.g. increasing labour supply) and promote fairness (e.g. gender equality) (Brough et al., 2008). The dynamic balance between these varied goals is critical for both short-term prosperity and long-term sustainability of all stakeholders. In order to examine the impact of worklife balance support from a multi-stakeholder perspective, we classified outcomes in relation to the main goals of key stakeholders at different levels: (1) work-related versus life-related outcomes at the private level; (2) profitabilityoriented versus social responsibility-oriented outcomes at the workplace level; and (3) efficiency-based versus fairnessbased outcomes at the societal level.
Our review draws on multidisciplinary research to develop a more comprehensive and nuanced understanding of stakeholders' competing pursuits. Specifically, we synthesize literature from five research disciplines: (1) management that offers managerial and employee perspectives; (2) industrial relations that addresses interrelationships between employees, employers, trade unions and governments; (3) applied psychology that explores interactions between individuals and their work/non-work environment through the psychology lens; (4) family studies that focuses on family characteristics, interactions and policies; and (5) sociology that focuses on broader societal concerns.
In summary, from a process-oriented perspective, we theorize ongoing interactions between individuals' work-life experiences and surrounding social systems in terms of a virtuous cycle of resource changes. Furthermore, we adopt a multilevel approach that unpacks the impact of work-life balance support via the constructive process of resource investment and return across multiple levels of social systems. Our multilevel conceptualization is based on the level of social system that provides work-life balance support.

Social support literature: A multidimensional typology of work-life balance
Our conceptual framework offers a novel way of unpacking iterative interactions between individuals and their environment, whereby individuals make sense of the environmental reality (e.g. actual presence of stressors and work-life balance support) through cognitive interpretation and these subjective perceptions enable individuals to react to, and transform, the environment. We are particularly interested in how the dichotomy between actual existence versus subjective perception of work-life balance support influences individuals' work-life balance and generates relevant outcomes. We draw on the social support literature to propose a multidimensional typology that distinguishes between actual existence versus subjective perceptions of work-life balance support.
Social support is defined as actual or perceived, tangible (e.g. money) and intangible (e.g. love) resources for stress management (French et al., 2018) within a person's social network, comprising personal (e.g. family members) and professional (e.g. co-workers) ties (Muñoz-Laboy et al., 2014). Our definition of work-life balance support builds on the social support literature, given the overlap between social support and work-life balance support addressed in more detail in our findings. Cohen and Wills (1985) classified dimensions of social support according to its structure (i.e. 'the existence of relationships') and functionality (i.e. 'the extent to which one's interpersonal relationships provide particular resources'). We adapt this classification into a multidimensional typology and propose that structural dimensions depict the actual existence and quantity of work-life balance support, while functional dimensions signify the perceptual and qualitive evaluation of work-life balance support.
Furthermore, we split the broad dimensions of support structure and functionality into more nuanced sub-dimensions. First, we utilize frequently studied facets of social support, that is, structural sub-dimensions of provision, utilization and timing and functional subdimensions of perceived availability, perceived necessity and perceived usefulness from comprehensive reviews on social support measurement (Bruhn & Philips, 1984;Shinn et al., 1984;Tardy, 1985). Second, we adapt negative career consequences and organizational time expectations from the composite concept of work-family culture (see Thompson et al., 1999) into two functional sub-dimensions, that is, perceived consequence and expectation. Third, we propose a new structural sub-dimension-intervention, which denotes the treatment of work-life balance support in experimental settings. Operationalized definitions and examples of specific dimensions in our proposed structure-functionality typology of work-life balance support are elaborated in Table 2.

METHODS
Our systematic literature review involved an exhaustive literature search, comprehensive synthesis and critical appraisal of extant studies according to pre-defined research questions 'by adopting a replicable, scientific and transparent process' (Tranfield et al., 2003, p. 209). We did not conduct a quantitative meta-analysis that extracts and analyses data statistically from reviewed studies for two reasons. First, a third of the articles used qualitative and mixed methods, rendering a quantitative meta-analysis impossible. Second, extant quantitative studies encompass a wide variety of support mechanism, antecedents and outcomes rather than concentrating on a few frequently used indicators. Therefore, our systematic review offers a qualitative synthesis of extant findings.

Literature search and selection
The literature search was conducted between 12 June 2018 and 3 December 2019. At the outset, all authors agreed on key search terms and the inclusion criteria, using the SPIDER protocol, given its applicability for a narrative synthesis of quantitative, qualitative and mixed methods research . The following inclusion criteria were applied: (1) sample (S)-people involved in both paid work and personal/family lives; (2) phenomenon of interest (PI)-work-life balance support; (3) design (D)-theoretical work, literature reviews and empirical studies; (4) evaluation (E)-concepts, antecedents, outcomes, contexts and other aspects of work-life balance support; and (5) research type (R)-quantitative, qualitative and mixed methods.
Additional inclusion criteria included: (1) restricting the publication outlet to peer-reviewed journal articles and scholarly book chapters, given that rigorous and valuable scholarly work is often published in these outlets (Littlejohn & Foss, 2009); (2) confining to English as the commonly used language in academia; (3) limiting the range of publication dates from 1 January 1960 (when worklife balance emerged as an academic discourse) (Naithani, 2010) to 3 December 2019 (the final day of our literature search) to capture relevant sources; and (4) focusing on five research disciplines encompassing different stakeholder perspectives at multiple levels (e.g. individuals, employers and governments) in line with our proposed conceptual framework (see Figure 1), including management, industrial relations, applied psychology, family studies and sociology.
Combining the inclusion criteria of sample 'AND' phenomenon of interest, we ran the search on article titles in the database of the Web of Science Core Collection using different combinations of search strings (see Figure 2). We focused on searching article titles since they usually contain the most crucial information (e.g. key concepts and studied measures), which enabled us to identify studies focused on work-life balance support.
The PRISMA (Moher et al., 2009) flow chart (see Figure 2) outlines the selection process in line with our preset inclusion criteria. During the identification stage, 688 results were generated. A total of 20 editorial materials were excluded for not complying with the design criteria, while the remaining 668 results were across five research disciplines of management (214), applied psychology (142), family studies (106), industrial relations (104) and sociology (102). During the screening stage, 187 duplicates were removed, while retaining 481 records. In the eligibility stage, full texts of these 481 articles were examined and a further 97 records were eliminated-59 records were irrelevant according to the SPIDER protocol, 31 texts could not be accessed, 5 records were in publication outlets not under consideration (e.g. news/magazine articles and editorial materials) and 2 records lacked critical methodological information. A final sample of 384 articles was coded and analysed.

