The effects of shaping employee work engagement and job satisfaction on company performance results: the mediating role of HRM outcomes in Central European MNCs

2.1 The general assumptions about the phenomena under study

EWE and JS have diverse definitions in the literature (Kahn, 1990, p. 694). EWE refers to active psychological engagement in roles, characterized by vigor, dedication and absorption (Schaufeli, Salanova, Gonzalez-Roma, & Bakker, 2002, p. 74). It includes commitment, attachment, involvement, effort, attitude, energy and attendance, impacting employee performance and organizational success (Shuck, Osam, Zigarmi, & Nimon, 2017). JS, defined by Locke, reflects overall job enjoyment and liking (Locke, 1976, p. 1304). It is influenced by needs and temporary events and is less predictive of business outcomes compared to EWE (Robertson-Smith & Markwick, 2009). Engaged employees may experience occasional dissatisfaction but remain committed due to higher values (Stor, 2023). Highly engaged employees consistently exceed expectations (Harter et al., 2013). EWE represents a multidimensional motivational state involving diverse personal resources invested in fulfilling organizational roles (Saks, Gruman, & Zhang, 2022), providing a holistic view of employee motivation.

In this article, EWE is defined as a specific attitude of an employee and the resulting behavior characterized by identification with organizational goals and values, taking actions consistent with the organization’s interests, willingness to belong to the organization, readiness to act, giving high rank to the company’s interests, undertaking activities that go beyond the standards, with simultaneous readiness for responsibility in the conditions of independent action (Stor & Haromszeki, 2020, p. 54). As far as job satisfaction is concerned, it is understood as an emotional state resulting from the employee’s perception of his or her own work as giving pleasure and providing what an employee considers important.

Shaping EWE and JS in organizations is crucial for employee and company performance. Both are influenced by employees’ organizational experiences (Plaskoff, 2017, p. 137), and recent research highlights changing employee expectations. Moving from a performance-centered to a human-centered approach is essential. Studies show that solely prioritizing results harms employee well-being and organizational performance (Boccoli, Gastaldi, & Corso, 2023). Therefore, in this article, it is assumed that SEWE&JS involves activities that are intended to stimulate employee engagement and job satisfaction in such a way as to achieve the organization’s goals and ensure its success by creating friendly working conditions (Stor, 2023, p. 97). In addition, it was also assumed that SEWE&JS is one of the essential subfunctions of HRM (c.f. Juchnowicz, 2014; Kim & LePine, 2019; Stor, 2023).

2.2 SEWE&JS and company performance results

Research shows that SEWE&JS positively impact organizational performance, with EWE and JS serving as predictors of performance (Kessler, Lucianetti, Pindek, Zhu, & Spector, 2020). Positive effects are observed in employee turnover, safety incidents, absenteeism, retention, performance, productivity, company productivity, quality standards, customer satisfaction, financial performance, profitability and organizational development (Schaufeli et al., 2002; Harter et al., 2013; Shuck et al., 2017; Sun & Bunchapattanasakda, 2019; Motyka, 2018; Taipale, Selander, Anttila, & Nätti, 2011; Farndale, 2017). These effects apply to both domestic and multinational companies.

For the purposes of the research presented in this article, the company’s performance results were divided into four categories: the scope of HRM, finance, innovativeness and quality. It should also be clarified that the concept of company performance results is used, which is not the same as company performance. It is based on the assumption that performance result means the final outcome to which certain activities lead. In contrast, performance alone is the act of executing certain activities (Stor, 2023, p. 43). Hence, the article focuses on the ultimate measure of the results of activities at a particular point in time, and not on the activities themselves.

As a result of the literature review, one main hypothesis and four auxiliary hypotheses have been formulated to describe the relationships between SEWE&JS and company performance results as follows:

2.3 The mediating role of HRM outcomes

In accordance with the assumptions mentioned earlier in the article, SEWE&JS is treated as one of the HRM subfunctions and, therefore, does not function in a vacuum but in connections with other HRM subfunctions (c.f. Juchnowicz, 2014; Stor, 2021). In practice, this may mean that SEWE&JS can also influence company performance results indirectly through HRM outcomes (Garengo et al., 2022). As a side note, it should be explained that in order to maintain terminological logic, HRM outcomes are identified here with the company’s performance results in HRM. However, the concept of outcomes instead of results is used when HRM takes the position of mediator. Regarding this indirect impact of SEWE&JS through HRM, there has been a long-standing consensus in the literature that the focus should be on HRM systems rather than on individual subfunctions, as the effects of HRM subfunctions may depend on other HRM subfunctions (Boon et al., 2019) and their results (Saks et al., 2022). The same applies to EWE and JS, which can be shaped at various organizational levels in connection with different HRM subfunctions that can create a positive and engaging working environment (Sivapragasam & Raya, 2018). Certain HRM services are considered resources for promoting engagement (Lee et al., 2020), and engaging leaders, including HR professionals and direct superiors, play crucial roles in these processes (Mazzetti & Schaufeli, 2022). Research across countries and organizations highlights that targeted HRM activities can enhance EE and JS levels (Knight, Patterson, & Dawson, 2017).

Against the background of the literature findings, one main hypothesis and three auxiliary hypotheses have been formulated, describing the mediating role of HRM outcomes in the relationship between SEWE&JS and company performance results as follows:

2.4 The pre-pandemic and pandemic context in MNCs

As indicated in the introduction, the research problem focuses on the relationship between selected variables in four specific contexts: pre-pandemic and pandemic periods, in HQs of MNCs and in their foreign subsidiaries. Similar to the Great Recession, the pandemic had a global impact, requiring organizations to adapt and address falling demand (Stor, 2011; Kim & Ployhart, 2014). The pandemic affected both the economy and people’s lives, emphasizing the need for mental health care, resilience and employee support (Mańkowski, Szmeter-Jarosz, & Jezierski, 2022; Donovan, 2022). Effective SEWE&JS during the pandemic relied on an appropriate organizational climate and support from the organization and direct superiors (Zeidan & Itani, 2020; Reinwald, Zimmermann, & Kunze, 2021). Employer branding, new compensation, empowerment, work autonomy, communication, training, incentives and technology support were crucial to boosting EWE and JS (Agarwal, Arya, & Bhasin, 2022; Pass & Ridgway, 2022). Overall, during the pandemic, SEWE&JS activities were closely linked with other HRM subfunctions.

All this means that during the pandemic, the relationships between SEWE&JS and other HRM subfunctions may be more intense than before the pandemic. At the same time, the aggregate results of these subfunctions may be a stronger mediator of the relationships between SEWE&JS and the organization’s performance results. For this reason, one main hypothesis and three auxiliary hypotheses have been formulated to describe the expected differences in the mediating role of the HRM outcomes in the pre-pandemic and pandemic periods as follows:

As already mentioned, in the research, these hypotheses will be tested separately for the HQs of MNCs and their foreign subsidiaries, which is consistent with the postulated differentiation of specific contexts (Meyer et al., 2011; Farndale, 2017) that may be shaped differently for SEWE&JS due to local nationalities, cultures (Shuck et al., 2021), economy (Dillard & Osam, 2021), level of a country’s development, law (Aftab et al., 2022) and many other factors.

3.1 The research sample, measures and data collection method

To emphasize the essence of the Central European research context, this part of the article deliberately has methodics in the title because it concerns only the components of the research process presented in this article and does not concern the science of research methods, which is what methodology deals with (see more: Stor, 2023, p. 27).

The empirical research took place at the end of Q1 2022 and covered 200 MNCs which were nonfinancial business entities, headquartered in a Central European country (Poland) and possessing foreign subsidiaries. Taking the official data of the Polish Central Statistical Office as a point of reference, it can be stated that the research sample constituted about 11% of the general population and made its representative sample in about 80% (Statistics Poland, 2022). Considering the type of economic activity according to NACE (the Statistical Classification of Economic Activities in the European Community), the size of the enterprise and the type of foreign direct investment (FDI), it can be concluded that the structure of the research sample largely corresponded to the general population. These characteristics were important insofar as they were used in the purposeful sample selection. It is worth noting that the MNCs under study employed a total of 76,740 employees worldwide and had over 400 foreign subsidiaries located in approximately 30 countries.

In the purposive sampling, two other criteria were applied. Namely, both the HQs and their foreign subsidiaries had to be predominantly owned by the Polish capital. Based on a review of research by other authors, it was assumed that the size of equity capital largely determines management activities in organizations. Therefore, its relative uniformity in the research sample may increase the comparability of the HQs and foreign subsidiaries. Thanks to this procedure, the context in the surveyed organizations can be perceived by the respondents in a similar way, and the measures used by the HQs to assess the results of foreign subsidiaries in different countries are understood in a similar way (cf. Farndale, 2017; Schlägel & Sarstedt, 2016). This is important in comparative research settled on benchmarking, and such a solution was used in the presented research. Another of the aforementioned criteria concerned the minimum period of operation on the market, which was four years, both in relation to the HQs and their foreign subsidiaries. It was assumed that this is the minimum time interval needed to identify causal relationships between variables of interest during the pre-pandemic and pandemic periods.

