Many, if not most of the employees nowadays will at some point of their careers engage in job search behaviours as a consequence of involuntary job loss and unemployment (Wanberg, 2012). Job search refers to a set of purposive activities conducted in pursuit of an employment goal (Kanfer, Wanberg, & Kantrowitz, 2001). This goal-oriented behaviour is generally seen as beneficial in that it increases one’s chances of reemployment (McKee-Ryan et al., 2005). At the same time, it often requires one to invest a considerable amount of time, energy and other resources without a guarantee of success. As such, unemployed persons engaging in job search behaviours are frequently confronted with high levels of uncertainty and unpredictability combined with many setbacks, disappointments and frustration (Wanberg, Zhu, & Van Hooft, 2010). For these reasons, it seems particularly advantageous to identify inter-individual differences that predispose unemployed persons to a proactive orientation towards reemployment and adaptability to unfavourable circumstances. The recently defined concept of dispositional employability subsumes dimensions that reflect both of these characteristics and, as such, may represent a valuable personal resource for the unemployed (Fugate & Kinicki, 2008).
Dispositional employability has been portrayed as ‘a constellation of individual differences that predispose individuals to (pro)active adaptability specific to work and careers’ (Fugate & Kinicki, 2008: 503). Fugate and Kinicki (2008) suggested that individuals who are able to adapt to their environment in a proactive manner are those who will more successfully pursue new career opportunities and adjust the situation to fit their own needs. For example, those who possess characteristics that predispose them to be more proactively adaptable may self-reliantly identify and pursue new job opportunities (e.g., a desirable job position) while at the same time remaining resilient to disappointment (e.g., rejections from employers). We argue, therefore, that unemployed individuals with high dispositional employability might more intensively engage in job search behaviour.
Notwithstanding its potential benefits, constructs related to proactivity and adaptability have only recently been addressed as antecedents of job search behaviour. For example, proactive personality and career adaptability have been demonstrated to positively relate to one’s engagement in job search activities (Brown et al., 2006; Claes & De Witte, 2002; Zacher, 2013). By examining dispositional employability as an antecedent of job search behaviour, we aim to extend the work of these authors in at least two ways. First, dispositional employability, as operationalized by Fugate and Kinicki (2008), explicitly integrates proactive and adaptable individual characteristics. As such, it represents a broader and more diversified construct than each of the two narrower dispositions alone (cf. Judge et al., 2003). Second, unlike more general notions of proactivity (e.g., proactive personality), it is a domain specific construct anchored to the work and career context. For these reasons, we suggest that dispositional employability may represent a promising, yet understudied antecedent of job search behaviour in an unemployment setting (for two exceptions, see Koen, Klehe, & Van Vianen, 2013; McArdle et al., 2007). Accordingly, the first aim of this study is to examine the relationship between dispositional employability and job search behaviour among unemployed persons.
In doing so, we aim to go a step further by accounting for the hypothesized mechanism that might explain the effect of dispositional employability on engagement in job search activities. Following the logic of the most recent employability models (e.g., Forrier, Sels, & Stynen, 2009; Forrier, Verbruggen, & De Cuyper, 2015), we posit that perceived employability has particular resonance in this regard because it concerns the individual’s perception of how easy it is to obtain new employment (Berntson, Sverke, & Marklund, 2006). Arguably, the postulated mediating role of perceived employability aligns with the subjectivity inherent in this approach. Specifically, an individual’s perception of employment chances should highly depend on one’s self-evaluation of work-related personal characteristics on the one hand (Vanhercke et al., 2014) while, on the other hand, the subjective interpretation of one’s labour market position, rather than objective reality, should guide people’s behaviours (Katz & Kahn, 1978; Lazarus & Folkman, 1984). Therefore, the second aim of this study is to test the mediating effect of perceived employability on the relationship between dispositional employability and job search behaviours.
By addressing these research aims, we believe the present study contributes to the literature in several ways. First, it extends current knowledge on job search antecedents by aligning with the most recent literature advocating the need for individual characteristics that foster proactive adaptability among unemployed (e.g., Fugate & Kinicki, 2008; Fugate, Kinicki, & Ashforth, 2004). Second, the concept of dispositional employability is anchored in a work context and, as such, offers a contrast to the more general dispositional measures that currently dominate job search studies (cf. Kanfer, Wanberg, & Kantrowitz, 2001). This choice reflects a greater match in specificity between the predictor and criterion that, in turn, has the potential to increase the predictive power of the results (see also Woo, Jin, & LeBreton, 2015). Third, the present study encompasses two related, yet conceptually different notions of employability–dispositional (Fugate, 2006) and perceived employability approaches (De Cuyper et al., 2012a) and thus enables empirical testing of the most recent employability models (e.g., Vanhercke et al., 2014). Fourth, while both conceptualizations of employability have thus far been predominantly studied in samples of employed individuals, the present study supplements the few existing studies that investigate either dispositional (i.e., Koen, Klehe, & Van Vianen, 2013; McArdle et al., 2007) or perceived employability (i.e., Chen & Lim, 2012; Onyishi et al., 2015) among unemployed persons. Fifth, in contrast to the majority of previous studies examining job search antecedents, the present study was conducted in an economy characterized by high unemployment rates (19.5% or 333,249 unemployed registered persons), a small number of job vacancies (12,175) and consequently, low work mobility (The Croatian Employment Service, 2015a).1 As such, it has the potential to generalize previous findings to a different contextual setting.
