PL refers to a leadership style that focuses on skillfully managing the inherent paradoxes within leadership and/or organizations ( Julmi, 2021). It is characterized by leader actions that appear to be in competition with each other yet are interconnected both concurrently and across time ( Sparr et al., 2022). In other words, paradoxical leaders are able to balance competing demands and find creative solutions to complex problems. PL is envisioned as a developing viewpoint aimed at comprehending how leaders assist followers in navigating paradoxes ( Zhang and Han, 2019). PL is a relatively new concept in leadership research, and it has attracted increasing attention in recent years. One of the key features of PL is handling paradoxes and tensions inherent in leadership and organizations by using seemingly competing yet interrelated behaviors to make choices that do justice to both sides of the coin. It can be used as a sense giving tool for fostering readiness for change and performance oriented towards change, and it involves the ability to enact these behaviors simultaneously and over time ( Zhang et al., 2015; Miron-Spektor et al., 2018; Sparr et al., 2022), some of the paradoxes that paradoxical leaders may need to manage include the following: Balancing accountability with rapport: Leaders must find a equilibrium between upholding accountability and nurturing positive relationships. Prioritizing rapport without enforcing accountability can hinder achieving peak performance ( Boemelburg et al., 2023). Balancing autonomy with monitoring: Leaders must allow their teams to have autonomy, yet they must also monitor their employees ( Julmi, 2021). Balancing control with autonomy: Leaders needs to balance the need for control and direction with the need for empowerment and autonomy ( Julmi, 2021).

The majority of studies on PL have concentrated on its favorable impact within organizations, leading to enhancements in various areas such as performance, creativity, commitment, competitiveness, workplace environment, and career success, as demonstrated by different researchers ( Amason, 1996; Smith, 2014; Knight and Harvey, 2015; Knight and Paroutis, 2017; Derksen et al., 2019). Moreover, studies have demonstrated that PL yields favorable outcomes for employees, including enhanced work attitudes, heightened work engagement, improved work role performance, increased creativity, and more innovative behaviors, as documented by multiple researchers ( Zhang et al., 2015, 2022a; Garg, 2016; Fürstenberg et al., 2021; Yang et al., 2021). In essence, due to its effectiveness in navigating the complex environments faced by contemporary organizations, PL has widely been regarded as a potent leadership style ( Zhang et al., 2015; Miron-Spektor et al., 2018). The minority of research on PL has studied its influence on employee Proactive Work Behavior (PWB). In our paper, we suggest a positive relationship between PL and PWB based on its beneficial impact on different aspects of employees work attitudes and performance.

PWB can be defined as “the act of taking initiative to improve current circumstances or create new ones, challenging the status quo instead of passively adapting to present conditions ( Bilal et al., 2021). It encompasses self-starting work behavior, implementing changes, and taking action to achieve goals ( Parker et al., 2010). Proactive behavior refers to an employee taking initiative in bringing about self-directed and future-focused changes in their work environment or job responsibilities ( Cui and Li, 2021). According to this definition, proactive behavior is characterized by three characteristics: “self-initiation, future focus, and change orientation”. Grant and Ashford (2008) illustrated that when employees are proactive at work, they Think, observe, strategize, fore-cast, and proactively act. According to Urbach ( Urbach et al., 2021), the PWBs are differentiated by various types such as personal initiative, taking charge, voice, and individual innovation. Essentially, research has consistently revealed significant correlations between PWBs and individual characteristics, such as cognitive ability and self-efficacy, as well as job characteristics, such as job control, and organizational factors, such as leadership that promotes transformation and an organizational environment that encourages change and forgives mistake ( Fay and Hüttges, 2017; Strauss et al., 2017; Zacher et al., 2019).

