Seeing the forest through different trees: A social psychological perspective of work addiction.

We live in exciting times for the scientific study of work addiction, given its increased relevance and the diverse perspectives one might take to approach this phenomenon. Simultaneously, this field does not appear to be unified as a result of several misleading myths, which are addressed by the debate paper of Griffiths et al. (2018). In response, we would like to complement this study by proposing that the construct of interest should be more precisely identified in the context of related constructs and that an integrative framework should be applied, which is able to take into account not just the micro-level characteristics (i.e., individual differences), but meso- (i.e., environmental factors) and macro-level (i.e., societal factors) ones as well.

Introduction

We welcome the debate paper of Griffiths, Demetrovics, and Atroszko (2018) in which they address key myths and issues related to the scientific study of work addiction (WA). In the following commentary, we would like to contribute to this discussion by proposing an integrative model (Figure 1) that reflects on several of the myths highlighted by Griffiths et al. (2018). In doing so, the conceptual similarities and dissimilarities are discussed between WA and several related constructs as this initial step of precise conceptualization is needed if we wished to better understand WA. Subsequently, several points addressed by Griffiths et al. (2018) are expanded to demonstrate the potential fruitfulness of seeing the forest (i.e., WA) through different trees (i.e., personality traits, norms, or economical changes).

.

The proposed integrative model

A Clearner Separation of Similar Constructs

Psychological research is sometimes characterized by overlapping constructs. As highlighted by Griffiths et al. (2018), these questions permeate the field of behavioral addictions with the simultaneous presence of similar, yet different constructs, such as WA, workaholism (Spence & Robbins, 1992), passion for work (Vallerand, 2015), and work engagement (Bakker, Schaufeli, Leiter, & Taris, 2008). While Griffiths et al. (2018) explicitly distinguished WA and workaholism (Myth 4), the presence of the other constructs could also cause confusion to researchers; fortunately, there have been some studies that addressed these questions.

First of all, Birkeland and Buch (2015) and Vallerand (2015; Curran, Hill, Appleton, Vallerand, & Standage, 2015) compared passion, workaholism, and engagement to one another and highlighted some important theoretical distinctions between the three constructs. With regard to passion and workaholism, it is worth noting that workaholics do not necessarily like, love, or enjoy the work that they are doing. In addition, the Dualistic Model of Passion (Vallerand, 2015) states that when one is passionate for work, this activity is part of one’s identity and it is perceived as being meaningful and valuable for the individual. However, this is not the case for WA. While passion is a dual concept that differentiates between the positive harmonious passion and the negative obsessive passion, WA is found predominantly negative by scientific research. Rather high correlations have been observed between different problematic behaviors and obsessive passion (e.g., Orosz, Vallerand, Bőthe, Tóth-Király, & Paskuj, 2016; Wang & Chu, 2007), suggesting that there might be fine-grained differences between the two constructs that is worth future investigations.

One potential difference between WA and other related construct might be the attitude toward work. Attitudes are positive (“I like my job”) or negative (“I hate my job”) personal evaluations toward a particular object. The detailed examination of these attitudes could reveal whether individuals who are addicted to work have positive attitudes toward certain aspects of work (e.g., “I like working together with inspiring colleagues”), but negative attitudes toward other aspects (e.g., “I hate doing the administrative parts of my job”). The positive aspects of work might be rewarding and might even strengthen WA. Such research could help not only in the distinction between obsessive passion and WA, but also can provide insights about the dynamics of WA development and maintenance. Moreover, the natural course of WA should be identified, given that the addiction process typically involves a transition from voluntary engagement with a particular behavior (i.e., passion) to a more compulsive engagement (i.e., WA), in which positive or even recreational feelings transform to rather negative feelings (Brand, Young, Laier, Wöfling, & Potenza, 2016). Overall, the close inspection of the ratio of positive versus negative aspects of work attitude might be informative regarding how obsessive passion can turn into WA.