In sum, we propose and aim to test a hypothesized mediational model that specifies dispositional employability as a predictor of job search behaviour via perceived employability. In doing so, we place the focus on job search intensity (i.e., the frequency of engagement in various job search activities, such as applying for available vacancies or contacting potential employers) as an exemplary representation of job search behaviour (e.g., Wanberg, 2012). In the following paragraphs, we further elaborate the hypothesized relationships between (1) dispositional employability and job search intensity, (2) dispositional employability and perceived employability, (3) perceived employability and job search intensity and (4) dispositional employability and job search intensity via perceived employability.
Dispositional employability and job search intensity
Dispositional employability has been conceptualized as a latent higher-order construct reflected in five dimensions: openness to changes at work, work and career resilience, work and career proactivity, career motivation, and work identity (Fugate & Kinicki, 2008). Each of these dimensions are considered critical and representative for a proactive orientation towards adaptability in the work context. Openness to changes at work entails personal adaptability and flexibility, characteristics that are essential for the identification of new job opportunities. Work and career resilience concerns one’s positive self-evaluations, optimism and positive expectations regarding the future (Fugate, 2006). Resilient individuals should be more able to perceive various career opportunities, interpret obstacles as challenges and persist in goal-oriented behaviours (Fugate & Kinicki, 2008). Work and career proactivity predisposes individuals to proactively acquire information regarding developments in their work environments (e.g., occupation, industry). As such, it also enables self-assessment of one’s value on a labour market (Fugate, 2006). Career motivation equips an individual with the ability to persevere in the attainment of career-related goals and persistence in periods of frustration, such as unemployment (Fugate, 2006). Finally, work identity refers to one’s self-definition in the work-related context and, as such, directs, regulates and preserves future behaviours (Fugate, Kinicki, & Ashforth 2004; Fugate & Kinicki, 2008).
Although each of the five dimensions is valid in its own right, it is their synergistic combination that reflects the added value of the dispositional employability construct (Fugate, 2004). As Fugate (2008: 506) noted, ‘[dispositional] employability is conceptually more abstract and has a meaning and influence in the work and career domains above and beyond that of any particular dimension’. As such, it should provide more conceptual and predictive power (Fugate, 2004). Accordingly, we align with this more parsimonious yet complete perspective on dispositional employability and operationalize it as a higher-order factor that underlies each of its dimensions (cf. Judge et al., 2003; Forrier, Verbruggen, & De Cuyper, 2015; McArdle et al., 2007).
We presume that people with higher levels of dispositional employability will more intensively engage in job-search behaviours to counter the demands of unemployment. This notion aligns with the basic assumptions of two stress theories, namely appraisal theory (Lazarus & Folkman, 1984) and conservation of recourses (COR) theory (Hobfoll, 1989). Dispositional employability might be considered a resource in line with both theoretical frameworks: it represents a constellation of personal characteristics that might serve as a means for attainment of valuable work-related resources (e.g., new job opportunities). According to the appraisal theory individuals with more resources (e.g., dispositional employability) should more intensively engage in problem-focused coping (e.g., job search) because they evaluate the problem of unemployment as solvable. By contrast, less resourceful persons are less likely to be oriented towards dealing with the problem because of their perception that nothing can be done about it (Latack, Kinicki, & Prussia, 1995; Lazarus & Folkman, 1984). According to the COR theory, individuals must invest resources in order to gain additional resources (Hobfoll, 1989). Those with greater resources are better able to recover from resource loss and orchestrate resource gain (Hobfoll, 2011). Building on these premises, we suggest that individuals with higher dispositional employability might be more capable to invest in job search behaviours to recover from job loss and to obtain new employment.
Accordingly, individuals with higher levels of dispositional employability might prepare in advance for any potential obstacle in the job search process. For example, they might broaden the scope of their job search channels by contacting employed acquaintances in order to gather relevant information about new job positions or by sending an e-mail to a desirable employer in order to present her/himself. In addition, individuals with higher dispositional employability may be more likely to persist in job search behaviours when confronted with rejection, a lack of job vacancies, or other obstacles. Koen, Klehe and Van Vianen (2013) found that adaptability, operationalized as career exploration and planning, was positively related to job search intensity one year later, above and beyond the influence of various barriers (e.g., physical problems). McArdle et al. (2007) found a positive relationship between dispositional employability and job search intensity in the context of the Australian labour market. However, unlike the present study, the authors in this study conceptualized dispositional employability as a second order factor comprising three dimensions – human and social capital, adaptability and career identity – in line with the previous work of Fugate and his colleagues (e.g., Fugate, Kinicki, & Ashforth, 2004). In consideration of the presented theoretical and empirical arguments, we set the first hypothesis as follows:
Hypothesis 1: Dispositional employability relates positively to job search intensity.