The research method was CATI (computer-aided telephone interview) based on a structured questionnaire. The selection of the respondents was also of a purposive nature because they had to possess knowledge about company performance and HRM and understand the measures applied within them (Richard, Devinney, Yip, & Johnson, 2009). As a result, the sample of personal respondents consisted of business owners (1%), managing directors/CEOs (2%), HR directors (50%), HR managers (46%) and HR business partners (1%). They provided information on two periods: (1) the pre-pandemic period of 2018–2019, and (2) the pandemic from the beginning of 2020 to the end of the first quarter of 2022, in which the interview was conducted.

Four variables were of research interest:

1. company performance results – evaluated by the respondents in four categories (i.e. in finance, innovativeness, quality and HRM) in a benchmarking process against the companies of similar business profiles;

2. the advancement level of SEWE&JS – assessed by the respondents in a benchmarking process by comparison to the best market practices;

3. the significance level of SEWE&JS – meant how important the activities and processes within the scope of this HRM subfunction were for the company performance;

4. the efficiency of employees’ performance – understood as employees’ work outcomes and assessed against the established standards in the meaning of KPIs (key performance indicators).

Each of these variables was evaluated on a 5-point scale, separately at the HQs and its biggest foreign subsidiary. The scales are presented in the lower part of Table 1.

3.2 The formulas applied for data conversion and statistical methods of data analysis

Before analyzing the collected research data, they were recalculated using unique formulas designed specifically for the research. The idea for such a conversion was born as a result of deeper considerations of the nature of the studied phenomena. Specifically, while numerous studies demonstrate the impact of HRM practices on employee performance results, which, in turn, directly influence company performance results (Cascio et al., 2019; Garengo et al., 2022), some studies enable the observation of incredibly surprising phenomena at times. For instance, even though a company’s productivity is high and its employees achieve high KPIs, it may not be profitable at the same time. This may be due to intense market competition, reduced consumer demand, legal and hygiene restrictions and so forth, which could be observed during the Great Recession or the COVID-19 pandemic (Kim & Ployhart, 2014; Minbaeva & Navrbjerg, 2023). Of course, greater productivity means that human capital resources have been effectively used, because of which the company gains the ability to generate above-average returns, but their potential for usage may be constrained by external factors. However, it should be recognized that increasing productivity is a crucial way to develop slack resources. These resources can then be employed to expand business operations, look into opportunities for new product innovations and get new customers. As a result, the company is able to pursue additional revenue-generating opportunities, thanks to a more efficient workforce (Kim & Ployhart, 2014; Cascio et al., 2019). Therefore, from an economic and management standpoint, empirical research appears to benefit from focusing more on the effectiveness and efficiency of the actions taken and the results achieved. However, efficiency and effectiveness are two distinct concepts. While efficiency refers to rates of resource utilization in attaining objectives (Rogers & Wright, 1998) and is understood as the ratio of output to input (Stor, 2012), effectiveness is a goal-oriented measure (Ostroff & Schmitt, 1993) that concerns the extent or degree to which desired goals are attained (Stor, 2012).

Given that there are numerous ways to develop and use slack HRM practices and slack human capital resources to improve the business’s operations, the aforementioned concerns have cast doubt on how to accurately capture the actual relationships between the variables under study. The author of this article proposes an innovative solution in response to the problem. The author of this article has previously applied such an approach to research on employee performance appraisal (Stor, 2023) and talent management (Stor, 2023). To capture the actual relationships between the performance results of employees and HRM activities with the organization’s performance results in the analysis process, the raw data of the variables were adjusted by the efficiency index (EI). Hence, the adjusted values of the SEWE&JS variable were calculated using formula (1), which expresses the ratio of the advancement level of SEWE&JS to the efficiency of employees measured by employee key performance indicators used in companies:

Where:

• EI_SEWE&JS – efficiency index of shaping employee work engagement and job satisfaction

• AL_SEWE&JS – advancement level of shaping employee work engagement and job satisfaction

• EKPI – employee key performance indicators

The adjusted values of the company performance results were calculated according to formula (2), which measures the ratio of the company performance results to the efficiency of employees measured by employee key performance indicators used in companies:

• EISCPR – efficiency index of company performance results

• (x) – one of the four categories of the company performance results, respectively, in human resources management (HRM), finance (F), innovativeness (I) and quality (Q).