As mentioned previously, there is no studies that examined a direct relationship between PL and PWB in the field of tourism and hospitality, but there are some studies that tested the rapport between PL and PWB indirectly and in other fields other than tourism and hospitality. some of the relevant papers are as follows. Sparr et al. (2022) have proposed that PL improves employee task performance by strengthening their proactivity. A study by Kundi et al. (2023) have indicated that PL exerts both immediate and mediated impacts on employee performance. This encompasses job performance within designated roles, alongside supplementary behaviors like innovative work conduct and expressing opinions. These effects are predominantly channeled through employee work engagement. A study by He and Yun (2022) has suggested that PL significantly contributes to positive outcomes in terms of unethical pro-supervisor behavior. The connection between PL and immoral pro-supervisor actions is partly intermediated by the supervisor-subordinate dynamic. Furthermore, follower mindfulness assumes a moderating function in the association between PL and supervisor-subordinate interaction. Bilal et al. (2021) conducted a study using the self-determination theory to test the extent to which proactive personality moderates the relationship between work un-certainty and proactive work behavior, as well as strengthens the relationship between entrepreneurial leadership and psychological well-being (PWB). Bilal et al. research findings indicated that entrepreneurial leadership initially diminished work-related uncertainties, leading to subsequent enhancements in employees’ proactive work conduct. The outcomes also underscored that the influence of entrepreneurial leadership on PWB was mediated by work uncertainty. Additionally, the connection between work uncertainty and proactive work behavior was influenced by proactive personality, with a significant correlation emerging only when proactive personality traits were less pronounced. Furthermore, through moderated mediation analysis, it was evident that individuals with lower levels of proactivity placed more confidence in the proactive nature of entrepreneurial leadership in comparison to their highly proactive counterparts within the workplace. Boonyarit (2023) carried out an investigation and revealed that self-leadership demonstrates a positive correlation with two forms of PWB: assuming initiative and personal innovation. Furthermore, self-leadership exhibited favorable connections with psychological empowerment and the breadth of roles one believes they can effectively perform (RBSE).

Based on what was previously mentioned, this study could provide evidence that demonstrates the link between PL and PWB. Accordingly, researchers would like to propose that H1.

PL is positively associated with PWB.

Psychological contract theory is built upon the principles of social exchange theory, which is a social psychological concept that explains how individuals engage in social interactions based on mutual benefits and obligations ( Aranda et al., 2018). Psychological contract fulfilment encompasses an individual’s personal convictions regarding the mutual commitments and advantages forged within a give-and-take association, such as employment within an organizational context ( Topa et al., 2022; Hammouri et al., 2022; Rodwell et al., 2015). According to Coyle-Shapiro et al. (2019), it involves shared beliefs, perceptions, and implicit obligations between an employer and an employee, which shape the practicality of the work to be performed and establish the dynamics of the relationship. Hafez Mahmoud Gharib and Ahmed Khairy (2019) declared that the psychological contract is an informal agreement that is not typically documented in writing. Its contents and obligations are fulfilled through employees’ interactions with their organization. Ultimately, psychological contract fulfilment is a crucial factor in the employee-employer relationship, which can influence the quality of the relationship and the employee’s outcomes.

The available literature indicates that there is a dearth of research on the association between PL and PPCF in the tourism and hospitality domain. Nevertheless, some relevant studies suggest that there could be an indirect correlation between these two constructs. For instance, Huertas-Valdivia et al. (2019) investigated the impact of various leadership styles on the job performance of hospitality employees and observed that PL exhibited a positive correlation with employees’ psychological empowerment. Although this research did not directly explore the relationship between PL and PPCF, it implies that PL could potentially influence employees’ attitudes and perceptions in a favourable manner. The findings of a study by He and Yun (2022) emphasized that PL has a direct positive effect on Employee Loyalty, and this effect is partially explained by the role of job satisfaction in mediating the relationship between paradoxical leadership and employee loyalty. While this study did not investigate the link between PL and PPCF directly, being Employee loyalty can be considered a component of psychological contract fulfilment as explained by Hart and Thompson (2007) may imply that there is an indirect correlation between PL and PPCF. Oorschot et al. (2021) suggest that leaders, and leadership styles in particular, have the ability to impact the psychological contracts of employees. Given that PL is widely recognized as an effective style of leadership styles navigating the complex environments encountered by modern organizations ( Zhang et al., 2021), this could suggest a potential correlation between PL and PPCF.