Dispositional employability and perceived employability
Introducing dispositional employability as an antecedent of perceived employability follows the conceptual logic of input- and output-based employability approaches. Namely, dispositional employability accounts for the individual characteristics, representing input, that enhance one’s probability of finding employment, whereas perceived employability reflects outcomes related to that probability (De Cuyper et al., 2012b). Like dispositional employability, perceived employability has been portrayed as a personal resource: it is an individual asset that promotes feelings of control in the work domain (Philippaers, De Cuyper, & Forrier, 2016). Since COR theory posits that individuals with greater resources are more capable of gaining additional resources (Hobfoll, 2011), higher level of dispositional employability might foster one’s perceived employability. In addition, previous research have outlined that perceived employability results from both personal (e.g., knowledge, abilities, attitudes) and situational (e.g., situation on a labour market) factors (Berntson, Sverke, & Marklund, 2006; Kirves, Kinnunen, & De Cuyper, 2014). Accounting for these arguments, it follows that the perception of one’s chances of finding a new job (i.e., perceived employability) should depend on one’s personal characteristics, such as dispositional employability.
This notion has already been portrayed in a process model developed by Forrier, Sels, & Stynen (2009) and adapted by Vanhercke et al. (2014), but has thus far gained only limited empirical verification (for an exception see Forrier, Verbruggen, & De Cuyper, 2015). For example, Kirves, Kinnunen, & De Cuyper (2014) confirmed the positive relationship between optimism and perceived employability among temporary and permanent employees. Wittekind, Raeder and Grote (2010) demonstrated that a willingness to change jobs significantly predicted perception of one’s employment possibilities. Maslić Seršić and Tomas (2015) found that dispositional employability related positively to perceived employability above and beyond the conceptually similar construct of core self-evaluations. However, while these studies were conducted in an employment setting, the present study aims to expand previous knowledge by testing the following hypothesis among unemployed people:
Hypothesis 2: Dispositional employability relates positively to perceived employability.
Perceived employability and job search intensity
Because perceived employability refers to the subjective interpretation of one’s own chances of finding a job, it should directly influence engagement in job search activities: an individual’s behaviour is driven by that individual’s perception of reality, rather than by reality itself (Lazarus & Folkman, 1984). Accordingly, unemployed persons who believe they have a high chance of finding a job are likely to act upon this perception and engage in more job search behaviours. This notion is also consistent with the COR theory: individuals with greater resources (i.e., perceived employability) are more capable to invest in acquiring new resources (i.e., new employment) by engaging job search behaviours (Hobfoll, 1989). Taking into account these theoretical arguments, we argue that perceived employability might act as a crucial determinant in one’s motivation to engage in the self-regulated and goal-oriented behaviours that a job search represents.
Studies examining the effect of perceived employability on job search behaviours among unemployed persons are few but do provide initial support for this line of thinking. For example, Onyishy (2015) found that perceived employability positively predicted preparatory (e.g., talking to friends or relatives about possible job leads), but not active (e.g., sending resumes to potential employers), job search behaviours among graduate students. In a similar manner, Chen and Lim (2012) demonstrated significant positive relationships between perceived employability and both preparatory and active job search behaviours. However, one could argue that the relationship between perceived employability and job search intensity is not linear: perceiving a high chance of finding a job may stimulate job search behaviours up to a certain point, after which high levels of perceived employability may imply that one does not need to engage in a job search since she/he will find a job anyway. From the perspective of control theories, it is plausible that the relationship may even be negative because individuals should engage in job search only if necessary (Powers, 1991). Nevertheless, in this paper, we align our hypothesis with the foremost argumentation dominating the employability literature, which is consistent with the existing empirical results presented above. As such, we set the third hypothesis as follows:
Hypothesis 3: Perceived employability relates positively to job search intensity.
Perceived employability as a mediator in the relationship between dispositional employability and job search intensity
All together, these assumptions form the basis for a mediational model that specifies perceived employability as the mechanism that explains the relationship between dispositional employability and job search behaviour. This assumption was inspired by recent employability models that differentiate between distal (i.e., personal strengths) and proximal (e.g., perceived employability) determinants of behaviour on the labour market (cf. Forrier, Sels, & Stynen, 2009; Forrier, Verbruggen, & De Cuyper, 2015). To date, however, these models have gained only limited empirical verification in an unemployment setting. In this study, we postulate that the subjective interpretation of one’s employment chances is what transmits the effect of personal characteristics on the behaviour of unemployed persons: dispositional employability nurtures perceived employability, which in turn predicts engagement in job search activities. Although none of the aforementioned studies, to our knowledge, have tested this mediational hypothesis, there are findings that have demonstrated the indirect effect of personal characteristics on job search behaviour via one’s self-evaluations. For example, Brown et al. (2006) found that the positive effect of proactive personality on job search behaviour was fully mediated by job search self-efficacy. Onyishy (2015) demonstrated a full mediation between core self-evaluations and preparatory job search explained by perceived employability. In contrast, McArdle et al. (2007) found a negative indirect effect of dispositional employability on job search intensity via self-esteem. However, taking into consideration the dominant empirical findings and the conceptual difference between self-esteem and perceived employability, we estimate a positive indirect effect and full mediation to be plausible and hypothesize that:
Hypothesis 4: The positive relationship between dispositional employability and job search intensity is fully mediated by perceived employability.