(Over)engagement and workaholism also have similarities to one another, given that both refer to heavy work investment. Empirical evidence also showed that workaholism and work engagement are only weakly and positively correlated (Schaufeli, Taris, & Bakker, 2006; Shimazu, Schaufeli, Kamiyama, & Kawakami, 2015). Gaming and TV series watching research also demonstrated that high engagement and problematic use are qualitatively different constructs (Deleuze, Long, Liu, Maurage, & Billieux, 2018; Tóth-Király, Bőthe, Tóth-Fáber, Hága, & Orosz, 2017). Engagement is not necessarily related to negative outcomes, whereas WA is. High engagement might be interpreted as a potential precursor of addiction as it only fulfills the peripheral – instead of the core – criteria of addiction.

Finally, it has also to be reinforced that time spent with work is not a sufficient criteria of WA (Myth 9). As highlighted by Griffiths et al. (2018) and other scholars (e.g., Bőthe et al., 2018; Chak & Leung, 2004), there is often only a small-to-moderate association between time spent with the activity and behavioral addiction. Future research may focus on quantity of work from the perspective of comparing individuals with high work engagement and WA. Overall, although sharing a number of similarities, these constructs nevertheless appear to be distinct, and researchers should put more emphasis on the precise identification of the construct of their interest as the initial step of research.

Personality Factors, Situational Factors, or Both?

With the precise identification of the construct of interest, we should turn our attention to the investigation of relevant correlates, which may stem from three sources: micro-level, meso-level, and macro-level. Upon reviewing the literature, one can see that previous studies predominantly investigated associations between WA and micro-level or individual variables such as, for example, Big Five personality traits (e.g., Andreassen, Hetland, & Pallesen, 2010), obsessive–compulsive personality (Mudrack, 2004), or self-esteem (Burke, 2004) (see the meta-analytic study of Clark, Michel, Zhdanova, Pui, & Baltes, 2016). These studies support the notion that there are some individual differences that might predispose one toward WA. Still, there is an additional micro-level characteristic that received less scientific attention from the perspective of WA: work motivations.

Motivations could also have important associations with WA by answering the simple questions of “Why do I work?” or “Why would I work hard?” (Gagné & Deci, 2005; Gagné et al., 2015). Previous studies highlighted the influential role of work motivations in relation to burnout (Fernet, Gagné, & Austin, 2010), organizational commitment (Howard, Gagné, Morin, & Forest, 2018), or job satisfaction (Howard, Gagné, Morin, & Van den Broeck, 2016). Taking the perspective of Self-Determination Theory (Ryan & Deci, 2017), a macro-theory of human motivation and development, one might work hard for the inherent enjoyment derived from the activity and others might work hard to achieve certain financial and social gains. While the relevance of motivations has been highlighted within the work setting, less scientific attention has been paid to their relations with WA. Some initial findings suggest that these motivations might not only predict WA, but they might also facilitate the separation of WA and work engagement (Stoeber, Davis, & Townley, 2013; van Beek, Hu, Schaufeli, Taris, & Schreurs, 2012). On the whole, we believe that focusing on attitudes and motivations might shed light on WA from different perspectives.

Motivations themselves might be influenced by many factors of which task significance is particularly relevant as it refers to the subjective judgments that work has a positive effect on others (Grant et al., 2007; Hackman & Oldham, 1976). When workers believe that, through their work, they contribute to the well-being of others, they are more likely to experience their work as being meaningful and valuable, subsequently becoming more involved in work and invest additional time and energy in it (Grant, 2008). In addition, perceived social impact (i.e., own actions contribute to others’ lives; Grant et al., 2007) and perceived social value (i.e., own actions are appreciated by others; Leary & Baumeister, 2000) could also contribute to overengagement with work or WA. This way, workers’ basic psychological needs of autonomy, competence, and relatedness (Ryan & Deci, 2017) are also satisfied which, in turn, might further increase intrinsic work motivation.