Literature coding and analysis
The selected articles were coded in an Excel spreadsheet according to the following pre-determined coding scheme: bibliographic information, research methods, concepts and measures of work-life balance support, theories and key findings. Each article was double-coded by two of the three authors, with discrepancies resolved through discussion between all three authors. We documented how each article conceptualized specific work-life balance support mechanisms (e.g. supervisor support) as well as their antecedents, outcomes and contexts.
As mentioned earlier, work-life balance support is a multilevel, multidimensional construct. Hence, we documented the conceptual level (i.e. source of support or level of the social system that provides support) and the operational level (i.e. unit of measurement or level of data analysis) of the measures of work-life balance support mechanisms and their antecedents and outcomes according to our process-oriented, multilevel conceptual framework (see Figure 1). We recorded conceptual levels for all types of research, but documented operational levels only for quantitative research or quantitative elements of mixed methods research. Outcomes were classified from the perspectives of individuals' (i.e. work/liferelated and cross-domain), employers' (i.e. profitability/ social responsibility-oriented) and governments' (i.e. efficiency/fairness-based) major goals.
Moreover, we coded specific dimensions of work-life balance support mechanisms for both quantitative and qualitative research according to our proposed operational definitions (see Table 2). The dimensions were coded based on the definition of specific work-life balance support mechanisms or the sample items/measures. For both quantitative and qualitative research, we also explored whether a study had examined the timing dimension by employing a longitudinal design or adopting a life-course or historical perspective.
In this review, we attempt to systematically synthesize key findings around the (1) multilevel and (2) multidimensional conceptualization and measurement of work-life balance support and the (3) antecedents, outcomes and contexts of work-life balance support in line with our process-oriented, multilevel and multidimensional conceptual framework. Table 3 captures the overarching characteristics of the literature considered. Figure 3 highlights the growing popularity of research on work-life balance support by discipline.

CONCEPTUALIZATION, MEASUREMENT AND EVALUATION OF WORK-LIFE BALANCE SUPPORT
Given the lack of a comprehensive definition for worklife balance support, we advance a definition for this term (see Introduction) by drawing on the PRA framework (Grawitch et al., 2010) and social support literature (French et al., 2018;Muñoz-Laboy et al., 2014). In this section, we discuss how extant research has conceptualized and measured work-life balance support as a mul-tilevel, multidimensional and process-oriented construct with reference to our proposed conceptual framework (see Figure 1) and structure-functionality typology (see Table 2).

Sources of work-life balance support: A multilevel construct
We adopted the socio-ecological systems theory's (Pocock et al., 2012) systematic, multilevel approach to examine the conceptual (i.e. source) and operational (i.e. unit of measurement/analysis) levels of work-life balance support.
We identified a discrepancy between the multilevel conceptualization versus operationalization of work-life balance support (see Table 1), whereby most research conceptualizes work-life balance support at the workplace level, while measuring and analysing support mechanisms at the micro level. Individual-level measures were frequently used to assess a higher-level concept without data aggregation or multilevel analysis. For instance, some studies measured organizational-level concepts such as organizational work-family culture at the employee level, without aggregating the data at the organization level (e.g. Next, we present how work-life balance support mechanisms have been examined at three conceptual levels and their interplay across levels.

Micro level
Social support was the dominant work-life balance support mechanism at this level, with a specific focus on spouses/partners (e.g. Pluut et al., 2018), some consideration of extended family members (e.g. children, siblings, parents and relatives) (e.g. Fan, 2009) and rarely friends and neighbours (e.g. Winston et al., 2019). Both tangible (e.g. career/childcare/housework assistance) and intangible (e.g. emotional and informational support) forms of social support for both work-related and liferelated demands were frequently examined. Several metaanalyses highlight the positive role of both micro-and workplace-level social support in abating work-family conflict and alleviating work/family stressors (French et al., 2018;Kossek et al., 2011;Michel et al., 2010).

Workplace level
At this level, formal and informal types of work-life balance support mechanisms emerged. Both workplace social support (denoting social support from professional networks) and organizational culture (referring to familysupportive or work-family culture) were commonly used for measuring informal organizational support.
We identified four major categories of formal support mechanisms at the organizational and team level: (1) flexible working arrangements that enable adjustment of work demands, such as flexible schedules (Fang et al., 2019) and phased retirement (Hill et al., 2011); (2) organizational interventions that help employees achieve their work/career goals, such as career development training (Batt & Valcour, 2003) and virtual office facilities (Kalysh et al., 2016); (3) family-friendly policies that accommodate employees' family demands, such as parenting courses (McDonald et al., 2013), leave arrangements for dependent care/marriage/bereavement (Moon & Roh, 2010) and childcare/eldercare services/facilities/subsidies (Butler et al., 2004;Wang et al., 2011); and (4) employersponsored provisions for other personal demands (e.g. health, leisure and education), such as pensions and life/health/medical insurance, canteens, on-site gyms, tuition reimbursement, travel services and allowances, recreational and social activities Rajan-Rankin & Tomlinson, 2013).
These formal organizational support mechanisms facilitate employees' fulfilment of life demands by rearranging work demands (e.g. time, location and workload) (Kossek & Lautsch, 2018) through family-friendly or life-friendly services, facilities and financial benefits (Fritz & van Knippenberg, 2018). In extant literature, the most frequently studied formal support mechanisms were flexible working arrangements and family-friendly policies (especially for childcare). Butts et al.'s (2013) meta-analysis indicates that employees' perceived availability and utilization of these organizational policies can improve their work-life balance and generate positive employee outcomes, such as increased job satisfaction and affective commitment. Also, Kossek et al.'s (2011) meta-analysis shows that worklife balance-specific supervisor support and organizational support have a stronger positive impact on employees' work-life balance than general support does. However, support mechanisms for other important life demands (e.g. eldercare, health, leisure, education and networking) were under-explored.