Participants and procedure
Data was collected in the spring of 2013 at the central regional branch office of the Croatian Employment Service in Zagreb. Individuals coming to the office for obligatory monthly reporting were randomly approached by trained researchers. Only unemployed persons with existing work experience were taken into consideration because the core study variables (i.e., dispositional and perceived employability) were anchored in previous work contexts and related to one’s experience while being employed. Researchers stressed the anonymity and voluntary nature of participation, as well as the importance of the study for all parties involved. Unemployed persons who agreed to participate in the study (N = 567) completed a paper-and-pencil survey at desks prepared for this purpose. After excluding nine participants for whom completed surveys contained missing values for more than 50% of the items on any scale (approximately 1.6% of the total sample) and 25 participants with missing data on the control variables (approximately 4.4% of the total sample), the effective sample size was 533 participants.2
Sample characteristics are presented in Table 1. In comparison with the general population of unemployed registered persons in Croatia (with the existing work experience) (The Croatian Employment Service, 2015b), the sample consisted of approximately equal proportion of male and female participants. In contrast, younger and more educated participants were overrepresented in the sample. Since older and less educated persons are more inclined to refuse to participate in research, the bias towards younger and more educated participants is a common characteristic of the convenience samples (e.g., Šverko et al., 2008). In addition, the sample consisted of the higher proportion of persons who were unemployed for less than one year, compared with the general population.
|Sample||Population of registered unemployed persons||Difference|
|N (%)||533 (100)||275,838 (100)|
|Gender, N (%)|
|Male||232 (43.53)||130,714 (47.39)||χ2 (1, N = 533) = 3.18, p > 0.05|
|Female||301 (56.47)||145,124 (52.61)|
|Age, N (%)|
|<35 years||339 (63.60)||99,884 (36.21)||χ2 (2, N = 533) = 183.19, p < 0.001|
|35–49 years||132 (24.77)||92,658 (33.59)|
|>49 years||62 (11.63)||83,296 (30.20)|
|Education, N (%)|
|Primary school or less||18 (3.38)||72,491 (26.28)||χ2 (2, N = 533) = 21.53 p < 0.001|
|High school||281 (52.72)||175,656 (63.68)|
|University education||234 (43.90)||27,691 (10.04)|
|Length of unemployment, N (%)|
|1 year||327 (61.35)||144,568 (52.41)||χ2 (2, N = 533) = 706.56, p < 0.001|
|1–3 years||139 (26.08)||78,028 (28.29)|
|>3 years||67 (12.57)||53,242 (19.30)|
All measures were translated into the Croatian language. Participants indicated their degree of agreement with a set of items using one of two response formats: (1) a five-point scale ranging from one (totally disagree) to five (totally agree) (dispositional employability and perceived employability) and (2) a four-point scale ranging from one (not at all) to four (every day) (job search intensity).
Dispositional employability. Dispositional employability was assessed with the 25-item dispositional measure of employability developed by Fugate and Kinicki (2008). This scale measures five first-order dimensions in which dispositional employability is reflected in: openness to changes at work (five items, e.g., ‘I feel that I am generally accepting of changes at work’), work and career resilience (eight items, e.g., ‘I am optimistic about my future career opportunities.’), work and career proactivity (three items, e.g., ‘I stay abreast of developments in my industry.’), career motivation (three items, e.g., ‘I have a specific plan for achieving my career goals.’) and work identity (five items, e.g., ‘The type of work I do is important to me.’). The items were transformed to past tense where necessary in order to make them applicable in an unemployment setting (e.g., ‘I stayed abreast of developments in my company.’). One item was dropped from the analyses because it loaded poorly on the corresponding latent factor in the preliminary confirmatory factor analysis (CFA) (i.e., ‘I define myself by the work that I do.’, intended to measure work identity [λ = 0.37]). The reliability assessed with the Cronbach’s alpha coefficient was 0.89.
Perceived employability. Perceived employability was measured with the seven-item self-perceived employability scale (Rothwell & Arnold, 2007). Some items were slightly adapted to fit the unemployment context. Again, three items were dropped from the analyses based on the results of the preliminary CFA: one item loaded poorly on the corresponding latent factor (i.e., ‘I have a good knowledge of the opportunities on the labour market even if they are quite different to what I have done so far.’ [λ = 0.20]) and two items had high cross-loadings with work and career resilience (i.e., ‘The skills I have gained in my previous job(s) are transferable to other occupations.’ [λ = 0.47] and ‘I could easily retrain to make myself more employable.’ [λ = 0.43]). The four items that were retained for the analyses were: ‘I could easily get another job similar to my previous one.’, ‘Anyone with my level of skills, knowledge and similar work experience is highly sought after by employers.’, ‘I could easily get a similar job to my previous one in almost any organisation.’ and ‘I could get any job, anywhere, so long as my skills and experience were reasonably relevant.’ The reliability of the scale was α = 0.77.