Considering the significance of both PL and PPCF in the service industry, it would be beneficial to explore the potential relationship between these constructs in the context of tourism and hospitality. The tourism and hospitality sector are renowned for its exceptional quality of service and consumer satisfaction, and effective leadership is crucial for managing employee expectations and obligations in this industry. Therefore, research in this field could shed light on the role of PL in enhancing PPCF in tourism and hospitality, which could contribute to improving job satisfaction, organizational commitment, and performance in this sector.

As previously noted, there is a gap in research regarding the specific relationship between PL and PPCF in the tourism and hospitality industry. Social Exchange Theory could be applied to address this gap. According to Ahmad et al. (2023), this theory posits that individuals are motivated by their expectations of rewards and outcomes. In the context of PL and PPCF in tourism and hospitality, this theory could be used to examine how PL behaviours influence employee expectations of rewards and outcomes, and how these expectations impact PPCF. Based on this discussion, the following hypothesis is proposed H2.

PL is positively associated with PPCF.

The search results reveal that there is a scarcity of research that explicitly investigates the connection PPCF and PWB within the tourism and hospitality industry. Nonetheless, a very few studies have indirectly highlighted the relationship between PPCF and work behaviour more broadly, as well as in other sectors. Here are some relevant academic papers:

One study by Noble-Nkrumah et al. (2022) investigated the interplay between job autonomy and employee trust in relation to the connection between PPCF and work behaviour within international non-governmental organizations. The research revealed a favourable correlation between PPCF and work behaviour, with job autonomy and employee trust functioning as moderators in enhancing this positive association. According to Hetet et al. (2023) and Bilal et al. (2021), work behaviour can be categorized as either in-role behaviour or extra-role behaviour, with proactive work behaviour falling under the latter category. Proactive work behaviour entails self-initiated and forward-looking actions aimed at enhancing existing conditions or establishing new ones. With this in mind, it can be deduced that there exists an indirectly positive correlation between PPCF and proactive PWB.

Another study by Yu (2022) investigated the effects of PPCF on employees’ work-related attitudes and behaviours (specifically work engagement, innate motivation, and emotional commitment) amidst the COVID-19 pandemic. The study found that psychological contract fulfilment positively influenced employees’ work behaviour during the pandemic. While this study does not directly refer to the relationship between PPCF and PWB, it may be concluded that psychological contract fulfilment positively influences proactive work behaviour, given that proactive work behaviour is a specific type of work behaviour.

Therefore, while there may not be specific studies directly examining the relationship between PPCF and PWB in the tourism and hospitality business, the existing research attempts to bridge the gap and explore the relationship between PPCF and PWB more comprehensively. These points lead to the following hypothesis: H3.

PPCF is positively associated with PWB.

Previous studies in the tourism and hospitality business have not explored the relationship between PPCF, PL, and PWB. However, based on the information provided about PL, PWB, and PPCF, the connection between PL and employee proactive work behaviour (PWB) can be elucidated through the mediating effect of PPCF. PL can have either positive or negative impacts on employee behaviour. On the one hand, PL can encourage employees to embrace ambiguity, think critically, and challenge traditional norms, which can foster PWB. On the other hand, the contradictory nature of its leadership practices may create tension or confusion. In this relationship, PPCF plays a mediating role. When employees perceive that their organization fulfils the promises and obligations outlined in their psychological contract, they develop a sense of trust and reciprocity. This builds their commitment to the organization and motivation to engage in proactive behaviours. Thus, higher levels of PPCF strengthen the positive relationship between PL and PWB. In essence, PPCF explains how PL influences employee behaviour and encourages proactive work. The inclusion of PPCF as a mediator between PL and PWB provides a comprehensive understanding of how leadership behaviours affect employees’ proactive work behaviours. It highlights the importance of organizational fulfilment of psychological contracts in promoting desirable employee outcomes. Based on the previous discussion, a possible hypothesis could be. H4.

