Defined as “the continuous fulfilling, acquiring or creating of work through the optimal use of competences” (Van der Heijde and Van der Heijden, 2006, p. 453), employability has become increasingly important to cope with fast-changing job requirements and to find career opportunities in internal or external labor markets more easily. In view of the aging world population, this also holds true for the category of workers being over 55 years old (cf. European Council, 2001) who have to continue working longer in order to make retirement systems more sustainable (Verbrugghe et al., 2016; Davies et al., 2017). Therefore, in the current scholarly debate, employability has been considered as a personal resource (cf. De Coen et al., 2015) that may influence workers' retirement decisions (Wang and Shultz, 2010; Feldman and Beehr, 2011).

Up until now, however, many employers tend to invest less in the training and development of their older workers compared to their younger counterparts, as they consider investments in them not to pay-off, just because older workers are closer to retirement. Older workers, in turn, may perceive a lack of support for learning in their job (i.e., the degree to which the job design promotes continuous learning and stimulates workers to acquire new knowledge and skills; Tracey and Tews, 2005) and may encounter negative age stereotypes (e.g., regarding older workers being less productive than younger ones; Posthuma and Campion, 2009) which may influence older workers' own expectations regarding the so-called ‘return on investment’ of their participation in HR-activities that are geared at enhancing their employability as well, and, in turn, their decision to stay or leave the labor market (cf. Greller and Simpson, 1999; Schermuly et al., 2014). More specifically, one of the most common age stereotypes regarding older workers is their alleged lower productivity and job performance (Posthuma and Campion, 2009). As a result, companies prefer to seek and hire younger workers and to invest in their further career growth. Furthermore, older workers may feel stigmatized and stereotyped and may opt, therefore, for (earlier) retirement (Schermuly et al., 2014). In other words, a self-fulfilling prophecy (Van der Heijden, 2005) may be generated in case older workers don't perceive their job and their organization to stimulate their ongoing personal development, and, as a result, they may even decide to seek new job opportunities elsewhere or to retire (prematurely) (cf. Armstrong-Stassen and Ursel, 2009; Zaniboni et al., 2010; De Wind et al., 2017).

It is well-known that the number of older workers has grown over the past decades, and that the more experience workers accumulate over the years, the more their needs vary in terms of the role work plays in their lives, leading to an ever-increasing workforce heterogeneity (Carstensen, 2006). Consequently, the moderating effect of age is a highly important, yet complex, issue to take into account in scholarly work in this field (cf. Rudolph, 2016; Van der Heijden et al., 2016). Up until now, only few empirical studies have looked into the association between, in particular, older workers' employability and their intention to retire (De Coen et al., 2015; Hennekam, 2015). Moreover, it can be questioned, whether the mechanisms or processes underlying the alleged relationships between job support for learning and negative age stereotypes (being the predictors), employability (being the mediator), and intention to retire (being the outcome) apply to different categories of older workers in similar ways. The present study aims to contribute to the scholarly debate on older workers' employability and their intention to retire (Van der Heijden et al., 2009; De Graaf et al., 2011) by studying the mediating role of older workers' participation in employability enhancing activities in the relationships between one organizational (‘perceptions of job support for learning’) and one job-related factor (‘perceived negative age stereotypes on productivity’—from now on called ‘perceived negative age stereotypes’) on their intention to retire, and the moderating role of age category in this regard. The following research questions will be addressed: Do older workers' perceptions of job support for learning' and of ‘negative age stereotypes’ in their organization influence their ‘intention to retire’? Are these relationships mediated by older workers' ‘participation in employability enhancing activities’? And are these relationships similar across workers in the age group ranging from 55 to 60 years old compared to those aged from 61 till 65 years old?

This study addresses some relevant issues. First, Van der Heijden's (2005) employability enhancement model hypothesized personal factors, job-related factors and organizational factors to be important determinants of workers' employability, which, in turn, predicts work and career-related outcomes (cf. Van der Heijden et al., 2009; De Cuyper et al., 2011; De Vos et al., 2011; De Coen et al., 2015). The integrative framework to employability enhancement as conceptualized in Van der Heijden's conceptual model builds upon insights from employability research, human resource management and work, and organizational psychology (see Van der Heijden, 2005) and is used as our theoretical basis.

