Interactive Effects of Approach and Avoidance Job Crafting in Explaining Weekly Variations in Work Performance and Employability

textabstractMeta-analyses on job crafting reveal that while approach-oriented job crafting (e.g., increasing job resources or challenging job demands) relates positively to employee performance, avoidance-oriented job crafting (e.g., decreasing hindering job demands) has either non-significant or negative implications for employee functioning. However, the joint effects of approach and avoidance job crafting remain an underdeveloped area of research. We administered a three-week diary survey among 87 employees to test interaction effects of approach and avoidance job crafting on employee (other-referenced and past-referenced) work performance and employability. Results revealed that decreasing hindering job dema


INTRODUCTION
Recent meta-analyses systematized the consequences of job crafting for employees and organizations (Lichtenthaler & Fischbach, 2019;Rudolph et al., 2017). A recurring pattern is that when employees display approach-oriented © 2020 The Authors. Applied Psychology published by John Wiley & Sons Ltd on behalf of International Association of Applied Psychology. job crafting (i.e. expand their job scope by increasing their resources or challenging demands) they attain high work achievements, while when they engage in avoidance-oriented crafting (i.e. narrow their job scope by decreasing their demands) they are less efficient (Zhang & Parker, 2019). This is because avoidance job crafting strategies cost resources (i.e. energy, time), while they are likely to be unsuccessful because job demands are given and cannot change easily (Lichtenthaler & Fischbach, 2019). While this stream of literature enables us to disentangle the favorable from the unfavorable job crafting strategies for employees, it neglects one possibility. Namely, that employees engage in both approach and avoidance job crafting simultaneously (Mäkikangas, 2018), which suggests that those who make their work resourceful and meaningful via approach crafting, may buffer the unfavorable consequences of avoidance-oriented job crafting.
Just like Mäkikangas (2018), we argue that job crafting behaviors are not mutually exclusive and can be deployed at the same time. To support this claim, Mäkikangas applied latent profile analysis to identify unique job crafting profiles based on different combinations of approach and avoidance job crafting strategies, and to explore how these profiles relate to work engagement. Extending this line of research, in the present study we investigate interactive effects between avoidance job crafting (i.e. decreasing job demands) and specific approach crafting strategies (i.e. increasing social and structural resources, and challenging job demands). In this way, we contribute to the literature by examining which specific approach job crafting strategies may buffer the unfavorable impact of avoidance job crafting, and whether all approach job crafting strategies are equally effective toward that end.
Considering that resources and self-regulation capacities are finite (Beal et al., 2005;Hobfoll & Shirom, 2000), it is relevant to understand whether certain approach-oriented job crafting strategies are more effective in mitigating the unfavorable consequences of avoidance crafting. By treating the different approach job crafting strategies separately, we contribute to the literature by acknowledging that increasing resources and increasing challenging demands may refill one's resource reservoir differently. Namely, empirical evidence (Demerouti et al., 2015;Tims et al., 2013;van Wingerden, Bakker, & Derks, 2017) implies that the strategy of increasing resources is more likely to enrich the work environment, while that of increasing challenges is more likely to refill employees' personal resources (e.g., self-efficacy). Although different strategies may refill different types of resources, they are expected to buffer the unfavorable impact of avoidance-oriented strategies through the same mechanism. This is because both job (Xanthopoulou et al., 2007) and personal (Xanthopoulou et al., 2013) resources have been found to mitigate demanding situations at work in the same way.
We test interaction effects between avoidance and different approach job crafting strategies to explain weekly work performance and employability orientation (i.e. defined as the willingness to actively increase one's employability as a means of increasing the flexibility of one's organization; for example, being prepared to change one's work activities if the organization expects an employee to perform different tasks; Van Dam, 2004). Work performance and employability are particularly relevant for our scope. Today's dynamic work environments require employees to constantly reinvent their work roles and proactively challenge themselves if they want to perform well or remain employable. As such, avoidance behaviors are largely incompatible with the contemporary image of the successful employee. However, due to the complex demands and threats of these dynamic environments, in practice, employees often have to display avoidance behaviors (e.g., protect themselves from stress and workload; Ybema et al., 2020). We, thus, find it relevant both for practitioners and researchers to examine when the negative effects of avoidance crafting become less pronounced. By testing approach-avoidance interactions with these specific outcomes, our study advances the literature on job crafting and work performance that has revealed inconclusive findings regarding the role of decreasing demands (Lee & Lee, 2018), and refines the previously examined links between job crafting and employability (Akkermans & Tims, 2017) by addressing interactive rather than main effects of job crafting strategies.
We conducted a weekly diary study to achieve our study aims. This agrees with literature highlighting the effects of simultaneous job crafting strategies as within-person phenomena (Mäkikangas, 2018). Different work weeks pose different demands (Bakker & Sanz-Vergel, 2013). It is in the most demanding work weeks that employees need to combine different job crafting strategies; that is, not simply to avoid tasks (i.e. decreasing demands) but to do so in a way that is constructive and compensates for the potential shortcomings of avoidance behaviors (e.g., by crafting challenges and resources). Because our study uses self-reports, and in order to help respondents use less subjective frames of references , work performance was assessed in comparison to (a) others (other-reference), and (b) to one's own past (past-reference).