Societal level
At this level, work-life balance support mechanisms can be categorized as (1) public policies, (2) public infrastructures and services, (3) private services, (4) support from communities and social groups/organizations, (5) national cultures or societal norms about work-life balance support and (6) international/transnational organizations and legislation. The most frequently studied measures were public policies addressing (a) dependent care assistance (e.g. maternity/paternity leave and child benefits), (b) work/employment flexibility (e.g. part-time employment and statutory flexible working requests) and (c) social security (e.g. health/unemployment insurance, housing benefits and pension). Research often presents Nordic examples as good practice, where it is easier to maintain good work-life balance than in other countries on account of their gender egalitarian norms and abundant state support for childcare/eldercare, employment and healthcare that encourage both genders to share work and domestic responsibilities (Leitner & Wroblewski, 2006).
Scholars repeatedly highlighted public childcare services (e.g. Misra et al., 2007;Riva, 2016) but ignored other public infrastructures and services, such as public eldercare services (Martin, 2017), public education systems (Barsoum, 2019), information technology infrastructures (Ladkin et al., 2016) and public transportation (Cook, 1989). Also, public health services, sports facilities and recreation infrastructures are indispensable elements of life but remain blind spots in existing literature. A handful of studies (e.g. Boye, 2011;Gronlund & Magnusson, 2016) discussed welfare state regime typologies that categorize countries according to their government ideology/disposition of market liberalism/efficiency versus state interventionism/fairness and/or the primary target of their public poli-cies and infrastructures for promoting gender equality versus reinforcing traditional gender roles.
Several studies considered private/marketized childcare (Crompton, 2002) and domestic services (Husu, 2005) but overall, the role of private services in promoting worklife balance was neglected. Support from communities and social groups/organizations was under-examined, barring some work on support from religious communities (Shai, 2002), trade unions (Berg et al., 2014), staff associations (Hyman & Summers, 2004) and works councils (Heywood & Jirjahn, 2009). National cultures or societal norms about work-life balance support were sparingly addressed in terms of gender egalitarianism (Cogin et al., 2018), familism (Riva, 2016) and work centrality . With respect to international/transnational organizations and legislation, only Oliver (2012) discussed the need for transnational policies aimed at facilitating scientists' work-life mobility across the European Union.

Cross-level interplay
Research on the interplay between different levels of worklife balance support mechanisms was limited and thus, we lack a systematic understanding of how work-life balance support is influenced by interactions between multilevel social systems. Lin et al. (2017) illustrated the interplay between support mechanisms at the workplace and micro levels by highlighting how the amount of support individuals offered to their colleagues influenced their provision of support to their spouses. Several qualitative studies explored the interplay between the societal and micro levels, underlining that people relied more on family social support given a national culture of familism and lack of governmental support and formal childcare (Annink, 2017;Grönlund & Javornik, 2014).
Scholars paid relatively more attention to the interplay between the national and organizational levels. Key findings highlight that employers are more likely to provide family-friendly initiatives given limited state support, a societal norm of considering work-life balance as an organizational responsibility and high union representation (Been et al., 2017;Budd & Mumford, 2006). Also, employees are more likely to utilize organizational initiatives in the context of gender egalitarian cultures, extensive national legislation, high unionization levels and advanced information technology infrastructure (Berg et al., 2013(Berg et al., , 2014Ladkin et al., 2016;Thebaud & Pedulla, 2016).
Some scholars offered a more holistic view of the crosslevel interplay, indicating a complementary relationship between these three levels of work-life balance support mechanisms in society (e.g. Leitner & Wroblewski, 2006;Xiao & Cooke, 2012). For example, British and American employees seek more work-life balance support from their social networks due to the 'limited and piecemeal' provision of national and organizational support (Warren et al., 2009, p. 126). In contrast, Swedish employees enjoy more generous state and employer support and rely less on their personal networks (Crompton, 2006).
In summary, there was a discrepancy between multilevel conceptualization versus operationalization, whereby most research conceptualized work-life balance support mechanisms at the workplace level but measured and analysed the mechanisms at the micro level. However, effective societal-and workplace-level support mechanisms are indispensable for individuals to cope with challenging macro-level stressors, given low individual control over macro contexts (e.g. economic recessions). This conceptualization-operationalization discrepancy may bias research findings, which cannot guide practice because key stakeholders (e.g. individuals, employers and governments) in multilevel environments hold common yet competing interests around work-life balance support that individual-level analysis fails to capture. Hence, we advocate investigations on support mechanisms at the societal and workplace levels, as well as the interplay across levels, and our conceptual framework adapted from Bronfenbrenner's (1986) socio-ecological systems theory can be particularly useful in supporting such endeavours.