PPCF mediates the association between PL and PWB.

Based on a review of previous literature and the study hypotheses that were proposed, a model for the study was designed, as shown in Figure 1.

To examine hypotheses, we gathered data from employees of hotels in Sharm El-Sheikh and Tourism companies and travel agencies in Cairo, Egypt, using a questionnaire tool. Because they had sufficient understanding to answer the survey questions, employees with at least 3-years of experience were qualified to participate. We instructed our postgraduate colleagues enrolled in our colleges and working in the study’s sample hotels and companies to connect with their HRMs to help us complete the questionnaire. We conducted “multi-wave and multi-source surveys” using the questionnaires, which reduced typical technique bias ( Podsakoff et al., 2012). A time-lagged study process was operated, with two data waves collected from our sample employees during two-month intervals. Data were assembled from February to April 2023, performing “convenience sampling and drop-and-collect” methods. Employees measured paradoxical leadership (PL) and demographic data in the first wave survey. The second was conducted one months later, and the Psychological Contract Fulfilment (PPCF) and employee proactive work behaviour (PWB) variables were measured. In both two waves of surveys, 500 questionnaire forms were distributed. After eliminating the incompetent responses, 338 were considered, with an efficient reply rate of 67.6%. The study sample consisted of 62 females (18.3%) and 276 males (81.7%). The participants’ ages ranged mostly between 24 and 52. The full set of data is available on Zenodo ( https://zenodo.org/records/11400950) and is freely available for non-commercial users ( Zenodo, 2023).

The literature was consulted and used to create questionnaires for all variables. All variables were assessed using a 5-point Likert scale. The 22 items suggested by Zhang et al. (2015) were used to operationalize the PL variable. The PL scale items are broken down into five dimensions, 5-items for each of “Treating subordinates uniformly while allowing individualization,” “Combining self-centeredness with other-centeredness,” and 4-items for each of “Maintaining decision control while allowing for autonomy,” “Enforcing work requirements while allowing flexibility,” and “Maintaining both distance and closeness.” The PAB was measured by a 13-item suggested by Parker and Collins (2010) distributed on four dimensions; 3 items for each of “Taking charge,” “Problem prevention,” and “Individual innovation,” and four items for “Voice.” For the PPCF, seven items were used from the study of Tseng and Wu (2017). Additionally, to increase their clarity and comprehension, the survey questions were transcriptions and edited. Nineteen individuals, including nine academics and ten business executives, tested the validity of the survey to verify its reliability. There was no alteration made to the survey’s content during these procedures.

Partial least squares (PLS) were used by operating the smartPLS version 4.0 ( https://www.smartpls.com/) to perform structural equation modelling (SEM) to test the suggested model (the software offer a trail free full version for one month). The PLS technique is fitting when the main investigation’s aim is to foresee one or more tested variables rather than to validate a formerly established theoretical model ( Hair et al., 2017). Therefore, this method is suitable for the current study because it tests relationships between the PL and PAB, considering the mediating effect of the PPCF. Additionally, the PLS is a productive tool across a broader range of sample sizes, and also it helps to construct a more complex and advanced model with fewer data constraints ( Hair et al., 2011). The PLS-SEM processing, according to Leguina (2015), passes across two steps, namely, (2) measurement modelling or outer model assessment and (1) structural modelling or inner model assessment.