• CPR – company performance results

• EKPI – employee key performance indicators

Hence, the formulas for adjusting the value of individual categories (2_A – 2_D) of the company performance results are as follows:

• EIHRM – efficiency index of company performance results in HRM

• HRM – company performance results in HRM

• EKPI – employee key performance indicators

• EIF – efficiency index of company performance results in finance

• F – company performance results in finance

• EKPI – employee key performance indicators

• EII – efficiency index of company performance results in innovativeness

• I – company performance results in innovativeness

• EKPI – employee key performance indicators

• EIQ – efficiency index of company performance results in quality

• Q – company performance results in quality

• EKPI – employee key performance indicators

The collected data were analyzed based on descriptive, correlational and mediation statistical methods. First, the Kolmogorov-Smirnov (with Lilliefors correction) and Shapiro-Wilk tests of normality were used. Then, Spearman’s rank coefficient was applied to examine the relationships between the variables. A paired-samples t-test was utilized to determine the significance of differences between means of variables in the pre-pandemic and during-pandemic periods. Additionally, Fisher’s z-transformation test was employed to compare the strength of correlations across these periods. As for the potential concerns of multicollinearity, the variance inflation factor (VIF) was calculated for each variable across different contexts. In this study, all VIF values were well below the standard threshold, that is <5 (Ringle, Sarstedt, Sinkovics, & Sinkovics, 2023), suggesting that multicollinearity is not a significant concern in the models. These analyses were conducted using the TIBCO Statistica v. 14.0.0.15 software.

In the last stage of the analysis, the author used partial least squares structural equation modeling (PLS-SEM) in R environment with lavaan package v. 0.6–12 software (Rosseel, 2012; Savalei & Rosseel, 2022) to verify the research hypotheses and assess mediating effects. The choice of PLS-SEM is mainly due to the fact that it works well for distributions of non-normal variables and is the preferred method when the research goal is to develop a theory and explain the variance (Hair, Hult, Ringle, & Sarstedt, 2022), which is exactly what happens in the presented studies. In addition, PLS-SEM provides the opportunity to analyze the differences between the coefficient paths of different sets of variables (Picón-Berjoyo, Ruiz-Moreno, & Castro, 2016) and has become an almost universally used method for examining the effect of HRM practices on organizational performance (Ringle, Sarstedt, Mitchell, & Gudergan, 2020). Finally, it is worth recalling that the correlation and path analysis were performed on variable values that were adjusted by IE.

4.1 The descriptive and correlational statistics

As shown in Table 1, both in the MNCs’ HQs and their local subsidiaries, the performance results in the four studied categories were comparable to those obtained by enterprises with similar business profiles, both before and during the pandemic. However, scores in HRM, finance and quality were slightly higher before the pandemic. In comparison, innovation scores were better during the pandemic.

As for SEWE&JS, its significance to performance results was lower during the pandemic for both HQs (x¯=3.21) and local subsidiaries (x¯=3.01) compared to before the pandemic (x¯ = 3.27; x¯ = 3.12, respectively). The advancement level of SEWE&JS was also slightly lower during the pandemic for both HQs ((x¯ = 3.08) and local subsidiaries (x¯ = 3.00) compared to before the pandemic (x¯ = 3.03; x¯ = 3.18, respectively). Surprisingly, employee performance was slightly better during the pandemic for both HQs (x¯=3.24) and local subsidiaries (x¯=3.19) compared to before the pandemic (x¯ = 3.00; x¯ = 3.05, respectively).

In turn, the correlation analysis in Table 2 reveals positive correlations between all variables in both types of organizations and during both time periods. The advancement level of SEWE&JS at HQs before the pandemic showed stronger correlations with finance (r = 0.63), innovativeness (r = 0.55) and HRM (r = 0.58) compared to during the pandemic (r = 0.53; r = 0.44; r = 0.47, respectively), while its correlation with quality remained the same (r = 0.55). In foreign subsidiaries, during the pandemic, all SEWE&JS correlations with company performance in HRM (r = 0.38), finance (r = 0.40), quality (r = 0.42) and innovation (r = 0.39) were stronger compared to before the pandemic (r = 0.37; r = 0.39; r = 0.36; r = 0.27 respectively).