Job search intensity. Job search intensity was measured with the six-item scale developed by Šverko et al. (2008). The items were derived from the job search literature (e.g., Schwab, Rynes, & Aldag, 1987; Wanberg, Kanfer, & Rotundo, 1999) and interviews with unemployed persons. The measure assesses the frequency of engagement in different job search activities during the last month, where job search activities include: reading and searching through paper advertisements, checking job offers from the Employment Service, searching through advertisements on the Internet, asking friends and acquaintances if they know of any job offers, personally calling or visiting employers, seeking out influential people and informal connections. A time-span of one month, as used and validated by the authors of the scale, was considered long enough to capture the representative frequency of engagement in various job search activities among the unemployed, yet short enough to enable its valid estimation (Šverko et al., 2008). The indicated time-span is often selected in W/O psychology as it is considered appropriate to capture relevant behaviours, cognitions and emotions in the work domain (cf. Griffin, Neal, & Parker, 2007; Van Katwyk et al., 2000). The Cronbach’s alpha of this measure was 0.72.
Control variables. We controlled for several sample characteristics that have been found to co-vary with the variables examined in this study (e.g., Berntson & Näswall, 2010; Šverko et al., 2008; Wanberg, Hough, & Song, 2002): gender (0 = male, 1 = female), age (years), education (0 = up to high school, 1 = university) and length of unemployment (1 = up to one month, 2 = one to six months, 3 = six to twelve months, 4 = one to three years, 5 = more than three years).
The hypotheses were tested using Structural Equation Modelling (SEM) in the AMOS 22.0 statistical package (Arbuckle, 2013), with the covariance matrix serving as the input for the analyses. Preliminary data screening did not reveal any violations of normality (skewness indices ranged from 0.11 to 1.16; kurtosis indices ranged from 0.05 to 2.33) nor multi-collinearity (none of the correlations exceeded 0.85) (Kline, 2005; Weston & Gore, 2006). Therefore, we used the maximum likelihood procedure to estimate the goodness-of-model-fit and model parameters. In order to increase the model parsimony and the robustness of the analyses, we parcelled scale items using the single-factor method (Landis, Beal, & Tesluk, 2000). Specifically, items for openness to changes at work, work and career resilience, work identity, perceived employability and job search intensity were subjected to an exploratory factor analysis with a specified one-factor solution. We created composites by pairing the items with the highest and lowest factor loadings until all items were assigned to a corresponding parcel. Items for work and career proactivity and career motivation were not parcelled as there were only three of these items.
Next, we followed the two-steps procedure recommended by Anderson and Grebing (1988). Accordingly, the first step included a comparison of a series of nested measurement models. First, we compared a hypothesized three-factor model (dispositional employability as a second-order factor comprising five first-order factors, perceived employability and job search intensity; M1) with two alternative models: a two-factor model in which all parcels/items intended to measure dispositional employability and perceived employability were specified to load onto one factor and job search intensity as a second factor (M2) and a one-factor model in which all parcels/items were specified to load onto a single factor (M3). In order to additionally test the discriminant validity of the scales used to measure dispositional employability and perceived employability, we conducted CFA with five dimensions of dispositional employability specified as separate factors, as suggested by an anonymous reviewer. Accordingly, we compared a seven-factor model (openness to changes at work, work and career resilience, work and career proactivity, career motivation, work identity, perceived employability and job search intensity; M4) with five alternative six-factor models, in which indicators measuring one of the first-order factors from dispositional employability and perceived employability were specified to load onto one factor (M5–M9).
The second step included the specification of the structural model. The significance of the total, direct and indirect effect was tested using the bootstrap method. We performed this analysis with 10,000 resamples and constructed bias-corrected (BC) confidence intervals (CI). This method uses the observed data to generate a sampling distribution, which is then used to derive asymmetric confidence intervals for the tested effects (MacKinnon, Fairchild, & Fritz, 2007). If the confidence interval of the corresponding effect does not contain zero, the effect is statistically significant (Preacher & Hayes, 2008). Compared to confidence intervals based on a normal distribution, the bootstrap method has an important advantage: it does not impose the normality assumptions of the sampling distribution of the indirect effect. As a result, it has higher statistical power and more accurate type-I error (MacKinnon, Fairchild, & Fritz, 2007; Preacher & Hayes, 2008).
We evaluated the overall goodness-of-model-fit with a combination of absolute (standardized root mean square residual, SRMR) and incremental (comparative fit index, CFI) fit indices and the index corrected for model parsimony (the root mean square error of approximation (RMSEA) with the corresponding 90% confidence interval). An acceptable fit between the hypothesized model and the observed data is indicated when: values of SRMR and RMSEA are close to or below .08 and the value of CFI equals or exceeds 0.90 (Bentler, 1990; Browne & Cudek, 1993; Hu & Bentler, 1999). We used the χ2-difference test to statistically compare the fit of the nested models.
In testing the hypotheses, we controlled for gender, age, education and length of unemployment. First, each dependent variable was regressed on the control variables. Only significant predictors were included in the structural model: gender was significantly associated with perceived employability and job search intensity while age was significantly associated with job search intensity. This choice was made due to the model parsimony.