More specifically, much in line with the plea for an integrative approach (De Vos et al., 2011), our theoretical lens will build on this model by looking into age (being an individual-based factor) (comparing employees of 55–60 vs. 61–65 years old), job support for learning (Tracey and Tews, 2005) (as a job-related factor), and perceived negative age stereotypes (Hassell and Perrewe, 1995) (as an organizational factor) in order to predict both workers' employability and intention to retire (being a career-related outcome). We will combine Van der Heijden's (2005) model with insights from the Job Demands-Resources (JD-R) framework (Bakker and Demerouti, 2007), the Conservation of Resources (COR) theory (Hobfoll, 1989), and the retirement literature (e.g., Resources-Based Dynamic Model for Retirement Adjustment; Wang and Shi, 2014).

Second, we will focus on workers in their later career stages (55 and over) in Italy, being one of the countries in Europe with the oldest (working) population (cf. Eurostat, 2015). Because of the serious lack of studies including workers in their late career (cf. De Lange et al., 2010; Müller et al., 2015) and the absence of research on different age groups within the older workers category, we will draw upon lifespan developmental theories (Baltes and Baltes, 1990; Carstensen et al., 1999) to investigate the moderating effect of age in the hypothesized relationships.

The present study aimed to contribute to the debate on older workers' sustainable employability and retirement considerations by thoroughly testing a new empirical model, based on Van der Heijden's (2005) employability enhancement framework, while incorporating additional key scientific literature. It was intended to unravel the relationships between older workers' perceptions of job support for learning and negative age stereotypes, their participation in employability enhancing activities, and their intention to retire. Using our full sample, empirical evidence for two mediation paths, wherein older workers' participation in employability enhancing activities acted as a mediator was found. More specifically, partial mediation was found for the relationship between perceived job support for learning and intention to retire, while full mediation was found for the relationship between perceived negative age stereotypes and intention to retire. Furthermore, empirical support was found for the moderation hypotheses predicting that these mediation effects differed among the two distinguished age groups. In line with our expectations, the mediation effect was stronger for workers aged 55–60 years old in comparison with their relatively older counterparts (61–65 years old). Below, an elaboration of the main results of our study is provided.

In line with related studies (Armstrong-Stassen and Ursel, 2009; Schreurs et al., 2011; Hofstetter and Cohen, 2014), our empirical work revealed that perceived job support for learning [viewed as a job-related resource (cf. Demerouti et al., 2001)] has the potential to reduce older workers' intention to retire. Further, in line with earlier scholarly work (cf. Von Hippel et al., 2013; Hofstetter and Cohen, 2014; Schermuly et al., 2014), it was found that perceived negative age stereotypes [viewed as an organizational demand (cf. Demerouti et al., 2001)], was positively associated with intention to retire. Hence, both the job-related and the organizational factor that we included in our model (i.e., positive perceptions on the room for development in their job, and negative perceptions on their organization's culture, in terms of the prevalence of negative age stereotyping) have proven to be important factors in older workers' retirement intentions (cf. Carr et al., 2016).

Investigating the direct relationship between older workers' participation in employability enhancing activities and intention to retire provided empirical support for the theoretical claim that employability [(viewed as a personal resource; cf. De Cuyper et al., 2012; De Coen et al., 2015)] has the potential to mediate relationships between job-related and organizational characteristics, such as job resources (i.e., job support for learning) and job demands (i.e., perceived negative age stereotypes), on the one hand, and older workers' intention to retire, on the other hand (cf. Bakker et al., 2003). Our path model revealed two important behavioral mechanisms operating between the work context and older workers' intention to retire.

On the one hand, building upon the JD-R model (Bakker and Demerouti (2007), COR theory (Hobfoll, 1989), and the Resources-Based Dynamic Model for Retirement Adjustment (Wang and Shi, 2014), the present research found a solid ‘positive gain spiral’ wherein older workers' perceptions of job support for learning motivates them to participate in employability enhancing activities (conservation of personal resources), which, in turn, reduces their intention to retire (cf. De Cuyper et al., 2011). Possibly, individuals who have accumulated more personal resources, as a result of them having had better chances to grow in their job, feel that their sustainable employability allows them to prolong their working life.