Interactive Effects of approach-and avoidance-Oriented Job Crafting
Based on Conservation of Resources (COR) theory (Hobfoll & Shirom, 2000), we argue that avoidance-oriented job crafting is unfavorable for work performance (Lichtenthaler & Fischbach, 2019) and employability because attempts to decrease demands may cost energetic, cognitive and emotional resources. Since job demands are usually given in a certain work environment, trying to reduce them may lead employees to realize that these efforts are inefficient or simply impossible (Demerouti & Peeters, 2018) and, thus, any resources invested toward that end are likely to be depleted with no actual gain. For instance, trying to make your work less mentally effortful may require investing cognitive resources for finding ways to remove obstacles (e.g., to process less information or have less workload) with highly uncertain outcomes. In this way, cognitive resources are no longer available to be invested in work tasks or in adapting to flexible conditions, thus impairing performance and employability, respectively. A way to compensate for this resource loss is by actively attempting to regain one's resource reservoir through the enhancement of social and structural job resources and challenges (i.e. approach crafting), at the same time. Hence, when efforts to reduce demands are combined with higher (vs. lower) efforts to gain job resources (that are likely to enrich the work environment) or challenges (that are likely to enhance feelings of meaningfulness), the unfavorable effects of avoidance job crafting (i.e. decreasing demands) will be mitigated because lost resources are compensated by approach crafting attempts.
Going one step further, when reducing demands is displayed on its own, it is more likely to indicate an overarching avoidance motivation or withdrawal behavioural pattern (Petrou et al., 2012). When, however, it is displayed together with approach crafting, it may hold a different meaning and have a less harmful impact. In other words, simultaneous avoidance and approach crafting ensures that the approach compensates for the harmful aspects of avoidance and that the avoidance is realized in a way that is less destructive and less cynical. As such, approach and avoidance do not exclude each other; rather, they help each other in preventing decrements in functioning (Nikitin & Freund, 2010).
Our proposition is, thus, two-fold: Reducing demands on its own deprives employees from the stimulating changes of their job and harms their performance (Rudolph et al., 2017). However, when applied together with approach crafting, it is less likely to hinder performance because of the gained resources and challenges. To illustrate, an employee, who tries to eliminate intrusions from difficult clients (i.e. decreasing hindering demands) but, at the same time, pursues training from older colleagues (i.e. increasing social job resources) or recruits new clients (i.e. increasing challenges) is less likely to exhibit performance impairments. A similar strategy should benefit one's employability. Flexibility or adjustment to new situations (cf. employability orientation) does not require only a growth mindset but also one that focuses on foreseeing and overcoming obstacles. Such a dual mindset allows employees to be in control and remain adjustable throughout different tasks and work environments (Miron-Spektor & Beenen, 2015). In other words, decreasing demands on its own impedes adjustment to change because it removes challenge from new situations (Petrou et al., 2012). However, when displayed by employees with a positive mindset, this unfavourable effect is offset (Demerouti et al., 2017). To illustrate, consider an employee who decides to answer e-mails less frequently (i.e. decreasing hindering demands) but, at the same time, they autonomously find and pursue a course on a software that will help them deal with their newly assigned tasks (i.e. increasing structural job resources) or they develop a new and innovative idea for a new product that they had (i.e. increasing challenges), and, thus, they do not experience decrements in their employability.
Hypothesis 1: The negative relationships between avoidance-oriented job crafting, on the one hand, and weekly other-referenced work performance (1a), pastreferenced work performance (1b), and employability orientation (1c), on the other hand, will be weaker when approach-oriented job crafting is higher (vs. lower).