Structural dimensions
Structural dimensions were primarily captured in quantitative research with a focus on individuals' utilization of organizational initiatives, national policies and childcare services, as well as the provision of these support mechanisms by employers and governments. The provision and utilization of social support from personal networks (e.g. family members and co-workers) was seldom addressed.
Among studies that explored the intervention dimension, half used a vignette experimental design to examine jobseekers', employees' or managers' reactions to the provision of organizational (e.g. family-friendly policies) or societal (e.g. public opinion support) support mechanisms in hypothetical scenarios (e.g. den Dulk & de Ruijter, 2008;Firfiray & Mayo, 2017;van Steenbergen et al., 2008). Four theoretical articles (e.g. , three field experiments (e.g. Moen et al., 2016) and three case studies (e.g. Gentilesco-Giue & Petrescu, 2008) conceptualized the implementation of support initiatives in organizational settings. Only one study examined the impact of familyfriendly policies enforced by the Spanish government over the economic cycle through a natural experiment (de la Rica & Gorjon, 2016). This stream of studies underlines the importance of sustaining investments, long-term followups and timely adjustments in implementing work-life balance support initiatives to ensure their effectiveness and avoid backfires.
A few experimental studies considered the timing dimension by tracking how national or organizational support interventions took effect over time (e.g. de la Rica & Gorjon, 2016;Moen et al., 2016). Considerable longitudinal qualitative and/or quantitative research investigated support mechanisms from other individuals (e.g. Kim & Hollensbe, 2018), organizations (e.g. Cheng et al., 2014) and governments (e.g. Bünning & Pollmann-Schult, 2016) over a period of time (from several days to multiple years). The long-term evaluation of organizational and societal support mechanisms is important because it may take 5 years or more for employers to enjoy returns (e.g. cost reduction in wages and lower turnover) on their provision of family-friendly benefits (Baughman et al., 2003). The daily diary or experience sampling method that collects longitudinal data at one or multiple time points per day over several weeks (e.g. Pluut et al., 2018) recently gained popularity. It helps capture the subtle, ongoing changes in the impact of family/workplace social support on individuals' work-life experiences within a short timeframe (Shockley & Allen, 2013). Limited but important qualitative work addressed the variation in the utilization and the effect of support mechanisms across individuals' life/career stages (e.g. Loretto & Vickerstaff, 2015), between different generations (e.g. Brandth, 2017) or over a period of history (e.g. Thörnqvist, 2006), highlighting that individuals' work-life experiences are constantly being shaped by time-varying contexts.

Functional dimensions
Functional dimensions were explored primarily in qualitative studies, partially because qualitative work often reflects more nuanced subjective evaluation on support mechanisms. The majority of extant research focused on employees' perceived availability of organizational initiatives and family/workplace social support. The expectation dimension was addressed predominantly by perceived organizational support (Gurbuz et al., 2013) and organizational cultures such as family-supportive culture (Bayazit & Bayazit, 2019) and ideal worker image (Mescher et al., 2010), occasionally by national/societal norms such as work centrality  and familism (Riva, 2016), and very rarely by family norms around domestic labour division (Krys et al., 2018). An under-examined but critical indicator of the expectation dimension was the perceived shared responsibility between individuals and surrounding social systems (e.g. families, organizations and the state) for achieving work-life balance, which likely influenced decisions on both governments'/organizations' provision and employees' utilization of support mechanisms (Peper et al., 2009;Remery et al., 2003). This indicator also remains central to our proposition that we need to examine and provide work-life balance support in a systematic, multilevel framework (see Figure 1), as suggested by the socio-ecological systems theory (Pocock et al., 2012).
The perceived usefulness dimension was primarily operationalized as employees' perceived benefits before or after using organizational support initiatives (e.g. Vyas et al., 2017) and, occasionally, as managers' or employers' perceived returns on support provision (e.g. McCarthy et al., 2010;Morris et al., 2011). Similarly, the perceived consequence dimension was largely examined as employees' perceived career consequences (e.g. lower chance of promotion) after using organizational support mechanisms (e.g. Cannizzo et al., 2019) and, occasionally, as managers' or employers' perceived negative outcomes (e.g. Been et al., 2016). Some theoretical articles offered a comprehensive overview of the perceived usefulness and perceived consequence dimensions in relation to employee use of organizational support (e.g. Bardoel & de Cieri, 2008;Beauregard, 2011;Perrigino et al., 2018).
The perceived necessity dimension was examined mainly in terms of employees' perceived demand for, or jobseekers' anticipation of, organizational support initiatives (Blair-Loy & Wharton, 2004;Mansour & Tremblay, 2018) and less often childcare services and public policies (Winston et al., 2019). Some scholars also conceptualized the importance of employers' provision of work-life benefits in response to societal changes in legislation, business environments, labour markets and employees' per-sonal demands (e.g. Abbott & de Cieri, 2008;Bretherton, 2008;Kossek, 2006;Roberts et al., 2004). Only Oliver (2012) underlined mobile researchers' demands for Europeanlevel legislation against the international context of flexible employment across borders.