On the other hand, a ‘negative loss spiral’ (Hobfoll, 1989) was found as well, in which older workers' perceptions of negative age stereotypes discourages their participation in employability enhancing activities, which, in turn, increases their intention to retire. Conceivably, older workers' perceptions of negative age stereotypes lower their self-evaluations, which, through a self-fulfilling prophecy, reduces their participation in employability enhancing activities (Van der Heijden, 2005). This, in turn, results in them wanting to leave their working organization sooner. With these outcomes, the present study follows Finkelstein et al. (2015) and calls for a broader perspective on age-related stereotyping, including more empirical work on meta stereotyping (i.e., better understanding of cross-age dynamics, and improving interactions in age-diverse work settings). That is to say, the daily pulse of interpersonal interaction in an organization requires attention for multiple parties involved (such as older workers, colleagues, and managers), which altogether have the potential to affect individual behaviors.

Finally, these findings also confirm the outcomes of a recent contribution by Zaniboni (2015) in which older workers who still had access to personal resources (i.e., learning ability and interest in development) desired to go on working longer in case they perceived a low amount of age discrimination.

Finally, the present research contributed to the debate on older workers' sustainable employability (Van der Heijden et al., 2009; De Graaf et al., 2011) by looking into the rather complex moderating effects of age within the studied relationships. Our multi-group SEM findings supported the assumption that there is a relationship between individuals' so-called ‘life time horizon’ (cf. Carstensen, 2006) and participation in employability enhancing activities, implying that those aged 55–60 years old have a more ‘open-ended future’ in comparison with those aged 61–65 years old, and are more inclined to invest in their further career development (cf. Kochoian et al., 2017) when supported by their jobs. Additionally, in line with lifespan developmental theories (Baltes and Baltes, 1990; Carstensen et al., 1999) and empirical evidence from previous aging research (cf. Baltes et al., 1999; Kanfer and Ackerman, 2004; Freund, 2008; Kooij et al., 2011; Kooij and Zacher, 2016; Moghimi et al., 2016; Rudolph, 2016), our multi-group SEM results confirmed that, with aging, people become more motivated toward maintaining the performance in their current job, instead of focusing on growth and development goals (i.e., being less learning-oriented), and tend to minimize losses (i.e., paying less attention to, focusing less on or avoid situations wherein they encounter negative age stereotypes, and as a result, probably, being less threatened by these).

In particular, our multi-group SEM findings supported outcomes from earlier work suggesting a stronger ‘prevention focus’ (Higgins, 1998) among older workers (Ebner et al., 2006), since the allocation of resources by individuals for acting upon growth goals tends to decline throughout the lifespan while, with aging, the allocation of resources for acting upon maintenance and regulation of ‘loss or prevention’ goals appear to increase (see Freund, 2008; Rudolph, 2016). Concrete, older workers may prefer to focus on what they might gain in their work setting rather than what they might lose. In doing so, they may shift their attention from focusing upon a further career growth to a possible alternative, in this case retirement. It is likely that in the meantime they may value their current position, in particular the positive relationships with their colleagues, and may be less bothered and/or threatened by negative age stereotyping. Moreover, in perceiving negative age stereotypes, older workers may re-evaluate the meaning of work (cf. Hobfoll, 1989; Baltes and Baltes, 1990), resulting in a decreased value that work has for them, which, in turn, may increase their likelihood to retire.

The revealed employee age group effects in this study may be explained by highlighting the 55–60 years old age group's relatively ‘open perception of the future’ in comparison with the 61–65 years old age group who will have a relatively ‘limited time perspective’ (Carstensen et al., 1999). That is, the former age group still has (almost) 10 years of work ahead, whereas the latter may be already more disengaged from their organization's prevailing culture. In fact, the relatively younger employee age group may see retirement more as a future event and, therefore, might be more discouraged by negative organizational factors, in our case perceptions of negative age stereotypes, as these may affect (harm) their actual mindset and immediate future. Alternatively, the 61–65 years old group's awareness of reaching retirement might contribute to striving for experiences that are more positive (for instance focusing on good relationships with close colleagues) in order to get rid of these negative emotions (Carstensen et al., 1999; Henry et al., 2017).