Sample and Procedure
Participants were recruited via research assistant network sampling (Demerouti & Rispens, 2014). Employees from different Greek organizations were invited to participate via three emails containing links to three online surveys that they had to complete at the end of three consecutive working weeks. Data were matched via an anonymized personalized link that was identical in all three questionnaires. Out of 283 invited respondents, 87 (55% women) filled in all three surveys (response rate = 31%) and formed the final dataset used for analyses; no data were missing. Their mean age was 31.1 (SD = 7.2) and they worked an average of 42.4 hours (SD = 8.5) per week. The most representative occupational sectors were commerce (12%), government (12%), business (11%), education (10%), health (10%), and finance (7%).

Instruments
Instruments were based on validated scales, adjusted to refer to the past working week. Unless indicated otherwise, answering scales ranged from 1 = never to 5 = always.
Job crafting was measured with the questionnaire by Tims et al. (2012), including five items for increasing structural job resources (e.g., "In the past week, I tried to develop my capabilities"), five items for increasing social job resources (e.g., "I asked colleagues for advice"), five items for increasing challenging job demands (e.g., "When there was not much to do at work, I saw it Note. Correlations below the diagonal are at the between-level and above the diagonal at the within-level; in the diagonal ranges are shown (low/high) for the alphas of all scales across the study measurements. *p < .05; **p < .01.
as a chance to start new projects") and six items for decreasing hindering job demands (e.g., "I made sure that my work is mentally less intense").
Work performance was measured with the World Health Organization Health and Work Performance Questionnaire (Kessler et al., 2003), including three items for other-referenced work performance (e.g., "How often during the past week, was the amount of work that you completed bigger than that of most colleagues of your position?"), and three items for past-referenced performance (e.g., "How often during the past week, was the quality of work that you conducted better than the respective period of the previous year").
Employability orientation was measured with four items from Van Dam (2004; e.g., "In the past week, I found it important to participate in development activities regularly" or "In the past week, I was prepared to change my work activities, if the organization needed me to perform different tasks"). Answering categories ranged between 1 = totally disagree to 4 = totally agree.