Table 2 presents means, standard deviations, reliabilities and correlations between all variables examined in the present study. It is noteworthy that zero-order correlations between the core variables followed the hypothesized pattern: dispositional employability, perceived employability and job search intensity were positively associated.
|4. Length of unemployment||–||–||–0.02||0.26**||–0.16**||–|
|5. Dispositional employability||3.90||0.53||0.10*||–0.06||0.09*||–0.06||(0.89)|
|6. Perceived employability||3.47||0.76||0.11**||–0.13**||0.01||–0.09*||0.54**||(0.77)|
|7. Job search intensity||2.13||0.59||0.16**||–0.18**||0.03||–0.04||0.17**||0.24**||(0.72)|
The CFA results of the three nested measurement models are presented in the upper part of Table 3. The hypothesized three-factor measurement model (M1) fitted the data reasonably well and significantly better than the alternative two- and one-factor models (M2 and M3, respectively). Other fit indices were consistent with this conclusion: SRMR, CFI and RMSEA values indicated that M2 and M3 exhibited an inadequate fit to the data. All indicators in the three-factor model (M1) were significantly and positively related to the corresponding latent factor (standardized regression weights ranged from 0.53 to 0.98). In addition, none of the correlations between the factors exceeded a value of 0.85, which substantiates the claim about the discriminant validity of the measures (Brown, 2006).
|Model||χ2||df||SRMR||CFI||RMSEA||RMSEA 90% confidence interval||Model comparison||Δχ2||Δdf|
1. three-factor model (M1)
(DE, PE, JSI)
|2. two-factor model (M2)
(DE + PE, JSI)
|1122.55***||151||0.074||0.74||0.110||0.104–0.116||M2 vs. M1||673.74***||7|
|3. one-factor model (M3)
(DE + PE + JSI)
|1502.69***||152||0.095||0.63||0.129||0.123–0.135||M3 vs. M1||1053.88***||8|
4. seven-factor model (M4)
(PE, RES, OC, PRO, CM, WI, JSI)
|5. six-factor model (M5)
(PE + RES, OC, PRO, CM, WI, JSI)
|582.23***||137||0.056||0.88||0.078||0.072–0.085||M5 vs. M4||162.57***||6|
|6. six-factor model (M6)
(PE + OC, RES, PRO, CM, WI, JSI)
|703.61***||137||0.057||0.85||0.088||0.082–0.095||M6 vs. M4||283.95***||6|
|7. six-factor model (M7)
(PE + WI, OC, RES, PRO, CM, JSI)
|591.82***||137||0.055||0.88||0.079||0.073–0.086||M7 vs. M4||172.16***||6|
|8. six-factor model (M8)
(PE + CM, OC, RES, PRO, WI, JSI)
|558.36***||137||0.052||0.89||0.076||0.070–0.083||M8 vs. M4||138.70***||6|
|9. six-factor model (M9)
(PE + PRO, OC, RES, CM, WI, JSI)
|599.15***||137||0.052||0.87||0.080||0.073–0.086||M9 vs. M4||179.49***||6|
|10. DE → PE → JSI||537.03***||178||0.051||0.90||0.062||0.056–0.068|
In addition, the results of the χ2-difference test showed that the seven-factor model (five separate dimensions of dispositional employability, perceived employability and job search intensity; M4) fitted the data significantly better than each of the alternative six-factor models (M5–M9) (see the middle of part of Table 3). Together, these results provide additional empirical evidence for the discriminant validity of the dispositional and perceived employability measures.
The fit indices for the structural model are presented in the lower part of Table 3. The model yielded a reasonable fit to the data. The tested effects are depicted in Figure 1. Unstandardized and standardized effects with the BC 95% CIs of the standardized effects are presented in Table 4. In line with Hypothesis 1, the total effect of dispositional employability on job search intensity was positive (β = 0.19, CI = 0.10–0.28). Furthermore, dispositional employability had a positive direct effect on perceived employability (β = 0.70, CI = 0.61–0.78), whereas perceived employability had a positive direct effect on job search intensity (β = 0.27, CI = 0.15–0.38). These findings are in line with Hypotheses 2 and 3, respectively. Finally, we found that dispositional employability was significantly related to job search intensity through perceived employability (indirect β = 0.19, CI = 0.10–0.28), as predicted in Hypothesis 4. The non-significant direct effect of dispositional employability on job search intensity leads us to conclude that perceived employability fully mediated the relationship between dispositional employability and job search intensity.
|Bootstrapping BC 95% CI|
|Unstandardized estimates||Standardized estimates||Lower||Upper|
|DE → JSI||0.21***||0.19||0.10||0.28|
|DE → PE||0.53***||0.70||0.61||0.78|
|PE → JSI||0.39***||0.27||0.15||0.38|
|DE → JSI||0.00n.s.||0.00||0.00||0.00|
|DE → PE → JSI||0.21***||0.19||0.10||0.28|
In regards to the control variables, gender did not significantly predict perceived employability (β = 0.06, CI = –0.03–0.14). However, it did significantly predict job search intensity (β = 0.15, CI = 0.06–0.24), implying that women engaged in more job search behaviours than men. Age significantly predicted job search intensity (β = –0.16, CI = –0.25–[–0.06]), with younger participants performing more job search behaviours. Moreover, dispositional employability was not significantly related to gender (r = 0.10, CI = 0.00–0.20), but was related significantly to age (r = –0.11, CI = –0.21–[–0.01]). In this case, younger participants reported higher levels of dispositional employability.