Overall, our multi-group results contribute to the existing literature by applying lifespan developmental theories in a study focusing on ‘older workers’ age brackets. In fact, the present research has demonstrated that, due to its heterogeneity, older workers' age categories should not be taken as one unit of analysis. Based on the effects of age that were found in this contribution, it can be concluded that further elaborating on age-related processes incorporating a ‘future time perspective’ is strongly needed in order to advance the HRM, career and employability literatures, since it allows scholars to better understand why and how age effects occur and how they affect organizational behaviors (cf. De Lange et al., 2010; Rudolph, 2016).

Despite the important contributions of our study as discussed above, there are several limitations that should be acknowledged. First, the data were cross-sectional. Therefore, our conclusions have to be interpreted with care. Second, as all variables in our study were measured using self-reports, a common-method bias may exist (Podsakoff et al., 2003). Moreover, although we believe that the scales reflect meaningful content coverage and uni-dimensionality, implying that the relatively low alphas are not a major impediment to its use (Schmitt, 1996), more reliable measures for perceived negative age stereotypes and for intention to retire might be used in future work. In response to the limitations that were reported above, future longitudinal research, preferably using multi-wave designs in order to better examine and analyze in-depth the causal relationships within the model (De Lange, 2005), and incorporating more objective data, such as, actual retirement age, is called for.

We are aware of the fact that our sample is characterized by a higher percentage of male workers, as in most banking organizations. That is, in future research, both gender and age could be taken into account in multi-group analyses. Moreover, it needs to be said that (in Italy) older workers are more likely to have a managerial role because of their seniority in the organization, being a possible reason for the relatively large amount of respondents who held a managerial position. Therefore, future work is needed in order to rule out possible effects of organizational roles and positions.

Our study has important implications for managers and other key stakeholders in organizations who are in search for measures to enhance older workers' career development, in order to improve both their workers' sustainable employability as well as the performance of their organization. Increasing older workers' added value throughout their lifespan strengthens their chances in both the internal and external labor market (Van der Heijden et al., 1998; Rocco and Thijssen, 2006; Zappalà et al., 2008). In order to achieve these multiple goals, organizations should provide all their workers, with older workers being no exception, with ample job support for learning and should combat a culture that is prone to negative age stereotypes. With regard to the latter, human resource policies should pay more attention to the process of aging and should focus on protecting sustainable employability throughout workers' entire career in order for older workers to feel ‘conserved’ instead of ‘depreciated’ (Peterson and Spiker, 2005; Beach, 2016). Moreover, our study shows that older workers respond differently to job-related and organizational factors than younger workers (cf. Zacher et al., 2017). Our results call for a reconsideration of reflections on ‘return on investment’ regarding the oldest workforce that are ubiquitous and solely based on negative age stereotyping rather than on actual strengths and weaknesses of the specific employee involved. Zwick (2011) highlighted that, often, organizations do not offer appropriate learning opportunities to older workers because firms may view older workers as being less effective; that is, on-the-job training is rarely available for the older counterparts of the workforce. This while the present results show that support for learning increases their participation in employability enhancing activities, and, as a result, reduces their intention to retire. Hence, job design models in working organizations should take into account workers' changing needs across the lifespan (see also Henry et al., 2015). Using a lifespan perspective and carefully paying attention to age-related processes (such as workers' changing future time perspective and their prevention focus), human resource specialists and line managers should focus on supporting older workers' development throughout their entire career.

This study was carried out in accordance with the recommendations of University of Pavia with written informed consent from all subjects. All subjects gave written informed consent in accordance with the Declaration of Helsinki.

PD: conceptualization, data curation, formal analysis, investigation, methodology, project administration, visualization, writing original draft, writing review, and editing; BV and PP: conceptualization, formal analysis, investigation, methodology, supervision, visualization, writing original draft, writing review, and editing; SK: data curation, formal analysis, methodology, software, writing review, and editing; PA: data curation, funding acquisition, resources, supervision, writing review, and editing.