Strategy of analyses
We used MlwiN to conduct multilevel analyses for each dependent variable separately. Intraclass correlations (i.e. variation at the between-level of analysis) were 47 per cent for past-reference performance, 57 per cent for other-reference performance and 51 per cent for employability orientation, revealing that adequate variation is left to be explained by within-level variables. Following previous theorizing and evidence revealing that younger employees and men may experience more employability or career success (e.g., Baltes, 1997;Nielsen, 1999; Van der Heijden et al., 2009), we have controlled for gender and age in the analyses. In each analysis we compared a null model to a series of nested models that comprised successively age and gender as control variables (Model 1), all four job crafting dimensions (Model 2), and all three possible interactions between decreasing hindering job demands and any type of approach-oriented crafting (i.e. increasing structural job resources, increasing social job resources or increasing challenging job demands; in three independent models, Model 3a, 3b and 3c). Following previous practice, all within-level predictor variables were centred to the person-mean. Significant interactions were probed with the simple effect approach and were plotted using +/-1SD of the moderating variables.  Step 3 for different clusters of variables (i.e. a four-factor model for job crafting, a three-factor model for the outcome variables, and a model distinguishing between increasing structural job resources and employability that were found to correlate highly; see Table 1). Results (Online Appendix Table S1) revealed the superiority of our measurement models compared to alternative models. Table 2 summarizes the results of the multilevel analyses. Three interaction effects were significant, namely: decreasing hindering job demands and increasing social job resources on other-referenced performance (see Figure 1a), decreasing hindering demands and increasing structural job resources on past-referenced work performance (see Figure 1b) and decreasing hindering demands and increasing challenging job demands on employability orientation (see Figure 1c). Simple slope tests revealed that the link between decreasing hindering job demands and other-referenced performance was positive and significant when increasing social job resources was higher (estimate = 0.43, S.E. = 0.14, p < .01) and non-significant when increasing social job resources was lower (estimate = −0.10, S.E. = 0.15, p = .95). The link between decreasing hindering job demands and past-referenced performance was positive and significant when increasing structural job resources was higher (estimate = 0.46, S.E. = 0.19, p < .05) and non-significant when increasing structural job resources was lower (estimate = −0.17, S.E. = 0.18, p = .34). Finally, the link between decreasing hindering job demands and employability was negative and significant when increasing challenging demands was lower (estimate = −0.31, S.E. = 0.09, p < .01) but nonsignificant when increasing challenging job demands was higher (estimate = 0.04, S.E. = 0.11, p = .73). These findings provide partial support to Hypotheses 1a, 1b and 1c. Rerunning analyses without gender and age as control variables did not alter the findings.

DISCUSSION
Based on the functions of different crafting strategies (Lichtenthaler & Fischbach, 2019;Rudolph et al., 2017) and the assumptions of COR theory (Hobfoll & Shirom, 2000), we hypothesized that avoidance-oriented job crafting will relate negatively to work performance and employability when approach-oriented job crafting is lower, while this negative effect will be buffered when approach-oriented job crafting is higher. Results partly supported these expectations. In the case of employability, findings showed that approach job crafting strategies buffer the negative relationship between avoidance crafting and employability, since this link was negative only when approach-oriented job crafting was low. However, in the case of work performance, results revealed that when avoidance-oriented job crafting strategies are combined with high levels of approach-oriented job crafting strategies, performance is boosted.