This study is one of a few that examined the role of dispositional employability in the engagement in job search activities among unemployed persons. In contrast to the majority of existing studies examining the determinants of job search behaviour, the focus in the present study was placed on personal characteristics that promote proactive adaptability in the work and career context (Fugate, Kinicki, & Ashforth, 2004). Accordingly, we tested a hypothesized model that specifies dispositional employability as a predictor of job search intensity via perceived employability. By doing so, we integrated two streams of research: that examining job search behaviour, where the importance of employability among the unemployed has only recently been recognized, and employability research, which has been primarily conducted within an organizational context.
Our results demonstrated that unemployed persons with higher levels of dispositional employability were more inclined to search for a new job, as outlined in Hypothesis 1. Consistent findings were also found in the Netherlands (Koen, Klehe, & Van Vianen, 2013) and Australia (McArdle et al., 2007). In contrast to these studies, the present study was conducted in a labour market characterized by high unemployment and a low number of job offers, thus enabling the generalization of previous findings to a different contextual setting. However, the effect of dispositional employability on job search intensity was weaker in the present study as compared to those obtained in the two previous studies. One plausible explanation for this difference might relate to different conceptualizations of employability. For example, in his operationalization of employability, McArdle et al. (2007) included social capital in addition to a proactive personality and boundaryless mindset (Fugate, Kinicki, & Ashforth, 2004), while our study aligned with the most recent conceptualization that exclusively reflects personal characteristics in terms of proactive adaptability (Fugate & Kinicki, 2008). Nevertheless, we believe that the significant and positive effects of dispositional employability on job search behaviour obtained in different contexts and with different scales substantiate the importance of individual characteristics that foster proactive adaptability among unemployed persons.
Furthermore, we aimed to investigate the explanatory mechanism that might account for the positive relationship between dispositional employability and job search intensity. Conditional upon the mediational effect, dispositional employability related positively to perceived employability (Hypothesis 2). This finding is consistent with studies that have demonstrated the positive effect of one’s movement capital on the perception of employability (e.g., Forrier, Verbruggen, & De Cuyper, 2015; Maslić Seršić, & Tomas, 2015). Although compatible with Fugate, Kinicki, & Ashforth’s (2004) notion of employability, these studies used a slightly different operationalization of movement capital. For example, Forrier, Verbruggen, & De Cuyper (2015) defined it in terms of human and social capital, adaptability and self-awareness. In addition, both studies were conducted among (mainly) employed individuals.
The present study also demonstrated a positive relationship between perceived employability and job search intensity, supporting the notion that the subjective interpretation of many employment chances relates to the higher intensity of engagement in job search behaviours (Hypothesis 3). A similar, but slightly higher effect was obtained with job search self-efficacy (rc = 0.27) in meta-analysis (Kanfer, Wanberg, & Kantrowitz, 2001). Job search self-efficacy reflects beliefs in one’s capabilities to successfully perform various job search activities and obtain employment. Although similar, perceived employability should be differentiated from job search self-efficacy for two reasons. First, it accounts for contextual factors in addition to one’s personal capabilities. Second, it might be reasonable to assume that perceiving very high chances of obtaining new employment (i.e., a high level of perceived employability) might non-significantly, or even negatively, relate to engagement in job search behaviours (i.e., one simply doesn’t need to search for a job in order to find it). Acknowledging nonlinearity in the relationships between the study variables is gaining increased attention in recent research (e.g., Le et al, 2011). As such, it might present a promising route for future research examining perceived employability and job search behaviours.
Finally, the statistically significant indirect effect of dispositional employability on job search intensity through perceived employability supported Hypothesis 4. The full mediation obtained in the present study places clear emphasis on the importance of one’s subjective interpretation of employment possibilities: people with higher levels of dispositional employability perceived greater chances of finding a job and in turn acted upon that perception by engaging in more job search behaviours. This is consistent with the findings of previous studies demonstrating that subjective evaluations of one’s employment chances fully mediated the effects of personal characteristics on job search behaviour (e.g. Chen & Lim, 2012; Onyishi et al., 2015).
Limitations and future research
The results of this study should be interpreted in light of several limitations. First, the cross-sectional research design limits the possibility of drawing solid causal conclusions. However, the direction of the relationships implied by our model was specified in line with dominant employability models (e.g., Forrier, Sels, & Stynen, 2009; Vanhercke et al., 2014). Moreover, it is consistent with existing results on the causal relationships between personal characteristics and job search behaviour found in longitudinal studies (e.g., Koen, Klehe, & Van Vianen, 2013). Nevertheless, it would be interesting to probe the reverse and reciprocal relationships between dispositional employability and job search intensity using a longitudinal research design. Specifically, although tentative, a reversed hypothesis might refer to the situation in which intensive engagement in job search behaviours enhances one’s personal resources (e.g., by exposure to different people and experiences). For example, Forrier, Verbruggen, & De Cuyper (2015) found that transitions across organizations increased person’s movement capital. However, this assumption would imply that there must be sufficient mobility on the labour market (e.g., job offers) that would enable unemployed persons to accumulate different positive experiences and enhance resources. This mobility is not a characteristic of the current Croatian labour market.