Interplay between dimensions
Some quantitative research explored the interplay between different dimensions of work-life balance support. One research stream focused on the impact of employers' and governments' provision and intervention, as well as individuals' perceived necessity, perceived availability, expectation, perceived usefulness and perceived consequence, respectively, on individuals' utilization of support mechanisms (e.g. Asiedu-Appiah & Zoogah, 2019;Dikkers et al., 2007). Another research stream concentrated on the impact of employers' and employees' perceived necessity, expectation, perceived usefulness and perceived consequence, respectively, on employers' provision of support mechanisms (e.g. Adame-Sánchez et al., 2018;Remery et al., 2003). Some research touched upon how employers' and governments' provision and individuals' perceived availability and utilization of support mechanisms, respectively, framed the expectation (e.g. organizational cultures) around support mechanisms (e.g. Butts et al., 2013;Parker & Allen, 2001). Only a handful of quantitative studies compared the differential impacts of the provision, perceived availability, utilization and perceived usefulness of support mechanisms on individuals' work and lives (e.g. Jones et al., 2008;Rajan-Rankin & Tomlinson, 2013).
Two key findings emerged. First, there is a provisionutilization gap whereby organizational/governmental provision of support mechanisms does not necessarily lead to individuals' effective utilization (Rajan-Rankin & Tomlinson, 2013). Second, individuals' perceived availability and perceived usefulness may be more important than their actual utilization of support mechanisms for generating positive work and life outcomes (Jones et al., 2008;Muse et al., 2008).
Furthermore, some qualitative studies (e.g. Daverth et al., 2016;McKee et al., 2000), literature reviews (e.g. Dengate, 2016;Thörnqvist, 2006) and theoretical articles (e.g. Martin, 2017; theorized the multidimensional interplay in a more nuanced and comprehensive way, formulating a series of unidirectional and bidirectional relationships between multiple structural and functional dimensions of work-life balance support mechanisms. Based on their findings and our objective of providing new insights into the interactions between individuals' subjective perception and the physical reality of their environment, we conceptualize the interplay F I G U R E 4 A process-oriented overview of the interplay between multiple dimensions of work-life balance support between multiple dimensions of work-life balance support mechanisms via three iterative processes-decisionmaking, realization and subjective feedback (see Figure 4).
Extant research has identified three major processes. Decision-making denotes the process of employers'/managers' or policymakers' policy design, planning and provision based on their subjective interpretation of the environment (Abbott & de Cieri, 2008;Adame-Sánchez et al., 2018); while for employees, it is the process by which they judge whether to use specific support mechanisms (Blair-Loy & Wharton, 2002). Realization refers to the process whereby plans of support mechanisms are translated into reality (Nord et al., 2002;. Subjective feedback depicts the process of how the provision or utilization of support mechanisms reshapes people's subjective perception (McCarthy et al., 2010;Ryan & Kossek, 2008). We summarize the interplay between multiple dimensions and the connection between these three processes below.
First, there is a unidirectional causal chain of dimensions, that is, the realization process, flowing from organizations' or governments' provision and subsequent intervention of support mechanisms to employees' perceived availability and following utilization of support mechanisms .
Second, some bidirectional relationships have been identified between elements of this causal chain (i.e. provision, intervention, perceived availability and utilization) and multiple functional dimensions (i.e. expectation and perceived necessity/usefulness/consequence).
On the one hand, research has articulated the decisionmaking process of employers' or policymakers' provision, as well as employees' utilization of support mechanisms based on their subjective perception (i.e. expectation and perceived necessity/usefulness/consequence) tied to the environmental reality (Been et al., 2016;McCarthy et al., 2010) prior to the realization process.
On the other hand, studies have conceptualized the subjective feedback process through which employers' and governments' provision and intervention, as well as individuals' perceived availability and utilization of support mechanisms, transform people's subjective perception (i.e. expectation and perceived necessity/ usefulness/consequence) of work-life balance support (Brumley, 2014;Ryan & Kossek, 2008;Thörnqvist, 2006) after the realization process.
Third, subjective feedback reshapes people's subjective perception (i.e. expectation and perceived necessity/usefulness/consequence) of work-life balance support and social environment, and can further improve employers'/managers' and governments' decision-making and subsequent realization of support mechanisms (Bardoel & de Cieri, 2008;Nord et al., 2002;Ryan & Kossek, 2008;Thörnqvist, 2006). Hence, the sequential processes of decision-making, realization and subjective feedback form an iterative spiral with the timing dimension embodied across all these iterative processes.
In summary, we use structural dimensions versus functional dimensions to conceptualize the objective reality versus subjective interpretation of work-life balance support. Building on extant literature, we theorize the multidimensional interplay of work-life balance support in terms of three iterative major processes, that is, decision-making, realization and subjective feedback (see Figure 4), providing a comprehensive framework for evaluating the effectiveness of the design, implementation and appraisal of support mechanisms.

Antecedents, outcomes and contexts of work-life balance support: A process-oriented and multi-stakeholder perspective
In line with our conceptual framework (see Figure 1), we synthesized the antecedents, outcomes and contexts in consideration of (1) the process of resource changes, (2) multilevel environment, (3) multi-stakeholder and multidisciplinary perspectives and (4) multiple dimensions of work-life balance support.

Antecedents
Limited research examined the antecedents of work-life balance support, resulting in limited knowledge about important societal, organizational and personal factors that shape decisions of employees' utilization and governments'/employers' provision of specific support mechanisms, which are also important for effective policy design and implementation. Scholars mainly studied the antecedents of employees' perceived availability and utilization of support mechanisms in terms of governments'/employers' provision of support mechanisms and individual and family characteristics, such as gender, occupation, income level and family structure (e.g. single/couple, one/two parent and with/without children) (e.g. Ollier-Malaterre & Andrade, 2016; Wharton et al., 2008).
The antecedents of the state's provision were investigated or conceptualized in terms of national characteristics, such as a country's culture (e.g. gender egalitarianism, paternalism and individualism/collectivism), level of industrialization and urbanization, ideology (e.g. neoliberalism, capitalism and socialism) and governmental/party affiliation, and international factors such as globalization and European Union legislation (e.g. Barsoum, 2019;Craig et al., 2010;Dona, 2012;Rajan-Rankin & Tomlinson, 2013).