Beyond the micro-level characteristics, meso-level attributes or environmental factors (e.g., workplace climate or relationship dynamics between colleagues) could also play an important role in WA. Of major interest are social norms as they play an important role in other problematic behaviors, such as heavy drinking (Neighbors, Lee, Lewis, Fossos, & Larimer, 2007), pathological Internet use (Liu, Fang, Deng, & Zhang, 2012), or problematic Facebook use (Marino et al., 2016). Workplace norms are mental representations of appropriate behaviors at the workplace. While many norms could be present in a workplace, three salient ones could be highlighted: (a) how much work is displayed in the relevant group (descriptive norms), (b) how much the relevant group approves work (prescriptive norms), and (c) information about how fellow employees behavior is changing over time (dynamic norms). Imagine that a graduate student gets a job in a company. The work group arrives at 9 a.m. and leaves at 8 p.m. (descriptive norm); his direct supervisor expects the student to work with full concentration and employees can have only short lunch breaks (prescriptive norm) and this student heard that in the past few months his immediate colleagues tend to spend more and more time with work (dynamic norm; Sparkman & Walton, 2017). In this context, the dutiful young employee who has just started the job might develop WA more easily.

Although not always in the focus of behavioral addiction research, there are some studies that support the importance of meso-level investigations, such as increased job demands (e.g., Andreassen et al., 2017), lower job resources (Molino, Bakker, & Ghislieri, 2016), competitive climate (Keller, Spurk, Baumeler, & Hirschi, 2016), managerial support (Mazzetti, Vignoli, Schaufeli, & Guglielmi, 2017), and job control (Schaufeli, Taris, & van Rhenen, 2008), which are positively related to WA. For these reasons, WA research might benefit from exploring not only the micro-level characteristics, but the work environment as well as other attributes can have differentiated effects on employees with different personality traits.

Finally, macro-level or societal characteristics refer to the global aspects of one’s environment that might include the effects of culture, economic situation, or average income. There has been an expansion of economic activity on the globe, which makes it possible to work in other countries. Regardless of crossing actual or virtual borders, macro-level characteristics are likely to influence one’s engagement with work, even if these effects are not that salient on a micro-level. For instance, one of the most common differentiations is between individualistic and collectivistic cultures (Hofstede, 2001; Hofstede, Hofstede, & Minkov, 2010). Individualistic cultures (typically Western countries) are characterized by individualistic values, such as self-fulfillment and personal autonomy; thus, the focus is on the individual. On the other hand, in collectivist cultures (typically Eastern countries), the group is in the focus instead of the individual; thus, interdependence and the interest of the group is prioritized (which can lead to stronger effect of the norms). Although work plays an important role in everyone’s lives (regardless of culture), the quality of one’s life has a higher importance than economic growth in individualistic cultures relative to collectivistic ones (Hu et al., 2014). This is also reinforced by the notion that people work harder in societies where survival values are more important than self-expression values (Snir & Harpaz, 2009). Interestingly, a couple of initial studies supported these hypotheses (Hu et al., 2014; Snir & Harpaz, 2009).

Conclusions

One last point that we would like to conclude is that the different levels in the aforementioned model do not function independently; in fact, these are likely to influence one another, akin to Vallerand’s (1997) hierarchical model of motivation. Economical changes (i.e., a financial crisis) are likely to influence one’s workplace (i.e., the supervisor requires the employees to work overtime), which is likely to be a fuel for the individual’s characteristics toward being more and more engaged with work (i.e., because it is important for the individual to do a thorough work). Consequently, although the effect of each level of generality is predictive of WA, their unique interactions might reveal interesting findings. Thus, in line with prior calls for a more comprehensive approach (Griffiths, 2005; Griffiths & Karanika-Murray, 2012) and for the avoidance of overpathologization (Billieux, Schimmenti, Khazaal, Maurage, & Heeren, 2015) future studies should put emphasis on attitudinal dimensions, situative, value and cultural factors, and their interaction with individual factors in investigating the causes and consequences of WA.