The second limitation of the present study stems from the fact that our data were based on self-report measures, which could result in inflated relationships between the study variables due to the common method variance. This methodological artefact is considered to be one of the main sources of systematic measurement error (Bagozzi & Yi, 1990). To alleviate this possibility, we followed the instructions proposed by Podsakoff et al. (2003). Specifically, we emphasized to participants that there were no right or wrong answers, stressed the anonymity of their responses and used different scale anchors for predictor and criterion variables. These arguments and the ongoing debate on the magnitude of inflation caused by the common method variance (Podsakoff et al., 2003) lead us to conclude that the main findings in this study were not affected by this potential limitation to any substantial extent. Nonetheless, future studies might benefit from the collection of data from other sources, that is, persons who have sufficient access to relevant information on the dispositional employability characteristics and job search behaviour of unemployed persons (e.g., partners, closest friends etc.).
Third, a potential concern might arise from the relatively high correlation between the dispositional and perceived employability. To test the discriminant validity of the scales used to measure both constructs, we compared a set of alternative measurement models. The results supported the notion that dispositional and perceived employability are related yet distinct constructs, the former referring to the constellation of individual characteristics that promote one’s employment possibilities and the later referring to the perceived probability of finding a new employment.
Fourth, the amount of explained variance in job search behaviour is rather small (13%). Accordingly, it might be useful to include additional employability characteristics (e.g., human and social capital, as specified in the first version of the model of dispositional employability) that could explain variations in job search intensity (Fugate, Kinicki, & Ashforth, 2004).
Fifth, we believe that future studies would benefit from the inclusion of additional measures that were not examined in this study. One possibility is the consideration of job search quality in addition to intensity, a criterion variable that could also be affected by dispositional employability. For example, proactive individuals might be more selective in performing job search behaviours by adopting more targeted and refined strategies. A second possibility for further consideration refers to reemployment success and quality. Although existing studies indicate that job search intensity predicts both of these outcomes, the results are not indisputable. For example, in a study closest to our own context, Šverko et al. (2008) found that job seeking intensity did not predict employment status in a follow-up. Therefore, it would be beneficial to extend the present findings by investigating the reemployment status of participants in a follow-up study.
Finally, it is necessary to address some issues related to the sample in this study. The first concern relates to the potential range restriction in dispositional employability, which might have diminished correlations between the study variables. It is reasonable to assume that persons with higher levels of dispositional employability were more inclined to participate in the study. Although we cannot entirely exclude this possibility, we attempted to minimize it by explaining the potential implications of the study for unemployed persons and thus providing motivation to all potential participants. The second concern in relation to our sample relates to the representativeness of the sample, which was comprised of a higher proportion of younger and more educated persons, as well as of a higher proportion of persons with shorter unemployment length, compared to the general population of unemployed persons in Croatia. Accordingly, it would be beneficial to replicate our results with representative samples as well as with a number of specific subpopulations of unemployed persons (e.g., long-term and low-skilled unemployed persons). It would also be interesting to compare the hypothesized model in different subpopulations (e.g. short- and long-term unemployed). Third, because it is reasonable to assume that dispositional employability might also boost job search behaviour among employed persons, future studies might aim to replicate the findings of the present study with this population as well.
The results of the present study have several implications for practitioners in the field of career counselling. In the first instance, programmes designed to boost one’s dispositional employability could complement the existing interventions aiming to increase job search activities, well-being and reemployment of the unemployed. Currently, most of the available labour market programmes are directed towards the improvement of occupational skills and/or problem-solving skills, self-efficacy, communication and networking skills (e.g., Brenninkmeijer & Blonk, 2011; Creed, Hicks, & Machin, 1998; Eden & Aviram, 1993). Our results suggest that activities designed to teach unemployed persons how to be more proactively adaptable will enhance their engagement in job search activities. Second, characteristics encompassed in dispositional employability may promote problem-focused coping not only among the unemployed, but also among new entrants to job markets and persons in periods of career transition. In this regard, educational institutions and organisations may offer individual career planning that is oriented towards enhancing one’s characteristics that promote proactive adaptability. Third, the finding that employability relates positively to job search behaviours in a context of high unemployment rates and a small number of job offers highlights the importance of individual resources that enable unemployed persons to remain resilient in the face of challenges and setbacks in an unfavourable labour market.
The characteristics encompassed by dispositional employability do seem to matter for the unemployed persons. The results of this study suggest that dispositional employability plays a role in job seeking behaviours via the subjective interpretation of one’s employment possibilities. Accordingly, we believe this study contributes to theoretical knowledge on the personal characteristics that are relevant for job search behaviour. From the practitioners’ point of view, investing in dispositional employability of unemployed persons might contribute to their well-being by enabling an individual to remain active and aware of labour market opportunities and preventing social isolation.