Outcomes
Different research disciplines emphasized different outcomes of work-life balance support mechanisms. Management literature primarily focused on the impact of organizational policies on employees' work-related outcomes and sometimes on organizational outcomes. Applied psychology literature mainly addressed the impact of workplace and family social support on employees' wellbeing. Industrial relations literature stressed the societal outcomes of national policies and unions, as well as the impact of organizational policies on organizations and employees. Family studies literature showed a broader interest in the impact of national and organizational policies and workplace/family social support on societal outcomes and working parents' wellbeing and parenting outcomes. These themes were also evident in sociology literature, albeit involving broader social contexts such as the cross-national comparison or historical development.
Most research across these disciplines examined microlevel outcomes. Support mechanisms generally generated positive micro-level outcomes, with very few studies highlighting the mixed impact of organizational family-friendly policies on individuals' personal resources for achieving work-life balance, such as the length and sovereignty of working time (e.g. Hildebrandt, 2006;Hill et al., 2010), parenting time (e.g. Reimer, 2015) and wages (e.g. Fang et al., 2019). Scholars focused primarily on cross-domain outcomes associated with individuals' wellbeing (e.g. work-family conflict/enrichment, depression and cardiovascular health) (e.g. Berkman et al., 2010) and workrelated outcomes conducive to achieving organizational profitability, such as work performance, job satisfaction and turnover intentions (e.g. Cook, 2009;Moon & Roh, 2010). Less attention was paid to life-related outcomes for individuals (e.g. family performance and parenting activities) (Estes, 2005;Las Heras et al., 2017) and their significant others (e.g. children's wellbeing and partner's parental satisfaction) (Matias et al., 2017;Millar et al., 2012). Life-related outcomes beyond family lives, such as leisure (Lin et al., 2013) and friendship (Pedersen & Lewis, 2012), were also ignored.
Workplace-level outcomes were examined less frequently, and findings were mixed. Scholars paid more attention to profitability-oriented outcomes, principally: (1) economic returns, such as organizational/team performance/profitability/efficiency/productivity, share price and wage reduction (Arthur, 2003;Baughman et al., 2003;Morris et al., 2011) and (2) human capital, such as customer satisfaction, service quality and employee retention/turnover/motivation/morale/absenteeism/ commitment (Cogin et al., 2018;Morris et al., 2011;Okechukwu et al., 2016). However, social responsibilityoriented outcomes, such as gender/motherhood pay gap (Fuller & Hirsh, 2019), perceived fairness/backlash of family-friendly benefits (Beauregard, 2011;Wilkinson et al., 2018) and public image (Morris et al., 2011), remained under-explored. Furthermore, limited research assessed the actual/estimated costs of employer-sponsored support, such as healthcare costs (Morris et al., 2011). No cost-benefit or cost-effectiveness analysis was conducted on organizational support initiatives, with the exception of Baughman et al.'s (2003) and Drago et al.'s (2001) implication that employers could potentially offset partial costs of family-friendly benefits through wage reduction.
Very few studies examined societal-level outcomes and revealed intricate results. Emphasis was placed on fairnessbased outcomes. The state and workplace support pro-moted the awareness of gender equality but controversially reinforced the gendered division of domestic labour and occupational segregation (Gronlund & Magnusson, 2016;Singley & Hynes, 2005). National family policies bridged the happiness gap between parents and nonparents (Glass et al., 2016), but might be ineffective to alleviate child poverty (van Mechelen & Bradshaw, 2013), gender/motherhood-related pay gap (Budig et al., 2016) and the gender, educational and urban-rural inequalities in policy use and work-life experiences (Fuller & Hirsh, 2019;Glauber & Young, 2015) due to problematic policy design. Scholars also revealed generally positive efficiencybased outcomes associated with the national economic competitiveness (Earle et al., 2011) and the maximization of current or future labour supply (Brough et al., 2008), such as maternal employment rates (Turki, 2017), mothers' labour force persistence (Baird & Burge, 2018), unemployment rates (Earle et al., 2011), fertility rates and childbearing intentions (Fahlen, 2013). However, the costs of state support, such as public spending on family benefits (Turki, 2017), were rarely documented, rendering a costeffectiveness evaluation impossible.
Extant research underplayed mutual and/or competing interests between multiple stakeholders and failed to align its stance (i.e. employer/employee/policymaker perspective) with key stakeholders' actual interests or demands, distancing research from practice. Such a research-practice gap may result in inefficient national and organizational investments in work-life balance support, rendering a lose-lose situation for employees, families, employers, governments and other stakeholders. Very few researchers explicitly highlighted that employeeled organizational support mechanisms were more effective than employer-driven ones (e.g. Ollier-Malaterre & Andrade, 2016). Most work is underpinned by the tacit assumption that organizational provision of support mechanisms is for the benefit of employees; but in fact, employers usually prioritize their business goals, labour market pressures and legal forces over employees' actual demands and preferences for support mechanisms (de Menezes & Kelliher, 2011;den Dulk, 2005;McDonald et al., 2013). Similarly, public policies such as maternity protection and child benefits are designed to facilitate individuals' worklife balance and thereby promote gender equality, increase labour supply and sustain long-term economic prosperity (Brough et al., 2008;Bünning & Pollmann-Schult, 2016), but problematic policy design and implementation may backfire.
Generally, outcomes were frequently measured as selfreported subjective perceptions (e.g. work-family conflict and job satisfaction) rather than objective or multisource assessments (e.g. share price and supervisorrated performance). Scholars also focused more on individual experiences rather than organizational and societal outcomes. Specifically, researchers considered more: (1) work-related than life-related outcomes at the micro level; (2) profitability-oriented than social responsibilityoriented outcomes at the workplace level; and (3) fairnessbased than efficiency-based outcomes at the societal level. Due to a lack of workplace-and societal-level examination and objective and multisource assessments, extant research evidence on the effectiveness of work-life balance support mechanisms may be inaccurate, biased and incomprehensive to provide useful and feasible guidance for practice, particularly at the workplace and societal levels.