4.2 Mediation statistics based on PLS-SEM

The path analysis (see Tables 3–5) confirms that SEWE&JS has a direct and positive effect on company performance results in HRM, finance, innovativeness and quality. This leads to the confirmation of the auxiliary hypotheses from H1a to H1d; however, hypothesis H1c can be accepted to some extent because not all effects are statistically significant.

Furthermore, there are variations in these effects between HQs and local subsidiaries, as well as between the pre-pandemic and pandemic periods. In HQs, the effect of SEWE&JS on company performance is stronger before the pandemic, while in local subsidiaries, it is slightly stronger during the pandemic. Similar patterns are observed for finance and innovativeness, with stronger effects before the pandemic in HQs and slightly stronger effects in local subsidiaries during the pandemic. The effect of SEWE&JS on quality is significant and stronger during the pandemic in both HQs and local subsidiaries.

Overall, the main hypothesis H1 is confirmed, indicating that SEWE&JS positively influences company performance results in various areas. However, not all effects are statistically significant, highlighting some variations in the impact of SEWE&JS across different contexts and time periods.

Mediation analyses reveal that company performance results in HRM positively mediate the relationships between SEWE&JS and company performance results in finance. In HQs, this mediation effect is significant in both the pre-pandemic (β = 0.82; p < 0.001) and pandemic (β = 0.85; p < 0.001) periods. The indirect mediation effect is slightly stronger before the pandemic (αβ = 0.48; p < 0.001) compared to the pandemic period (αβ = 0.44; p < 0.001). In foreign subsidiaries, the mediation effect remains consistent across both periods (αβ = 0.32; p < 0.001). Therefore, auxiliary hypothesis H2a is confirmed, while H3a is rejected. Figure 1 provides a graphical representation of these findings.

Company performance results in HRM positively mediate the relationships between SEWE&JS and company performance results in innovativeness. In HQs, this mediation effect is significant in both the pre-pandemic (β = 0.67; p < 0.001) and pandemic (β = 0.76; p < 0.001) periods, with a consistent indirect mediation effect (αβ = 0.39; p < 0.001) across both periods. In foreign subsidiaries, the mediation effect is significant in both the pre-pandemic (β = 0.67; p < 0.001) and pandemic (β = 0.81; p < 0.001) periods, but the indirect mediation effect is significantly stronger during the pandemic (αβ = 0.31; p < 0.001) compared to pre-pandemic (αβ = 0.25; p < 0.001). Therefore, auxiliary hypothesis H2b is confirmed, while H3b is partially supported. Figure 2 illustrates these relationships.

HRM outcomes positively mediate the relationships between SEWE&JS and company performance results in quality. In HQs, this mediation effect is significant in both the pre-pandemic (β = 0.56; p < 0.001) and pandemic (β = 0.50; p < 0.001) periods, with a stronger indirect mediation effect in the pre-pandemic (αβ = 0.32; p < 0.001) compared to the pandemic (αβ = 0.26; p < 0.001) time. In foreign subsidiaries, the mediation effect is significant in both the pre-pandemic (β = 0.55; p < 0.001) and pandemic (β = 0.56; p < 0.001) periods, with a slightly stronger indirect mediation effect during the pandemic (αβ = 0.21; p < 0.001) than during pre-pandemic (αβ = 0.20; p < 0.001). Auxiliary hypothesis H2c is confirmed, while H3c is partially supported, specifically in relation to foreign subsidiaries. Figure 3 visually presents these phenomena.

Summarizing the results of mediation analyses, it can be stated that they allow for a positive verification of auxiliary hypotheses H2a, H2b and H2c, which ultimately entitles us to recognize the main hypothesis H2 as true. This is based on the findings that company performance results in HRM positively mediate the relationships between SEWE&JS and the other three categories of company performance, regardless of the type of the company (HQs or foreign subsidiaries) and the time period under consideration. As for main hypothesis H3, it can be considered only partially confirmed. This is due to the fact that none of the auxiliary hypotheses has been confirmed for both the HQs and the foreign subsidiaries. Hypothesis H3a was positively verified only for the HQs, and hypotheses H3b and H3c are true only for the foreign subsidiaries.

The explanatory capability of all the models is juxtaposed in Table 6. In general, it can be said the amount of variance explained in the company performance results ranges from weak to strong (Ringle et al., 2020; Hair et al., 2022).