The measuring model assesses the “convergent validity” (CV) and “discriminant validity” (DV) to determine the quality of the data. The CV evaluates the connections between the study’s indicators by performing “Cronbach’s alpha” and must exceed the threshold of 0.50 ( Leontitsis and Pagge, 2007), “composite reliability (CR)”, which is greater than 0.60 ( Fornell and Larcker, 1981), “Average Variance Extracted” (AVE), which should not be less than 0.50 ( Hair et al., 2011), and “factor loading”, also, preferably greater than 0.50 ( Afthanorhan, 2013) as assessment standards. Additionally, DV claims that when using various methods to measure other factors, the measured results must be distinguishable. According to Fornell and Larcker (1981), If a factor’s AVE is greater than the association between that factor and other factors in the suggested model, the factor meets the statistical requirements for “discriminant validity." In addition, a number of scholars evaluated the “Heterotrait-Monotriat” ratio of correlation (HTMT) to determine the “discriminant validity” in response to the many criticisms made of “Fornell and Lacker’s criterion.” ( Gold et al., 2001).

Table 1’s CV values display that every specified minimum and/or maximum level was accepted, proving the validity of the proposed outer model. Also, the values of AVE and HTMT, which are portrayed in Table 3 and Table 4, respectively, also met the recommended norms, proving that their “discriminant validity” is adequate. Also, Table 2 shows that “an item’s loading within its factors is greater than any cross-loadings with other factors, guaranteeing the DV”.

The “variance inflation factor” (VIF) is utilised in this study to examine the collinearity regards to ascertain the presence of collinearity issues among variables and to stop the impact of variables on the offered model’s contribution. According to Hair et al. (2011), reducing multicollinearity is unnecessary for VIF values under 5. Additionally, the “coefficient of determination” (R2) and “Stone-Geisser’s” (Q2) were operated to gauge how satisfactorily the regression model explained the data. R2 results of 0.20 are regarded as adequate level in behavioural studies ( Hair et al., 2011). The Q2 scores also met the suggested norm value of 0.0 ( Hair et al., 2014). The VIF, R2, and Q2 findings are presented in Table 5.

Unlike “covariance-based structural equation modelling” (CBSEM), the PLS approach lacks a range of statistical parameters for validating models, such as X2 and other indicators of model fit ( Henseler and Sarstedt, 2013). To solve this problem, “Goodness of Fit” (GoF) has been proposed as a successful model validation technique ( Tenenhaus et al., 2005). According to Mital et al. (2018) and Tenenhaus et al. (2005), the GoF can be calculated by taking the value of the square root of the average variance extracted multiplied by the R square value.

As per Tenenhaus et al. (2005), GoF scores of 0.1, 0.25, and 0.36, respectively, denote modest, medium, and high GoF. The GoF value for the model constructed in this study is 0.558, indicating a notably high GoF index. The “Standardised Root Mean Square Residual” (SRMR), which compares variances in observed correlations, was used to further assess the model’s validity. Model fit is considered acceptable when the SRMR value is less than 0.1 ( Hu and Bentler, 1998). The proposed model’s SRMR value is 0.095, which indicates a satisfied model fit.

Upon validating both the external and internal models, the study progressed to judge the formulated hypotheses. Utilizing SmartPLS v.4, a bootstrapping process with 5000 iterations was conducted to calculate (β) the regression weights, t-scores, and P-values associated with the direct, the indirect, and the mediating paths. The analysis encompassed four main hypotheses, comprising of 3 direct hypotheses and 1 mediating hypothesis, as presented in Table 6.

Data portrayed in Table 6 and Figure 2 demonstrated that the PL has a positive significant influence on PAB (β = 0.293, t = 3.388, p<.001) and PPCF (β = 0.680, t = 13.328, p< .001), confirming H1 and H2. The results further supported H3 by showing that the PPCF variable positively influenced PAB at β= 0.420, t=5.637, and p.001. Furthermore, at β= 0.286, t = 4.800, and p< .001, the PPCF variable mediated the link between PL and PAB, supporting H4.