Contexts
Extant research was largely conducted in industrialized or post-industrialized, economically developed, capitalist countries in Europe and North America. Therefore, populations subsisting in diverse cultural, economic and political contexts in the Global South were largely underrepresented. This narrow focus of extant work on research contexts limited its theoretical implications and practical impacts for broad and diverse populations.
Social and historical contexts were under-explored and the international level was considered mainly with respect to globalization, economic recessions and international migration (e.g. Been et al., 2016;Krys et al., 2018). Political and social stability was the default societal setting in the existing work. However, large swathes of the global population do not live in affluent and peaceful societies. Hence, more research is needed to understand how to better support people's work and lives, particularly their recovery from turbulent and complicated societal contexts, such as wars, political and social unrest, economic uncertainty, disease outbreaks and natural disasters.
Extant literature on organizational, family and personal contexts primarily studied the impact of national and organizational support for juggling working and parenting on middle-class, heterosexual, dual-earner parents in managerial or professional jobs. Limited research attention was paid to low-income workers (Kossek & Lautsch, 2018), elderly employees (Loretto & Vickerstaff, 2015), immigrants (Rudolph et al., 2014), single parents (Konrad & Yang, 2012), single and/or childless people (Skinner & Pocock, 2011), disabled people, LGBTQ+ groups and religious groups. Also, research primarily focused on childcare support but downplayed support for other important components of life, such as dependent care for the elderly/disabled/sick, housework, healthcare, education, recreation, networking and philanthropy.

A RESEARCH AGENDA FOR WORK-LIFE BALANCE SUPPORT
Based on extant literature, our conceptual framework highlights two major research avenues in work-life balance support.
First, a multilevel and process-oriented investigation of the spiral process of interactions between individuals and their environment through resource changes, specifically highlighting the: (1) influence of stressors and work-life balance support from multilevel social environments on individuals' personal resources; (2) impact of individuals' investment of personal resources and employers' and governments' investment in work-life balance support on multilevel environments, that is, personal, family, organizational and societal outcomes; and (3) effect of these personal, family, organizational and societal outcomes on resource expansion, accumulation or replenishment. For example, following Bronfenbrenner's (1986) socio-ecological systems framework, scholars can explore how the level of industrialization (societal contexts) shapes work and family demands and public/private childcare/eldercare services (support mechanisms), which in turn influence individuals' time allocation for work, family and leisure (personal resources) and subsequent personal, family, organizational and societal outcomes longitudinally in one country or horizontally across multiple countries.
Second, a multidimensional and process-oriented investigation of three iterative processes of interactions between individuals' subjective perception (i.e. functional dimensions) and physical environment (i.e. structural dimensions) of work-life balance support, as follows: (1) the decision-making process of evaluating the environment (e.g. analysing environmental stressors and individuals' corresponding personal demands) and planning subsequent utilization or provision of work-life balance support mechanisms; (2) the realization process of translating plans of support mechanisms into the environmental reality (i.e. provision → intervention → perceived availability → utilization); and (3) the subjective feedback process of reshaping individuals' subjective perceptions of the environment and providing new information for the next iteration of decision-making process. For instance, researchers can compare the similarities and differences in the decision-making, implementation and effectiveness evaluation processes of national and organizational worklife balance support provision in liberal (e.g. the USA) versus socialist (e.g. China) economies.
Our review identifies some major gaps in the extant work-life balance support literature. From a multilevel perspective, as recommended by Bronfenbrenner's (1986) socio-ecological systems theory, there is limited examination of higher-level factors, such as societal-and workplace-level support mechanisms and their family, organizational and societal antecedents and outcomes. Also, little attention has been paid to broader societal and historical contexts. From a multi-stakeholder perspective, there is a lack of systematic consideration of different stakeholders' (i.e. individuals, employers and governments) shared and competing interests. The narrow focus of extant research on micro/individual-level analysis limits its scope and capacity for guiding societal and organizational practice concerning broad, diverse populations holding vastly different interests.
From a process-oriented perspective, there is insufficient understanding of the long-term impact and sustainability of work-life balance support mechanisms. From a multidimensional perspective, limited research has addressed the iterative, bidirectional interactions between individuals' objective and subjective work-life experiences and their social environment. The predominance of static, oneoff analysis using cross-sectional designs fails to capture how the iterative interactions between reality and perception shape the effectiveness of work-life balance support mechanisms from perspectives of different stakeholders in the long term, which holds critical implications for governments' and organizations' effective design, implementation and improvement of their support mechanisms.
We propose three major research directions in terms of (1) cost-effectiveness analysis, (2) pluralist thinking and (3) context specification to address these gaps.
Cost-effectiveness analysis. Extant literature rarely compared the costs and returns of work-life balance support mechanisms and hence, we have limited insight on their cost-effectiveness. Therefore, we recommend a costeffectiveness, or equivalent, assessment of key stakeholder interests in future research.
Scholars should consider individuals' costs in terms of their investment of personal resources (e.g. time and money) and returns in terms of the degree of fulfilment of work/life demands and other work/life-related and crossdomain personal outcomes. Additionally, future research could capture governments' and employers' interests by weighing their costs (i.e. investments) of providing and implementing work-life balance support mechanisms (e.g. expenditures and human resources) and returns encompassing profitability-and social responsibility-oriented organizational outcomes and efficiency-and fairness-based societal outcomes. However, it should be noted that not all costs and returns can be quantified or monetized on a comparable scale and hence, some functional dimensions of work-life balance support mechanisms (i.e. perceived usefulness/consequence) can be used as proxy measures to estimate their cost-effectiveness.
Moreover, cost-effectiveness concerns remain central to the interactions between individuals' objective and subjective work-life experiences characterized by three iterative processes-decision-making, realization and subjective feedback. By incorporating sensemaking, decisionmaking and stress and coping literature into work-life balance support research (e.g. Golden, 2009;Powell & Greenhaus, 2006;Somech & Drach-Zahavy, 2012), scholars can investigate whether employers' and policymakers' needs assessment, policy design, implementation and subsequent outcomes are aligned with their business and societal goals, and whether individuals' needs assessment, utilization and associated personal outcomes of work-life balance support mechanisms successfully meet their worklife demands. More investigation on the iterative processes is needed, particularly on how employers/policymakers adjust and refine their decisions and subsequent implementation of support mechanisms when their intervention is not aligned with their goals and/or employees'/citizens' needs.