Theoretical and Practical Implications
Recent empirical evidence suggests that resources and challenges interact with hindrance demands in explaining employee functioning (Riedl & Thomas, 2019;Tadić et al., 2015). However, literature has yet to uncover whether crafting job resources or challenging job demands (i.e. approach-oriented fIgURE 1. (a) The link between decreasing hindering job demands and otherreferenced work performance moderated by increasing social job resources; (b) the link between decreasing hindering job demands and past-referenced work performance moderated by increasing structural job resources; (c) the link between decreasing hindering job demands and employability moderated by increasing challenging job demands job crafting) interacts with crafting hindering demands (i.e. avoidanceoriented job crafting) to determine employee behaviours, although it has been recognized that avoidance and approach job crafting strategies may operate simultaneously (Mäkikangas, 2018). The main contribution of the present study is that it advances this line of research, suggesting that approachavoidance interactions are particularly important for explaining weekly variations in employee performance and employability. Our study highlighted the role of approach-oriented job crafting as a boundary condition that determines when avoidance job crafting is unfavourable and when it is favourable for employees and organizations. In that sense, our findings offer validation to literature suggesting that successful employees nowadays need to be ambidextrous, in other words, not only able to grow and develop, but also to do so in a way that shows awareness of dangers and a concern for efficiency (Kao & Chen, 2016). Such employees can reach maximum performance (Lee et al., 2019) and become assets for their organizations (Ybema et al., 2020).
A second contribution of the present paper is that, in contrast to expectations, it uncovered different interaction patterns for work performance and for employability. Results for employability were in line with our buffering hypothesis, suggesting that avoidance crafting related negatively to employability at lower levels of approach crafting (i.e. increasing challenges), while this link was buffered at higher levels of approach job crafting. In contrast, interaction effects regarding work performance revealed a boosting rather than a buffering effect. Avoidance crafting was unrelated to performance at lower levels of approach crafting, but it related positively to performance at higher levels of approach job crafting (i.e. increasing social and structural resources). A possible interpretation may be that avoidance is more necessary for performance (i.e. in-role behavior) when combined with attempts to increase resources, and less relevant for the development and employability of employees (i.e. extra-role behaviour). Resource allocation theories propose that regulating one's attentional resources is vital in achieving performance (Beal et al., 2005). In other words, to do their job well, employees need to focus on the essential and pay less attention to the non-essential. However, when it comes to becoming flexible, mobile, and ready to take up new responsibilities (i.e. employability orientation), avoidance is naturally more harmful. One cannot explore new terrains with an avoidance mindset (Brenninkmeijer & Hekkert-Koning, 2015).
Interestingly, different types of approach-oriented job crafting moderated the relationship of avoidance crafting with performance (i.e. increasing structural resources), and employability (i.e. increasing challenging demands). This may suggest that, although expected to behave in similar ways, approach crafting strategies may play different roles for different outcomes. It could be argued that while work performance benefits more clearly from additional resources (Beal et al., 2005), employability is more clearly possible when employees are ready for new challenges (Van Dam, 2004). Another pattern was that the hypothesized interaction was significant for other-referenced performance with increasing social resources and for past-referenced performance with increasing structural resources. While we can only speculate, one may argue that to outperform others (i.e. other-referenced performance), employees need to contact others (cf. increasing social resources), while to outperform their own standards (i.e. past-referenced performance), they need to focus on their own capacities (cf. increasing structural resources).
While theory is informed by the present findings, these are also relevant for organizational practice. We suggest that a successful way to prevent the prominent negative consequences of avoidance-oriented job crafting (Lichtenthaler & Fischbach, 2019;Rudolph et al., 2017) is by simultaneously facilitating approach-oriented job crafting. Previous studies have shown that job crafting interventions are successful in guiding employees' in increasing both avoidance-and approach-oriented job crafting strategies (Gordon et al., 2018). In light of our findings, any intervention attempts should make sure to aim at increasing both types of job crafting strategies because avoidance without approach job crafting can be proven detrimental for both employees and organizations.

Limitations and avenues for future Research
This study is not without limitations. First, we used only self-report measures that can be a source of common method biases. Even if evidence for interaction effects partly counteracts this concern, it is important for future studies to replicate our findings by sampling other-ratings of work performance and employability. Another limitation is that we have only collected information about employees' employability orientation and not employability activities. According to Van Dam (2004), employability activities are employees' actual attempts to become more employable by, for example, being informed about job vacancies or actively managing their career. Employability activities could be perceived to be parallel to career crafting and would be interesting to be tested vis-à-vis job crafting strategies. Also, the scale we used to test pastreference performance asked employees to recall their past year's performance in order to evaluate their weekly performance, which leaves room for biased estimations. Despite these limitations, this is the first study that provides evidence for the interactive effects of approach-and avoidance-oriented job crafting in explaining employee weekly performance and employability orientation. Our results reveal that one way to deal with the unfavourable consequences of avoidance job crafting strategies, such as attempts to reduce one's job demands, is to combine these with efforts in making one's work more resourceful and meaningful.