Pluralist thinking. We recommend a pluralist approach to work-life balance support that not only acknowledges the diversity of goals and interests across different stakeholders but also emphasizes the synergy of shared goals and reconciliation of competing interests in order to maximize the overall benefits for all stakeholders involved (Gomez et al., 2004).
As previously mentioned, each stakeholder seeks longterm sustainability by balancing competing goals (e.g. individuals balancing work vs. life demands). Therefore, we suggest that scholars bridge the research-practice gap by aligning their investigation of the impact and effectiveness of work-life balance support with stakeholders' actual demands, interests and/or goals. Following our recommended pluralist approach, scholars should also recognize that shared and competing interests within and between stakeholders across different levels of social environment systematically influence the effectiveness of work-life balance support mechanisms. Two typical scenarios and relevant examples are given below. Some support mechanisms may enable a win-win situation by achieving the mutual interests of multiple stakeholders. For instance, family-friendly practices may allow employers to strengthen employer branding, improve productivity and reduce costs (e.g. lower entry-level wages and fewer labour conflicts) and facilitate individuals' work-life reconciliation (Baughman et al., 2003;Faria & Machado, 2018).
However, other support mechanisms may fail to reconcile the conflicting interests of multiple stakeholders by achieving partial goals or one-sided interests. For example, the government legislates delayed retirement to increase the current labour supply and reduce social welfare expenditures (e.g. pensions), which can decrease grandparental childcare and maternal labour force participation (Belan et al., 2010;Du et al., 2019).
Context specification. Consistent with previous literature reviews (Gatrell et al., 2013;Ozbilgin et al., 2011), we found that current work largely focused on working parents through a Western lens, especially professional mothers situated in economically developed and industrialized countries. However, a majority of the global population works and lives in vastly different contexts. Scholars need to carefully consider the extent to which worklife balance support themes studied in Western countries can be applied in other countries and contexts. European countries enforce generous family benefits and delayed retirement to increase female and elderly people's workforce participation and secure future labour supply given the ageing crisis. However, these public policies are not applicable in many Global South countries with large, predominantly young, unemployed and under-educated populations. The policy priorities of these countries may be improving the quality rather than the quantity of labour supply by offering better educational and job opportunities.
Work-life balance support scholars need to explicitly consider the different priorities and objectives of stakeholders in specific contexts (accounting for variations in populations and associated family, organizational and societal/national dynamics), with cautious generalization across contexts. Below, we give specific recommendations to achieve context specificity with respect to individuals, organizations and countries.
First, individuals have different work and/or life priorities and demands at different stages of life/career and in different family and community contexts. More research is needed on support mechanisms aimed at helping individuals from non-traditional families (e.g. singles, homosexual couples and dual earners without children) at the early/middle/late stage of their life/career and situated in social/religious/ethnical communities with unique cultural norms, traditions and rituals.
Second, employers' provision and employees' utilization of support mechanisms vary substantially across different organizational contexts. These differentials in organizational contexts need to be unpacked in terms of the size, workforce composition (e.g. gender and age), industry (e.g. manufacturing and technology), sector (i.e. public/private), degree of unionization and business/industry lifecycle.
Finally, the provision and utilization of work-life balance support mechanisms are also influenced by a country's social, economic, political, cultural and historical contexts. Some important contextual factors are the level of industrialization and urbanization, the level of economic development, political ideologies and regimes, cultures, infrastructures and public security. National contexts are also shaped by a country's role in the global economy, relative national power and geopolitical influence, and the idiosyncrasies of international environment (e.g. economic recessions and wars). More importantly, a country's national context needs to be considered longitudinally along its own historical development and horizontally in comparison to other countries. Future research can utilize a series of recommended national-level structural, economic, social and cultural measures (e.g. den Kossek, 2015;Ollier-Malaterre & Foucreault, 2017;Ollier-Malaterre et al., 2013;Trefalt et al., 2013).
The temporal context (i.e. timing dimension) is also critical for the effectiveness of work-life balance support mechanisms, because key stakeholders prioritize different interests and goals over time. We have a series of recommendations for addressing temporal contexts.
First, we recommend a short-term longitudinal design (i.e. daily diary method) to examine the impact of social support from professional and personal networks on individuals' fluctuant personal resources (e.g. energy) and personal outcomes (e.g. health).
Second, a long-term longitudinal design (e.g. natural/field experiments lasting several months/years) is recommended to track the impact of organizational and national support mechanisms that may take longer to generate a return.
Third, the psychological life-span theory and the sociological life-course theory (Shanahan & Porfelli, 2002), as well as life history interviews (Jessee, 2019), can be applied to capture changes in individuals' work-life experiences across different life stages and across generations.
Fourth, life history interviews and ethnographic approaches (Singh & Dickson, 2002) can help track systematic changes of work-life balance support (e.g. both institutional and cultural changes) in organizational settings over the business lifecycle or in national/societal settings across different periods of history.

CONCLUSION
We synthesized extant findings in line with our proposed conceptual framework that unpacks the critical role of work-life balance support in iterative interactions between individuals' objective and subjective work-life experiences and their multilevel social environments through resource changes and reality-perception transformation.
Our review contributes to work-life balance research by providing a multilevel, multidimensional, processoriented framework for systematically analysing how (1) the interplay between individuals and multilevel social environments, (2) the interactions between individuals' objective and subjective work-life experiences and (3) multiple stakeholders' shared and competing interests shape the effectiveness of work-life balance support mechanisms. As such, we challenge and expand the traditional understanding of work-life balance as an individual-level phenomenon into a societal concern that necessitates systematic support from a wider range of stakeholders, such as families, organizations, trade unions, governments, communities, public and private services. To conclude, our conceptual framework and research agenda pave the way for multilevel, multidimensional and process-oriented investigation of work-life balance support in the future.

A C K N O W L E D G E M E N T S
We highly appreciate the editors and reviewers for a very constructive review process. Furthermore, we would like to extend our sincere gratitude to the participants at the conferences for their valuable comments. Previous versions of this paper were presented at the European Asso-